Print Page | Close Window

Adventures In Transmission Line Loudspeakers

Printed From: Graham Slee Hifi System Components
Category: And the rest
Forum Name: Transducers, Speakers, Mics...
Forum Description: Interested in bi-amping, tri-amping, crossovers, speaker design, miking techniques, EQ - the list goes on
URL: https://www.hifisystemcomponents.com/forum/forum_posts.asp?TID=4097
Printed Date: 29 Mar 2024 at 3:30pm
Software Version: Web Wiz Forums 12.01 - http://www.webwizforums.com


Topic: Adventures In Transmission Line Loudspeakers
Posted By: Graham Slee
Subject: Adventures In Transmission Line Loudspeakers
Date Posted: 30 Dec 2016 at 8:24am
A long time ago I remember there were a few loudspeaker companies advertising acoustic labyrinth loudspeakers. Since then I haven't heard much about them. It seems it is much easier and cost effective for them to now make bass reflex or sealed enclosures.

The trouble with bass reflexes is they are tuned to one particular frequency so you get what's called "one note bass". This might suit some but bass is actually more tuneful, and you can't have a one note tune (can you?).

Sealed cabinets are OK I guess, but it would seem are room position sensitive, in that in some placements the room contributes more (or less) bass than in others. Some like the LS35A are panel tuned to resonate and reinforce their bass output, or so it seems.

Then there are horns of all kinds of shapes and sizes, including the frugel horn, which some here have taken a liking to. Originally a horn was a compression driver plus a flare, and horns of different frequencies were arranged with magnets in the same plane so as not to produce allignment distortion. These were often seen in old cinemas which had the room to accomodate the long length of the bass horn.

There are also open baffle designs which to me are a little like brute force. Air pressures cancel out and to get any meaningful bass require large (12 - 18") bass drivers. This is OK if you have a big enough room (I don't).

The acoustic labyrinth or transmission line loudspeaker is a different approach. In a way it is like an open baffle because output from the rear of the driver is coupled to the open air. However, this is via a long pipe whose length is a quarter wavelength of the "boost frequency", and because it is quarter wavelength its output is 90 degrees out of phase with the cone front output, and so no cancellation occurs.

In addition other frequencies either side of the "boost frequency" are near to 90 degrees and so are not cancelled either, and that differentiates it from the bass reflex in that the bass isn't as one noted.

The "boost frequency" is usually set to the resonant frequency of the driver. It is far easier using an online calculator such as this: http://www.1728.org/freqwavf.htm" rel="nofollow - http://www.1728.org/freqwavf.htm , than getting the math wrong, and I punched in 70 hertz which gives me a wavelength in air at 20 degrees C of 193 inches (490 cm).

The quarter wavelength is therefore 193/4 which is 48.25 inches (1.225 metres). So the "pipe" needs to be that length.

This would be unmanageable - a 4 foot pipe on the back of a driver - so a cabinet will have to be made such that a labyrinth of dividers inside produces a folded pipe having similar cross sectional area to the area of the cone, and is 4 foot in overall length.

The "port" can emerge anywhere you want, but if facing forward might lead to objectionable results - two outputs 90 degrees out of phase with one another - and early experiments having porting to the rear sounded good to me.

I will offer more background on this next post.




-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps



Replies:
Posted By: KJFAudio
Date Posted: 30 Dec 2016 at 9:26am
This is great stuff Graham, maybe worth pasting your plan to this page so that it doesn't get lost?

FWIW a BR can be made to not be one note, is a design choice really. Most modern designs are made to have a pronounced hump at the tuning frequency to give the impression of more bass but this does just lead to the one note sound you describe.

Also with noting that all speakers are just as much as a pain in the arse to get positioned correctly!

-------------
UK/EU dealer for Mark Audio Drivers and Manufacturer of speaker kits


Posted By: KJFAudio
Date Posted: 30 Dec 2016 at 10:14am
Oh, and if you can hang fire until mid week next week, I'll make you some prototype speaker boxes up.

Stefan

-------------
UK/EU dealer for Mark Audio Drivers and Manufacturer of speaker kits


Posted By: Graham Slee
Date Posted: 30 Dec 2016 at 1:30pm
Originally posted by KJFAudio KJFAudio wrote:

Oh, and if you can hang fire until mid week next week, I'll make you some prototype speaker boxes up.

Stefan


I know you do the Hurricanes but can you make to my design (approx.)? I'd like to experience just how good or bad my own effort is, and it would be educational for this topic (please).


-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: BAK
Date Posted: 30 Dec 2016 at 1:59pm
http://www.t-linespeakers.org/readme1st.html#TL" rel="nofollow - http://www.t-linespeakers.org/readme1st.html#TL

This site has links to most of the TL design masters.

-------------
Bruce
AT-14SA, Pickering XV-15, Hana EL, Technics SL-1600MK2, Lautus, Majestic DAC, Technics SH-8055 spectrum analyzer, Eminence Beta8A custom cabs; Proprius & Reflex M or C, Enjoy Life your way!


Posted By: KJFAudio
Date Posted: 30 Dec 2016 at 3:14pm
Originally posted by BAK BAK wrote:

http://www.t-linespeakers.org/readme1st.html#TL" rel="nofollow - http://www.t-linespeakers.org/readme1st.html#TL

This site has links to most of the TL design masters.


Yes, that's what I'll make for you. I'll just do it with MDF as a TL is well braced anyway.

Stefan

-------------
UK/EU dealer for Mark Audio Drivers and Manufacturer of speaker kits


Posted By: BAK
Date Posted: 30 Dec 2016 at 5:16pm
Originally posted by Graham Slee Graham Slee wrote:

Originally posted by KJFAudio KJFAudio wrote:

Oh, and if you can hang fire until mid week next week, I'll make you some prototype speaker boxes up.

Stefan


I know you do the Hurricanes but can you make to my design (approx.)? I'd like to experience just how good or bad my own effort is, and it would be educational for this topic (please).


I believe Stefan was replying to Graham's quote above with...

Originally posted by KJFAudio KJFAudio wrote:

Yes, that's what I'll make for you. I'll just do it with MDF as a TL is well braced anyway.

Stefan

Stefan



-------------
Bruce
AT-14SA, Pickering XV-15, Hana EL, Technics SL-1600MK2, Lautus, Majestic DAC, Technics SH-8055 spectrum analyzer, Eminence Beta8A custom cabs; Proprius & Reflex M or C, Enjoy Life your way!


Posted By: Graham Slee
Date Posted: 30 Dec 2016 at 6:06pm
Background to all this was me trying anything, with what I had available, to get close to the sound I want (and some of you know how pedantic I can get regarding that...).

All I had at hand was some carpet offcuts, bits of crafting felt, glue and rubber bands. Not much to go at.

At which time I had no idea about transmission lines at all. I rolled the carpet offcuts up which made a tube (well damped) and shoved them in the back of the half open cabinets (see pics).

After a good listen I concluded this was akin to a church organ pipe tuning a particular frequency. Wanting to make sense of this I did what we always seem to do these days - I googled until I started to see similarities, which led me to the transmission line.



Please feel free to roll about laughing.



Excuse my appearance. Picture taken using my phone.

And the design I came up with which is what I think Stefan will be building (dreading the cost of all this ... )

Today I had a listen to Chris Firth's FH3's fitted with Alpair 7M's (thanks for bringing them Chris, it was good to see you again).

These have to be heard to be believed. Chris brought some of his digital files on his Squeezebox, and we spun some vinyl via the Accession MC. The vinyl included Porcupine Tree's Incident album, and Chris had another Porcupine Tree album on Squeezebox, and so we both know the FH3 can handle rock rather well, which as we're both partial to the stuff that's a good job.

I'm not blessed with reviewers tongue so I won't wax lyrical and leave the story there (Chris might like to chip in though).




-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: Drewan77
Date Posted: 30 Dec 2016 at 6:38pm
Originally posted by Graham Slee Graham Slee wrote:



Headline: Genesis, Carpet Crawlers blows Graham Slee's socks off Shocked

(or on the LH side floor.... blew his brains out ShockedShockedShocked)


-------------
Older than I once was, younger than I'll be
.............................
Andrew


Posted By: Chris Firth
Date Posted: 30 Dec 2016 at 7:06pm
Well I rather enjoyed myself, and that's despite my deflicted left ear.

I had a list of questions, top of which was "will Graham like what these do?".
I didn't need an answer, as the broad grin spoke volumes, so we have another victim to the charms of the FH3s.
The sound was magnificent, harmonically rich, rhythmically adept and fleet of foot, and above all truly engaging and entertaining, a testament to Graham's electronics, and to the FH3s sporting Alpair7s.

I'm particularly partial to a bit of Brand X, and their album "Unorthodox Behaviour", and the first two tunes, "Nuclear Burn", and "Euthanasia Waltz" were rendered in all of their strange, melodious beauty - this is "no mates" music of the highest calibre, and Phil Collins has never sounded so good on drums.

On the vinyl front we had Foxtrot from Genesis, and The Incident by Porcupine Tree, and both were truly engaging with their complex instrumental arrangement and occasional bombast.

Other material from the usual suspects (Dire Straits, Steely Dan, Daft Punk and Jeff Buckley) sounded phenomenal.

My questions were how does the Accession/Majestic/Proprius sound - I know now, and I'm left wondering how much I could get if I sold my son for medical experimentation .....Shocked

I enjoyed my visit - thanks Graham Big smile



Posted By: ICL1P
Date Posted: 30 Dec 2016 at 9:57pm
My son certainly enjoys the FH3s we built together just over a year ago.

Stefan, how do the Pensil 7s compare to the FH3s?

-------------
Ifor
=====
Reflex M & ACCESSION M, CuSat50, Majestic DAC, a Proprius pair.


Posted By: Richardl60
Date Posted: 30 Dec 2016 at 10:09pm
Graham I wish the very best with your TL design I have been a huge fan for over 40 years. They are sadly not very fashionable with commercial manufacturers there are an odd manufacturer in Germany and USA but other than pmc very thin on the ground. Back pressure/loading is important so once you arrive at a working model sure you will find difffering materials and density will impact on the sound quite a lot. Will watch with interest as small TL designs have been virtually non existent (until now).


Posted By: ICL1P
Date Posted: 30 Dec 2016 at 10:23pm
I had been considering %20" rel="nofollow - http://www.iplacoustics.co.uk/ipl_acoustics_____transmission_l.htm , but like the simplicity of a single driver.

-------------
Ifor
=====
Reflex M & ACCESSION M, CuSat50, Majestic DAC, a Proprius pair.


Posted By: KJFAudio
Date Posted: 30 Dec 2016 at 11:10pm
Did you prefer the FH3 to your carpet tube? If so do you think they took up to much floor space for your needs? I could make you a pair of those as well You'll have more speakers than me soon

I love that photo of the carpet sticking out of the back of that speaker, I've got it as my screen saver at the moment!

Stefan

-------------
UK/EU dealer for Mark Audio Drivers and Manufacturer of speaker kits


Posted By: BAK
Date Posted: 30 Dec 2016 at 11:31pm
Originally posted by Graham Slee Graham Slee wrote:


Here is my idea for a TL for the Alpair 7. I wish I was good at woodwork! (hint)

Graham has the TL terminus (port or TL open end) on the back of the cabinet.
With the exception of the TL terminus location,
  This is very similar to the Radford S.90 as shown here... http://www.t-linespeakers.org/classics/radford.html" rel="nofollow - http://www.t-linespeakers.org/classics/radford.html    ...
     and the Bailey found here...    http://www.mh-audio.nl/bailey_tml.asp" rel="nofollow - http://www.mh-audio.nl/bailey_tml.asp  ...


Here is another idea with the port NOT on the front or back...  http://www.mh-audio.nl/TML.asp" rel="nofollow - http://www.mh-audio.nl/TML.asp   ...
 BUT on the TOP ! with only 1 fold in the TL.

The calculators on the mh-audio.nl site I found are good for a very rough approximation only... getting into the "ballpark". Some box size adjustments are still needed here.

HINT:
There may be other calculators that are a closer approximation to Martin J. King's and Dr A.R. Bailey's designs.

There are many speaker designs on the DIY websites... all types.
But, as with many other things, there are also many misdirected ideas as well.

I find the FH3 to have a very strong following and it has been designed with very close approximation to an
accepted design methodology.
(You will find many vague references like the one I just gave in this topic of TLs. Many that have the knowledge
 do not wish to dishonor or violate copyrights of others when making confirming statements. There are so many theories and factual evidences.)

The following is quoted from ... http://www.frugal-horn.com/" rel="nofollow - http://www.frugal-horn.com/ ...
"An estimated 1,000+ http://www.frugal-horn.com/FH3.html" rel="nofollow - have been built by the diy hifi community. FH3 been shown to work with a wide variety of drivers. The FH3 has been followed by a larger http://www.frugal-horn.com/FHXL.html" rel="nofollow - and smaller http://www.frugal-horn.com/FHL.html" rel="nofollow - . Independent cottage industry http://www.frugal-horn.com/get-FH.html" rel="nofollow - and bespoke finished Frugel-Horns has started up in Canada (North America), UK, Australia, & The Netherlands. More are being encouraged."

Note that the FH3 is the middle model. The FH XL being larger and the FH Lite being smaller.


-------------
Bruce
AT-14SA, Pickering XV-15, Hana EL, Technics SL-1600MK2, Lautus, Majestic DAC, Technics SH-8055 spectrum analyzer, Eminence Beta8A custom cabs; Proprius & Reflex M or C, Enjoy Life your way!


Posted By: BAK
Date Posted: 30 Dec 2016 at 11:51pm
Originally posted by BAK BAK wrote:

http://www.t-linespeakers.org/readme1st.html#TL" rel="nofollow - http://www.t-linespeakers.org/readme1st.html#TL

This site has links to most of the TL design masters.


The above "Read Me First" page was written by:
mailto:tlinespeakers@mac.com" rel="nofollow -


Posted By: BAK
Date Posted: 30 Dec 2016 at 11:52pm
Originally posted by Drewan77 Drewan77 wrote:

Originally posted by Graham Slee Graham Slee wrote:



Headline: Genesis, Carpet Crawlers blows Graham Slee's socks off Shocked

(or on the LH side floor.... blew his brains out ShockedShockedShocked)

 All that was left is his shoes !!!!


-------------
Bruce
AT-14SA, Pickering XV-15, Hana EL, Technics SL-1600MK2, Lautus, Majestic DAC, Technics SH-8055 spectrum analyzer, Eminence Beta8A custom cabs; Proprius & Reflex M or C, Enjoy Life your way!


Posted By: KJFAudio
Date Posted: 31 Dec 2016 at 8:45am
Originally posted by Richardl60 Richardl60 wrote:

Graham I wish the very best with your TL design I have been a huge fan for over 40 years. They are sadly not very fashionable with commercial manufacturers there are an odd manufacturer in Germany and USA but other than pmc very thin on the ground. Back pressure/loading is important so once you arrive at a working model sure you will find difffering materials and density will impact on the sound quite a lot. Will watch with interest as small TL designs have been virtually non existent (until now).


That Pensil is the only one I haven't actually built! I imagine it compares in a similar fashion to the others. The FHXL goes lower than the Pensil for the 10 series drivers but have a similar character other than that. The way the driver is loaded with air extremely tunable in both speakers with the stuffing. When Scott designed the Pensil range it was a study really in creating a new form factor that made it as easy as possible to build a speaker box with all the benefits normally associated with a TL. My Pensil 11 kits have just been built by a reviewer who thought they were amongst the most transparent speakers he'd had and they do bass aplenty to

-------------
UK/EU dealer for Mark Audio Drivers and Manufacturer of speaker kits


Posted By: Graham Slee
Date Posted: 31 Dec 2016 at 8:54pm
Having further thoughts about my prototype idea, I think it better to have a removeable side than a removeable baffle. This will mean adjusting the dimensions of the pipe depths to make them uniform because there would no longer be the battens to hold the baffle on. Also I guess the slight taper might be a hindrance to tuning, so it needs to be constant. Hopefully Stefan hasn't started making the boxes...

Some TL designers reckon the Vas has a bearing on the performance and the air mass of the line should be Vas or Vas x 0.7. The volume of the box as it stands is about 2.5 times the driver's Vas.

Making the box smaller would shorten the pipe which needs to be 1.2m (48") to tune to 70 Hz at 1/4 wavelength.

Perhaps the solution would be to stick further thicknesses of MDF on the inside - this would certainly reduce the volume, or maybe the design should be started over?

Here we need the opinion of the specialist (over to you Stefan).




-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: Graham Slee
Date Posted: 01 Jan 2017 at 3:37am
Having seen so much of the theory being based on actual pipes, and the Alpair 7 fitting a 103mm hole, it occurred to me that 4" (101.6 ID/110mm OD) soil pipe and 92.5º double socket bends could be utilised for practical assessment rather than going to all the trouble making trial and error cabinets.

Were you expecting a LOL at the end of that paragraph?

The seated ear height being approx. 44" (1.1m) would allow for a single downward bend in a 48" overall pipe with the open end down facing. The resulting pipework could be attached behind the baffle of the existing cabinet, which then leads to the question: of what use is the actual baffle (speaker front)? What real acoustic role does it play?

If it plays nothing then the cabinet would then only need to comprise of the pipework, with the speaker stuck in the front, the bottom end of which could be attached to a block of wood acting as a base to stand the contraption on.

Maybe the above is just a daft thought? (my first of 2017 - expecting a humorous answer from Drewan77... Wink)



-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: Graham Slee
Date Posted: 01 Jan 2017 at 4:21am
Ah, the things you find surfing...

http://audiojudgement.com/speaker-baffle-design-and-diffraction/" rel="nofollow - http://audiojudgement.com/speaker-baffle-design-and-diffraction/

Now I understand baffle-step.

Put quite simply the distance between the speaker and the edge of the cabinet will defract at some frequency increasing output above that frequency. In crossover design baffle step compensation is used to combat the increased output. Another way to reduce this boost in output is to cover the baffle with felt...

So here's why the LS35A have a felt surround to the tweeter, and why my hobby felt (see earlier photo) did make an audible difference, and I was not kidding myself.

So, to get minimal upper frequency boost the baffle should be non-existent, and so I am basically making a Teletubbies speaker!

So are the response plots of the Alpair 7 made on a baffle or not? And if so, what was the baffle dimensions? Without such info we are all in the dark.




-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: Graham Slee
Date Posted: 01 Jan 2017 at 5:52am
I also found https://www.google.com/patents/US6411721" rel="nofollow - https://www.google.com/patents/US6411721 which explains how my carpet roll works.

It sets the "boost" frequency at its 1/4 wavelength which is the internal pipe length plus the distance from the back of the carpet roll to the front plane of the cabinet - which keeps it in-phase - thus providing the boost. Therefore it isn't the 22" I had thought but twice that which is 44". This gives a 1/4 wavelength of 77Hz - not far away from the resonant frequency of the driver: 71Hz.

The output from the rear of the "pipe" also hits the rear wall providing further room reinforcement which will be at 150 - 160 Hz as calculated, and frequencies above that will also reflect with some decay as frequencies increase - maybe compensating for the driver "suck-out" in the lower mids as per its response plot.

With the bottom of the box stuffed with sound absorbing material, the box is not playing much of a part, except for baffle-step (see above), whose effect is attenuated by the hobby felt.

Having stuffed the "pipe" with sound absorbing material (polyester wool) it didn't sound right, much in accordance with what the above patent describes. And the open carpet roll sounded best, again as per the patent description.

The best way to understand the patent is by reading the notes for figures 1 then 18 - 20.

The "pipe" didn't do much (at 1/8th wavelength) to load the resonant frequency of the driver however, as evidenced by considerable cone movement.

So, in a way, this isn't a true TL, but still can be made into an acoustic labyrinth.

Is it viable commercially? Not if one doesn't want to pay royalties, but the patent expires in 12 months time.



-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: morris_minor
Date Posted: 01 Jan 2017 at 9:55am
Do I detect a GSP Speaker launch coming in 2017 (or 18 or 19)? LOL

-------------
Bob

Majestic DAC/pre-amp
Accession MC/Enigma, Accession MM, Reflex M, Elevator EXP, Era Gold V
Solo ULDE, Novo, Lautus USB and digital, Libran balanced, CuSat50
2 x Proprius + Spatia/Spatia Links


Posted By: Drewan77
Date Posted: 01 Jan 2017 at 11:20am
Originally posted by Graham Slee Graham Slee wrote:

...it occurred to me that 4" (101.6 ID/110mm OD) soil pipe and 92.5º double socket bends could be utilised for practical assessment rather than going to all the trouble making trial and error cabinets.....expecting a humorous answer from Drewan77... Wink)


Maybe it's not such a sh*t idea Graham (slurry overflow as tuning port)  Confused

 



-------------
Older than I once was, younger than I'll be
.............................
Andrew


Posted By: Graham Slee
Date Posted: 01 Jan 2017 at 1:39pm
Further notes regarding baffle step effects:

The protruding Alpair 7 surround gives some diffraction at 900 Hz, and the baffle width at 600
Hz. This coupled with room reinforcement for the bass, gives something like a 4-5dB boost at 600 Hz and perhaps a dB or two at 900 Hz - quite objectionable (try listening to a 700Hz squarewave for more than a few seconds).

Rather than using a baffle-step correction filter - basically a crossover for one driver - by changing the baffle width the diffraction frequency can be brought lower towards where room resonances take over.

Widening the baffle width to 12" reduced it to 380 Hz, which improved things substantially, but still there is some "whistle noise" which could be the protruding surround. One solution to that could be rear mounting the drive unit and flaring the hole.

I think now the construction of the box needs to be halted whilst all these variables are worked out.




-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: KJFAudio
Date Posted: 02 Jan 2017 at 10:43am
The use of pipe for prototyping isn't to bad an idea, of course you can't control the taper or cross section but if your calculations don't need that then it works well.

Try and see the baffle step the opposite way round that you are seeing it, so rather that giving you gain is gives you a loss. When the wavefront leaves the speaker cone it travels outward in all directions. A baffle only allows it to move in a hemispherical way or 2Pi, once the wavefront gets to the edge of the baffle is able to move in a spherical way or 4pi. Once we are in 4pi we are loosing 6dB at frequencies below the wavelength of the baffle width. Another thing a happens at the edge which is diffraction or illumination. Put simple this means that the edge acts as a new source, picture a ripple spreading out in a pond, as the ripple reaches an obstacle like a twig it will created a new wavefront spreading outward from the twig (baffle edge) this wavefront will interact with the original, causing some cancellation and some reinforcement. In our speaker that causes +-3dB ripples over a narrow Q centred just before the baffle step. 

Graham you mentioned the baffle step correction circuit, that allows the speaker designer to attenuate frequencies above the chosen point by 6dB. This circuit very rarely gets used for a few reasons, the primary reason is that it adds components to the crossover circuit which adds expense for the MF. Secondly we almost never want a full 6dB of attenuation as we will always have some room gain. If you visualise your speaker with its back against the wall the wall is an extension of the baffle and will continue to support the wavefront in a 2pi radiation pattern so in this case there will just be a ripple caused by the baffle edge illumination and no losses. Lastly it is easier to just adjust the value of the bass driver coil (in a 2 way) to give a bit of attenuation, this can give us 3dB or in a 4 way just adjust the level of the bass driver and cross to it at the baffle step frequency. In the states where larger rooms are more common and the speakers will be out into the room more, baffle step is more common or a full baffle step design like a 2.5 or 3.5 way design.

Tips, rounded edges help with edge illumination, felt on the baffle also help with illumination, off centre driver placement help spread the effect of edge illumination, avoid sharp edges (so flush drivers (dont rear mount them with a flared baffle, this will horn load them which is a whole other load of maths probably best avoided!)), tubular enclosures will create a perfect standing wave so be careful with the pipes 

Driver manufacturers all measure their drivers on an IEC baffle, that is a large open backed baffle where the driver is sited off centre and the wavefront is supported over most of the bandwith and measurements are done at 1 meter at 1 Watt or 1 Volt. That means the graph you are seeing shows the driver radiating into a 2pi space.

I'll hold off making you a prototype Graham until you finish  your experiments with pipes. I'm hoping to pass you on the 21st on my way to a bake off so could drop in with a prototype box then if that suits? 

Re the Hurricane, yes that was the speaker I was talking about, again it's designed by Dr. Scott Lindgren the foremost thinker on driver loading for wideband drivers. Scott will ahve taken all the things you are talking about here into account and more. He also has the benefit of access to measurement equipment and an anechoic chamber. The comments about rock are not just confined to this particular speaker but is the reviewers feeling about wide band drivers. To be fair if pushed hard with music that is a 'wall of sound' a driver covering a wide bandwidth will harden and start to distort, hence the use of increasingly multi way speakers. Used sensibly within their design criterion though all is good as you have found so far.

One more thing to say on the use of vented boxes of any sort, TL included. When a driver becomes unloaded (below the tuning frequency) it will be subject to a lot of excursion without actually making any sound, this is obviously not good for a driver and at best will sap the dynamics of the bandwidth you are able to hear. It's best to try and limit the level below the tuning frequency somehow. I'm active so that's easy, some use a filter, some just don't play so loud, some alter the corner frequency of their amps or feed back.......

Oh and download a bit of software called 'the edge' it allows you to model the baffle step and edge diffraction easily.

Hope some of that helps?




-------------
UK/EU dealer for Mark Audio Drivers and Manufacturer of speaker kits


Posted By: Graham Slee
Date Posted: 02 Jan 2017 at 10:48am
See my article on baffle steps (my blog bit...)

I can now see that maybe the transmission line approach to getting a precision response from these drivers is possibly the wrong direction, although I will not discount it fully yet. Replacing the backs, making it a sealed box again, and also stuffing the area behind the driver to remove reflections from the inside rear panel, plus carpeting the entire front (which damps the upper mids - thanks to Bruce for that), revealed that there was sufficient bass for what I need. It's just that the mids and highs are in your face.

It seems that the baffle step response is the real problem here and that the cabinets are fundamentally wrong for this full range driver, without any "crossover" to deal with this real and audible step.

The shape may be trendy and is what most cabinets look like these days, but the baffle does the drivers no favours at all. Diffraction proper starts at around 600 Hz due to the distance between the driver and the cabinet edges, except for the distance between the driver and the bottom of the speaker, which gives some lower frequency boost, in fact hardly any.

Add the below 600 Hz loss to the 400 Hz suck out in the driver response and a good proportion of the music is missing. It is quite obvious that the baffle size needs to accomodate lower frequencies, say down to 200 - 300 Hz where the room should be starting to take over.

If we took 250 Hz as a starting point the baffle width would need to be 18 inches, and in fact I experimented by giving the existing cabinet cheeks making it 12" which took it to 380 Hz, and the sound was much better.

The plastic speaker surround on these Alpair 7's is not helping either and needs to be recessed into the baffle or treated with some felt, or both, as it does its own diffraction, and hence gain at just under 1kHz. The driver itself is its own baffle when you think about it, and that will accentuate its own output at around 1500 Hz, but there's not much you can do about that.

Because of my lack of woodworking tools I think it possibly best to reinstate the Pluvia 7 drivers which will recess in with this cabinet. Then it should be down to deciding which baffle width best suits.

-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: Drewan77
Date Posted: 02 Jan 2017 at 11:38am
Regarding baffle step and accentuated frequencies: 

On my existing OB speakers, an initial slight issue with treble annoyed me so I tested with various lambs wool felt treatments around the ribbon tweeters as below. 

Cutting a thin card template & then gluing the material using contact adhesive created clean shapes & by varying the card, different depths could be tried temporarily with blu-tak. 

As well as listening, I was able to take verification measurements - the final double-layer shown is the end result. Being able to adjust the different drivers actively was also a big help.









-------------
Older than I once was, younger than I'll be
.............................
Andrew


Posted By: Ash
Date Posted: 02 Jan 2017 at 9:31pm
Not gonna lie. I don't understand 90% of the content of this topic but really wish I did.

-------------
We do not see things as they are. We see things as we are.


Posted By: Graham Slee
Date Posted: 03 Jan 2017 at 9:29am
Update on baffle step correction:

Armed with some CLS timber and Mitre Bond the baffle dimensions were increased as per the images below...




The choice of CLS enabled adding 1.5 or 2.5 inches by sticking them endways or sideways.

Still with the Alpair 7 fitted, and with the backs on (sealed cabinets again) with a sheet of stuffing material right at the back of the cabinet, to absorb rear reflections only...

I started by sticking the CLS pieces endways (like the piece at the top). This gave a calculated width F-3dB of 365Hz, and height F-3dB of 315Hz. This is very close to a square baffle which is not considered good, but gave a much better tonal balance than the original cabinet. More of the lower mids were audible, which meant the volume could be turned down, and so it sounded a lot less piercing, in fact quite pleasant. The downsides being timing and a not too stable image.

Having used dots of Mitre Bond it was easy to knock the side pieces off with a swift tap of a hammer. These were then reattached sideways on...

The width F-3dB was calculated to be 434Hz, the height F-3dB remaining at 315Hz. The tonal balance improved further and the timing was felt to be much improved and on par with the LS35A (as it was used previously in the workshop).

However, the bass was a bit overblown with some music, and also lacking with other music. I felt this was due to a cabinet volume mismatch with the Aplair 7. I therefore decided to refit the Pluvia 7 drivers, which are suited to a sealed box of this volume (8 litres according to Stefan).

This cured the bass, and the sound is now very much to my liking Smile

Technically, the height width ratio is close to 1.4:1, which is akin to the ways used in designing 2nd order electronic filters, such that the frequencies are far enough apart so as to cause minimum ripple. This is probably why the sonic picture was less confused, making the timing much improved.

The lower baffle stop frequency at 315Hz is also 1.4:1 above 225Hz, which will be in the region where room modes take over to reinforce the bass output. The centre of the drive units being 15" from the side walls and 30" from the rear wall. The structure although using dot and dab, is constructed to building regulations part E, and so similar results should be obtainable in rooms of properties built in recent years (as long as they were built to the regs).

I would now like to be able to hear a 8 litre cabinet made to these baffle (speaker front) dimensions, to be able to conclude that what I have done has only been due to baffle step correction, and nothing to do with anything else.

The chamfer to CLS timber is only very slight, so there will still be some boosted artefacts due to corners, and these may well be at the -3dB frequencies stated, such that they might not be as low as -3dB, and if so, all the better IMO.

So now, where are we at sound wise? The image is much improved; much more of the music can be heard; any confusion in the music is just about gone, and lyrics are much clearer and easier to make out...

An example being Pink Floyd's Final Cut track Not Now John, in the ending where there is a lot of shouting, and now Water's can clearly be heard shouting "Hey, where's the f*****g bar gone!"

And that's what I call high fidelity!

So far, so good.


-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: Fatmangolf
Date Posted: 03 Jan 2017 at 9:05pm

Thanks for sharing this Graham. It is exciting and educational! Thumbs Up



-------------
Jon

Open mind and ears whilst owning GSP Genera, Accession M, Accession MC, Elevator EXP, Solo ULDE, Proprius amps, Cusat50 cables, Lautus digital cable, Spatia cables and links, and a Majestic DAC.


Posted By: Graham Slee
Date Posted: 04 Jan 2017 at 6:41am
Originally posted by Fatmangolf Fatmangolf wrote:

Thanks for sharing this Graham. It is exciting and educational! Thumbs Up



Possibly too educating...

Later on I made a further change by adding pieces of small offcuts to take the height width ratio to just over (root 2):1.

(root 2):1 is a fraction over 1.4:1, in fact 1.4142135...:1, it's so much easier calling it (root 2):1.

Nature seems pretty specific if not pedantic about its ratios because, even though this adjustment wasn't all that accurate, it was closer to (root 2):1 than before, and made an audible difference I thought sounded better.

However, due to extending the top, we now have a driver that's very nearly offset from being dead centre. The actual baffle height step frequency is lower or higher depending where you measure the half height from the top or from the bottom. So we have a multiplicity of height frequencies, and my partly mathematical brain doesn't like that.

It has a "gut" feeling that the driver needs to be dead-centre, but that would mean building a new cabinet.

Now if nature is a pedantic thing, then we might have to delve deeper into other things. One being the question "is a sealed cabinet really all that good?". Having forgot to replace all the back panel screws, and when I did, the sound changed, made me think what was going on.

The cabinet without all screws in place was obviously ported, but for what frequency? The bolts are approximately 6mm through T-nuts, and two were missing. Using the same formula as for baffle step, each hole has a half wave frequency of 18240 Hz. Is that 9120 Hz for two?

The higher of those two frequencies won't be heard by the older adult, but the lower one can be. But here we're talking about a miniscule quantity. What the cabinet seems to be without all screws in, is what's called an "acoustic suspension" speaker, which is a closed box vented to the outside air, such that it doesn't have any effect at boosting any audible frequency. The thought being that barometric pressure can offset a driver, which then obviously cannot "swing" around its at rest position, because it's offset.

All sorts of techniques seem to be used for acoustic suspension, but to my mind if you drill a small hole in the cabinet somewhere, it's going to equalise pressure. A 2mm hole has a wavelength around 40kHz - and I won't be able to hear that! Surely it won't upset the acoustic spring of the box??

Having tried it and not noticing anything untoward, I can tick that box as having been done, and removing another variable.

So, to recap, we now have a frontal area size which is (swapping to metric here) 265mm wide by 375 high, which seems to be working in the room, but with no porting or TL loading, is a little tight on the bass side.

And penultimately for this instalment, room positioning is getting to be rather tedious to be able to obtain a precision stereo image. The room is the problem because one half is not the mirror image of the other - but who has a room that is? Speakers should be made to work in real life so all can enjoy. Obviously there must be some further tricks involved. Perhaps Stefan can throw some light on this?

(Edit: the structure of the baffle extensions are not particularly stiff which won't be helping)

Lastly, I have not yet tried a speaker zobel network (posh way of saying a series resistor and capacitor across the driver - like a snubber circuit) to compensate for the rising inductance at high frequencies. Calculation suggests an R of 9 Ohms, and a C of 470nF. I shall hunt around for these values and try that next.



-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: Graham Slee
Date Posted: 04 Jan 2017 at 4:45pm
Although acting at 37.6 kHz the zobel network seems to have tidied up the sound somewhat, and is worth having IMO.


-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: Graham Slee
Date Posted: 05 Jan 2017 at 2:37am
Time to take stock of developments so far.

This project started due to the LS35A refusing to perform in my new part E constructed office. I recall Ash's descriptions of his sound in his "small" room from his kit LS35A, and although not identical in every way to ready built units, the crossover electronics should at least be common to both.

The LS35A have performed admirably in rooms with stud partition walls - stuffed with rockwool as are many studio walls. I have always used them along a wall at least 12 feet long, but now I'm limited to the 8 foot 6 width of this new office, and the walls are 15mm soundbloc plasterboard over masonry cavity wall. I wonder what the walls are at Ash's home?

Now with Sound Service's bass traps in position, the sound is still muddy where previously one visitor described them as bright - that being in the other place, the one with stud wall construction.

I expect they will also perform quite well with US customers whose homes tend to be wood frame construction (but not in every case), and in the UK there are many wood frame constructed homes, but this office is a complete departure from that.

Moving onto the full range drivers, and cabinetry so far. These have been a lot brighter in their presentation but lacking in substance (the lower mids and bass). I have tried to correct for this forward presentation by experiment, one being the one which led to the formation of this topic.

Although mightily impressed by their resolving powers, particularly with the very real presentation of timbre and timing, they are far too bright to be a success beyond the sale of say a few thousand to people like me.

I found myself falling asleep through fatigue. This I have found has been one tool I have in the development of my electronics - if I fall asleep during listening tests then I got it wrong, and it's back to the drawing board. It's not because they fail in their brilliance, it's because of the fatigue that brilliance results in - I fall asleep like I would if suffering from something, it's a natural response. Where younger souls might have more stamina, I, at 61, don't.

I have tried to correct the response through adjustment of baffle step, because the change from hemispherical (4 pi) to half wave (2 pi) which occurs due to baffle width, boosts the upper section of frequencies by +6dB (as if your twice as close to the drivers at higher frequencies than at lower). Unfortunately this hasn't yielded the expected results.

The Frugal Horn 3 (FH3) succeeded however, and I can only put this down to the baffle height being 3 foot tall, leading to a 127 Hz low end baffle step frequency. This, when added to the horn output makes the design "work", and it produces a quite delightful sonic quality, ever so slightly marred by panel resonance, which I'm sure could be stiffened.

Here I am being hyper critical - the FH3 has been the only speaker capable of giving great results in this room yet!

So how do we get around the imbalance between highs and lows in a stand mount cabinet? I think Stefan begged me not to use electronic baffle step correction earlier (or words to that effect), but there seems no other option.

Baffle step correction is simply a series inductor like what you have in a two way crossover in line with the woofer, but with a bleed resistor across it to allow the highs to be shelved - brought back to flat once the desired attenuation is reached.

But where to apply the turnover? Perhaps it needs to clone the estimated 127 Hz of the FH3? We are in the realms of guesswork right now. The only thing to do would be to try it.

However, inductors of the type required don't seem to be easily available off the shelf. One would have to construct them. It's as simple as winding transformer wire onto a wooden bobbin - lots of turns - I've had the displeasure before.

It could make sense to have a multi-tapped inductor for different frequencies, with which could be tried a range of resistors  in the hopes of being able to hit that sweet spot. This could give some sort of reference point using the "sealed" cabinet, and be a jumping off point for the transmission line.

Taking stock further, perhaps some existing cabinet design would do the job? The FH3 certainly did. I cannot believe for one moment that the small cabinet designs offered online can do any better than the one Stefan supplied. Perhaps the Pensil 7 could be the answer? It's height being the same as the FH3, should offer a similar 127 Hz baffle height step frequency?

I guess the only thing putting me off are some negative comments around the DIY forums where it seems some people are forever engaged in adjusting the stuffing.

The Pensil 7 is a form of transmission line which doesn't have a labyrinth of internal baffles. It doesn't seem to be braced all that well, and that might be the reason for the forum stuffing issue.

I also noted that the FH3 benefited greatly by its outrigger spike mounts. Perhaps the Pensil 7 should have these too?

Having now got a pair of Alpair 7s doing nothing it is tempting to try the Pensil 7 cabinets...

Message to Stefan: any possibility? and do they come with outriggers?


-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: Graham Slee
Date Posted: 06 Jan 2017 at 12:36am
Originally posted by I I wrote:

Although acting at 37.6 kHz the zobel network seems to have tidied up the sound somewhat, and is worth having IMO.


Components do run in after all, and after some time of use distortion of the highs were detected, and after removal of the zobel network the distortion was no longer present.

Originally posted by I I wrote:

Moving onto the full range drivers, and cabinetry so far. These have been a lot brighter in their presentation but lacking in substance (the lower mids and bass).


This led me to discover this report from the BBC research department from 1977...

http://downloads.bbc.co.uk/rd/pubs/reports/1977-03.pdf" rel="nofollow - http://downloads.bbc.co.uk/rd/pubs/reports/1977-03.pdf

Page 19 starting third paragraph right column sheds some light regarding where the brightness could come from. Thicker material having higher Q pushes up the resonant frequency, where thinner material pushes it downwards.

There is also much discussion of damping or deadening material, and it seems that car panel deadening material (deduced) fits the bill.

It might be sensible at this juncture to build a new cabinet having the frontal dimensions found by experiment earlier in this topic, with 8 litre volume, made from 9mm material with deadening material adhered to it.

What I'm looking for is a reference point where everything sounds as uncoloured as possible before embarking on the transmission line for real. Thus far I find the colorations to be too great to proceed.

I also believe that rockwool was used as "stuffing" instead of polyester, but have access to the latter, so will have to listen to results with that first.


-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: Dave Friday
Date Posted: 06 Jan 2017 at 1:11am
I remember tapped coils used In ( Bbc?)speakers and tissue paper put in front of tweeters ( to take the edgey ness off )
I like the bit about timing,now waiting for the next bit.

-------------
lp12,oc9mk3,ca610p,krimson40watt pa,kef105.4


Posted By: KJFAudio
Date Posted: 06 Jan 2017 at 1:10pm
Hi all,

Sorry missed a lot of what has happened here, been crazy busy!

I wanted to touch on a couple of points.

Re the baffle step. I did mention this in an earlier post but the 6dB loss is a bit of a red herring, you'll loose about 3 dB because of the proximity to the walls. Although the Pluvia is flatter at the bottom end the Alpair has a bit of a bass hump intended to compensate a little. The LS3 5/a while it doesn't have a full baffle step circuit either does have the level of the tweeter set lower than that of the woofer which give a small amount of baffle step and gives the famous BBC dip in the mids, which may explain you finding these a bit brighter.

Your experiments with baffle with and height have been very interesting so here are a couple of models to show you what is happening and to challenge your gut feeling that the driver benefits from being central. You can see that the green line of the original box with its higher f3 has a smaller hump just before the fall off that the larger box with the lower f3. This is because the driver is a different distance from all the edges except 2 of them, this 'averages' the gain caused by the edge illumination. in the last example I moved the driver position so that the offset was .707 from the bottom and .707 from the side, this averages the edge gain (illumination) much better and you can see that the repeating harmonics of that are flatter as you move up in frequency. Another piece of acoustics where those mathematical constants come in very handy!




Adventures with panel thickness may turn out to be a long and dark rabit hole that you crawl down and never get back out of. The BBC paper is great, however, the BBC have (had) huge resources at thier disposal for research. Building a lossy cabinet brings with it other problems which need to be cured usually in a crossover filter. The panels will resonate more easily as they get thinner, and that is fine so long as you can account for that elsewhere. The BBC needed to produce a LOT of cabinets as cheaply as possible and so it was worth doing the research to develop a speaker that worked despite it's cheap build quality and it works great. However unless you have the time and inclination for building lots of prototypes and taking lots of acoustic measurement and accelerometer measurements of the panels you are better off with a panel that doesn't resonate. Of course it is impossible to make a box completely inert but Baltic birch ply is very stiff as it is easy to make a box stiffer with increasing thicknes (stiffness increases roughly four fold for a doubling in thickness) here we are using 18mm ply with ring braces in a very small box with 25mm front baffle, so lets assume its plenty stiff enough to only get excited when things are pretty loud. As you pointed out the Q of a stiff panel is pretty steep so a little easier to deal with than a wider Q spike. Usually in a sealed box having the stuffing up against the wall is enough to damp this but with your experimentation it may be worth adding something to the walls with a much better modulous like bitumen pads ( sound deadening pads for car vehicles)

I am more than happy for you to use some sort of electronic correction for the drivers, of course the single driver purists will hate you but mehh just done go ofr the full 6dB and maybe get your self a mic like the UMIK-1 so you can see a bit more what is hsppening.

The zobel experiments are pretty interesting again, however the point of a zobel in a cross over network is to flatten the impedance of the driver so that the crossover filter has a constant to look at other wise the filtering would alter as the impedance of the driver altered so you could never get a apir of filters to sum correctly. Without a crossover though and the amp driving the speaker directly that is of much less consequence assuming that and out put filters in the amp (and cables) aren't significantly affecting the shape of the signal when seeing a varying load. Most modern amps don't but older amps and tube amps might.

I've a feeling that now you are getting a few hours of messing about under your belt and are starting to see what changes do and how they affect each other that you might be best going back to the Alpair which is the better driver and stating again. I could either build you a box to your dimensions at the correct volume (sealed) or one of your TL ideas. The box you have got need to be about 1.5 litres less. The baffle needs to be cut o that the driver sits flush or you could just go for the Pencil 7 which will fill out the bass a little (BTW it's not really a TL it's a resonant air coupler with some TL properties and the driver positioned to cancel some unwanted nodes in the box) This can be built with braces to stiffen it a little more. You will be experimenting with the stuffing, but not endlessly I think, or certainly no more than you currently have.

I'm passing near you again on the 21st I think so can stop by with speaker boxes if you want?

Here is a model of the pensil 7 baffle step




-------------
UK/EU dealer for Mark Audio Drivers and Manufacturer of speaker kits


Posted By: Graham Slee
Date Posted: 06 Jan 2017 at 3:19pm
Quite interesting looking at the frequency response models for what I did. At least it showed the calculations work, but in this case not to my favour. I will give the 0.707 offsets a try (somehow) and report back later. I ordered some car panel deadening sheets of the type you mention yesterday, and will try them in the existing box. I would much rather use 18mm MDF if constructing boxes myself because I find it easier.


-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: Graham Slee
Date Posted: 06 Jan 2017 at 5:14pm
The 0.707 placement sounds better to my ears.

However, room modes aren't taking over all that much which leaves me with little output below the 200 - 300 Hz region. Taking out 5 of the 6 rear panel screws does absolute wonders to the bass! Not boomy or colored at all to my ears. In fact these cabinets sound really full range and the imaging on Waters Amused to Death is WALK THROUGH in terms of the image not collapsing when you "look" at the Q-sound effects, which it always did with the LS35A (collapse that is).

The only thing wrong with removing the screws is that most of the rear panel resonates (felt by touching the panel), and that is a big no-no.

Except that it does wonderful bass!

Need an explanation.



-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: KJFAudio
Date Posted: 06 Jan 2017 at 11:58pm
Well it becomes a Bass reflex basically, with 5 ports. I've tried to model it and although difficult, as I've no idea how how clear the pathway from the driver to the vents is, and obviously the vents have also got the threaded inserts in them, Taking a stab at a few assumptions and plugging the number in anyway, you've got a 55-60Hz f3 with a +2 ish dB lift at about 70-80 Hz

Port velocity would be quite high because of the narrow size and group delay wouldn't be as good as the sealed but if it sounds good to you then that's great news.

Maybe after all this you would be good with a ported golden ratio box for the Alpair, take a look at the design which is listed on the Mark Audio website. It's approaching the shallower and wider approach you have ended up at here.

Stefan

-------------
UK/EU dealer for Mark Audio Drivers and Manufacturer of speaker kits


Posted By: ICL1P
Date Posted: 07 Jan 2017 at 12:26am
Originally posted by KJFAudio KJFAudio wrote:


Maybe after all this you would be good with a ported golden ratio box for the Alpair, take a look at the design which is listed on the Mark Audio website. It's approaching the shallower and wider approach you have ended up at here.

Stefan

http://www.markaudio.com/Plans/item/download/43_dcb268cfa838ba135e58afa3edd41202" rel="nofollow - This?

-------------
Ifor
=====
Reflex M & ACCESSION M, CuSat50, Majestic DAC, a Proprius pair.


Posted By: KJFAudio
Date Posted: 07 Jan 2017 at 8:20am
Yeah, that's the one. Offset the drivers as per discussions above and you are good to go. I've never actually modelled that speaker, might have a look tonight if I get time.

Stefan

-------------
UK/EU dealer for Mark Audio Drivers and Manufacturer of speaker kits


Posted By: Graham Slee
Date Posted: 08 Jan 2017 at 5:12am
Originally posted by KJFAudio KJFAudio wrote:

Well it becomes a Bass reflex basically, with 5 ports. I've tried to model it and although difficult, as I've no idea how how clear the pathway from the driver to the vents is, and obviously the vents have also got the threaded inserts in them, Taking a stab at a few assumptions and plugging the number in anyway, you've got a 55-60Hz f3 with a +2 ish dB lift at about 70-80 Hz

Port velocity would be quite high because of the narrow size and group delay wouldn't be as good as the sealed but if it sounds good to you then that's great news.

Maybe after all this you would be good with a ported golden ratio box for the Alpair, take a look at the design which is listed on the Mark Audio website. It's approaching the shallower and wider approach you have ended up at here.

Stefan


I calculated an equivalent port size of 16mm dia x 30mm long (thickness of rear panel plus threaded insert "fillet".

Replacing just one more screw would reduce that to 14mm x 30mm, and the difference is quite staggering in a negative sense (bass collapses).

This is why I thought panel resonance was partly responsible. The additional screw stiffens the panel.

But, from what you say, there being a 2dB lift at 70-80 Hz due to port velocity (?), then increasing it's size may improve it?

TBH the sound is a tad colored.

Wanting to keep the Pluvia 7's in the existing box which is 8 litres, I came across this for the Pluvia 7: https://www.madisoundspeakerstore.com/pdf/Pv7-0.25ft_3.pdf" rel="nofollow - https://www.madisoundspeakerstore.com/pdf/Pv7-0.25ft_3.pdf , which is 7 litres, and https://www.madisoundspeakerstore.com/pdf/Pv7-9Ltrs.pdf" rel="nofollow - https://www.madisoundspeakerstore.com/pdf/Pv7-9Ltrs.pdf , which is 9 litres.

The ports differ by just 1mm in length...

I can obtain some 1.5" plastic waste pipe to make the ports??




-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: Graham Slee
Date Posted: 08 Jan 2017 at 6:38am
Anyway, back to the transmission line idea, and the carpet tubes were doing 1/8 wavelength (150Hz), and if you add the distance from their exit at the rear back round to the front of the cabinet, you get 1/4 wavelength at 70Hz as near as damn it.

But they weren't doing anything to flatten the driver response, so resonant frequency (70Hz) output power wouldn't have been much to start with.

Then I installed corner bass traps in the room, and the bass was much reduced, and that basically led to stumbling on the "accidental bass reflex" above, due to being a good deal dissatisfied...

So will a transmission line proper work?

The harmonic labyrinth patent will have you believe that a rear opening will lower the frequency because it has to travel from back to front, thus increasing the wavelength, and hence lowering the frequency.

If I tune a line (pipe) to the driver resonance and very lightly fill with stuffing, it should give some mechanical stiffness to the cone, which lowers its electrical resistance, which should mean more power is developed??

However, a sealed cabinet is supposed to do that, but if what I heard was the driver's full output at its resonant frequency, well, what can I say? It wasn't much.

I think we're now getting close to some sort of conclusion; that the air displacement of such a small speaker isn't sufficient to make any substantial volume in the bass region.

The little bass output there is requires boost to equalise it to what our human hearing expects, and room modes haven't been doing that, or should I say, when pushing the speakers close to the wall, it may have been helping, but the loss in sound staging and tonal colour made for a lack lustre rendition.

Bringing speakers out into the room does improve matters greatly, but then the bass reinforcement isn't there.

So will a transmission line really suit these drivers? I am having my doubts. The HiFi Pig review of the Hurricane got me a bit concerned when it refered to rock music. The sound balance has to include some bass or rock music falls apart. I didn't realise how important this was before, and so the journey so far has helped me learn.

One favourite track is Cinema Show from Selling England By The Pound (Genesis). This is dominated by lots of melody, but if you listen you will hear it is underpinned with a bass line and percussion. The melody stands out, but it needs that bass foundation.

Without it, the left-right balance goes. It all falls to the right. Yes, there are central images, but there is little left side output, and it simply doesn't sound right. For a perfectionist like me it becomes quite annoying.

The accidental bass reflex obtained by "having a mess" put the balance back. Imagery started happening. Water's barking dog appeared right and forward of the right hand speaker; Razz appears to be talking on a radio or TV well to the left and forward of the left hand speaker - you can gaze these out and the images don't move.

However, the two guys on the talkback (Water's Amused to Death: Watching TV) did not appear to be right next to my right hand side. They are supposed to be there, and the first time you hear that you could easily have an accident in your trousers!

(the shock of finding two (invisible) guys sat next to you in what you thought was just you in the room)

And the lack of that suggests the bass output has a lot of ripple, and Stefan seems to make that clear.

The Stirling LS354A V2 could do all those Q-sound effects with ease, but not in this room. And after hearing what Ash had to say about his, I guess many rooms share the same problems. So, here what I'm trying to obtain is something better than the LS35A - better than the BBC!

So if my adventure into transmission line is about to grind to a standstill, where do I go from here?

There only seems to be horns and bass reflex capable of producing the bass my "universal sound balance" needs. And it has to be universal because I don't know what my customers are listening to.

I know how easy it is to make highly satisfactory sounds to fit one particular genre, such as a quartet, there is one (or similar) on "Give My Regards to Broad Street" (McCartney), but the rest of the album is in other genres, and so it has to work with all of it, otherwise it's wasted music.

You know, there are guys out there who think their products are God's own gift, and they may be, provided they're playing the guy's favourite music style - but when you request Led Zeppelin's Kashmir on the guy's classical tweaked system (or Bjork maybe?), it simply goes flat and washed out.

My job is to ensure what I make works for Led Zeppelin, Bjork, Classical orchestra, Regae, Jazz quartets, etc etc etc.

And so the speakers have to work on all those as well.

Perhaps it has to be bass reflex, or horn loaded to get the movement of air required for the bass?? I may be asking too much of a full range speaker here?? The thing is, with no crossovers, such a speaker will not be hiding or masking or otherwise changing the phase (electronically that is), and leading my efforts down the wrong path.

Hopefully we will shortly be homing in on a speaker that will work for me. Sometimes you just have to iterate several times before the curve converges on the correct result.



-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: Ash
Date Posted: 08 Jan 2017 at 8:42am
Graham, my small bedroom has hard artex walls and a ceiling that appears to be a hard thin type of polystyrene. Far from ideal materials. The fairly thick floor underlay and carpet and my overhead mattress would help to absorb/scatter sound, along with the big metal (some kind of hard painted steel) bed frame, which dominates the room increasing sound scatter, I'm sure.

The LS3/5A "kit" speaker pair I had were ok but nothing special. They didn't sound promising at any volume. They used the same type of thin-walled "reference" cabinet with screwed on backs as the V2 version I believe and the crossover network was advertised as the "latest". The drivers were different though. Monacor woofer and Scanspeak tweeter, I think. It was intended to replicate or give a taste of what the LS3/5A spec could do. The guy who bought them from me liked their sound when we listened to them in the larger living room downstairs, speakers seated away from the walls, sat on "really useful box" stacks. He didn't like headphones and didn't like the sound of the HD540II I had to hand as a listening reference. He preferred the more coloured and spatial LS3/5A rendition. I had kept them looking like new and hadn't bothered to fully run them in as I knew I still wouldn't be satisfied. I got a more enjoyable sound from the Equator D5 but they were far from pure sounding as well.

I have gone back and forth between speakers and headphones over the past 8 years and headphones have overall given me more transparent and enjoyable results. But now I have experience with how the various headphones sound, I don't need them anymore, including the K1000, because I have a clear idea of how the sound should translate. Mark Audio drivers, for me, have made leaving headphones a necessity; speakers are no longer just a second choice. I will enjoy them more. I have a good gut feeling about them after hearing Alpair 12PW/7P pairing and the 12P. Finally got some speakers that will be much more than a boomy box with an unresolving emitting surface.

-------------
We do not see things as they are. We see things as we are.


Posted By: KJFAudio
Date Posted: 08 Jan 2017 at 10:27am
The port size is based on cross sectional area and volume and isn't proortional so you can quite just add it all up.

Here are some thoughts on this bass reflex and continuing the use of the Pluvia 7 driver (still think if you are building another box you might be best basing this on the Alpair, it's a better driver as you discovered earlier.

Take a look at this image.


The yellow line is the original sealed enclosure when you didn't think you were that bothered about the bass (funny what we can discover to challenge our preconceptions along the way) It's a bit thin but has excellent group delay and excursion handling. The green line is the same box with the 5 screws missing and is pretty good, the tuning hump is only a dB or so, I was wrong earlier and is quite broad based, helping with the Baffle step a little up to nearly 500 Hz. You can see a faint blue line almost overlaid on that line, that is a calculation for just one port to give you what you have got now which is a 12 cm piece of your inch and a half pipe. The orange line is the observation that you have placing an extra screw back in the hole, a loss of bass indeed. The Grey line is a different box completely, it's an 11.5 litre golden ratio box (the one I spoke about earlier) but tuned for this driver with the same inch and a half pipe at 8 cm length. The port velocities of the single port boxes are much better meaning there won't be any port noises audible so the port is probably best on the front where it can sum with the primary better. The screw holes by comparison whistle at about 4KHz. Anything below about 12 here is pretty good



Lastly the group delay as pictured below is obviously much much better for the sealed box, however I', not so sure it matters here, firstly the group delay starts to become a problem below the tuning point where output is already dropping off, and TBH the room will have messed up the group delay long before that!!



So what can a TL give you that this BR can't? Well probably not much more that a headache designing it and trying to get it right especially without the necessary tools to model and measure it. If you do go that route, potentially you can get an even lower tuning. You can get a group delay as good as the sealed box as the driver doesn't get unloaded  at the tuning frequency (although it eventually does at say 10Hz) this means that you get more usable excursion so can go a bit louder.

Remember these graphs don't include the baffle step or the room just the box profile.

Looking at all this and all you have said I think using the Alpair in a golden ratio box, with driver offset and tuning to suit what you have found you like here would suit the best.

Stefan


-------------
UK/EU dealer for Mark Audio Drivers and Manufacturer of speaker kits


Posted By: Richardl60
Date Posted: 08 Jan 2017 at 2:27pm
Ash, out of curiosity how do you mount/support your drivers? Graham has ill used to the difficulties getting it right, what is your solution?


Posted By: BackinBlack
Date Posted: 08 Jan 2017 at 2:36pm
The "loose" back could be behaving as either an ABR (Auxiliary Bass Radiator) or ARU (Acoustic Resistance Unit) latterly developed as the Variovent by Dynaudio and others. I would think it is more likely akin to the ARU/Varivent effect.
Interestingly, the ARU was first described by EJ (Ted) Jordan in 1956, who went on to develop the Jordan range of full range drivers which use(d) spun metal cones, usually aluminium.
You'll need to find Wireless World Jan and Feb 1956 for his article, or for a full analysis see Richard Smalls' Thesis - Direct Radiator Electrodynamic Loudspeaker Systems (Sydney University May 1972) Chapter 7.
I find these observations fascinating.
After quite a few revisions of the stuffing in my FHXLs with Alpair 10.3, I thought I had the best sound. I'm now tempted to try some dedshete type damping to the inside of the throat below the driver where the largest panel areas could be resonating, even though they're built with 18mm birch ply. Perhaps this feature is an integral part of Scott Lindgren's design?


-------------
Just listen, if it sounds good to you, enjoy it.


Posted By: Ash
Date Posted: 08 Jan 2017 at 3:14pm
What listening I have done with MA drivers has been on cardboard baffles at low to moderate volume in a small room, from Proprius and Majestic. Sounded pretty good to me! Needs some improvements but I find it hard to be too disappointed at this stage. I'm a lot less critical than Graham and others because I don't rely on this stuff for a living/livelihood. If there is audible  room interaction, I just turn the volume down and/or move closer to the sound source until I no longer hear the issues. Embarrassed LOL I'm no audiophile really. I just like a clear, well-resolved and well-balanced spatial sound then I'm 90% satisfied.


P.S and Graham, Mark Fenlon himself said that heavier music like some rock/metal etc requires the use of a heavier duty driver (i.e a larger emitter to move more air). The 7 of both Alpair and Pluvia are quite diminutive compared to the larger MA drivers so at similar listening distances from them, they are inevitably going to sound smaller. The sound balance at higher listening volumes with the 7s would also tend towards elevating the higher frequencies IMO.


-------------
We do not see things as they are. We see things as we are.


Posted By: Ash
Date Posted: 08 Jan 2017 at 3:36pm
and for another reference, I am going to try a pair of the Alpair 5 Gen 2 spiderless drivers playing at very close proximity or as a K1000-esque headphone setup (mount them on a homemade headband and look like a complete and utter idiot) so I can monitor what colourations are being introduced to the bigger speakers out in the room.

-------------
We do not see things as they are. We see things as we are.


Posted By: KJFAudio
Date Posted: 08 Jan 2017 at 4:13pm
Hey Ash, you will just need to check that your amp has a reasonably high damping factor. Are you using one of Graham's? Graham can you chime here? The small Alpair has got no spider and so very little mechanical damping, especially if you are using it open backed. So you need an amp that has got a good grip on it to act as a brake. So mostly not a little tube SET, Naim (or other RCA) circuits are not a good idea either. In response to your PM, that is fine on this one occasion.

Stefan

-------------
UK/EU dealer for Mark Audio Drivers and Manufacturer of speaker kits


Posted By: KJFAudio
Date Posted: 08 Jan 2017 at 4:14pm
Originally posted by Ash Ash wrote:

and for another reference, I am going to try a pair of the Alpair 5 Gen 2 spiderless drivers playing at very close proximity or as a K1000-esque headphone setup (mount them on a homemade headband and look like a complete and utter idiot) so I can monitor what colourations are being introduced to the bigger speakers out in the room.


Graham is very close to these speakers so the amount of air that can be shifted isn't really to much of a problem.

-------------
UK/EU dealer for Mark Audio Drivers and Manufacturer of speaker kits


Posted By: KJFAudio
Date Posted: 08 Jan 2017 at 4:20pm
Originally posted by BackinBlack BackinBlack wrote:

The "loose" back could be behaving as either an ABR (Auxiliary Bass Radiator) or ARU (Acoustic Resistance Unit) latterly developed as the Variovent by Dynaudio and others. I would think it is more likely akin to the ARU/Varivent effect.
Interestingly, the ARU was first described by EJ (Ted) Jordan in 1956, who went on to develop the Jordan range of full range drivers which use(d) spun metal cones, usually aluminium.
You'll need to find Wireless World Jan and Feb 1956 for his article, or for a full analysis see Richard Smalls' Thesis - Direct Radiator Electrodynamic Loudspeaker Systems (Sydney University May 1972) Chapter 7.
I find these observations fascinating.
After quite a few revisions of the stuffing in my FHXLs with Alpair 10.3, I thought I had the best sound. I'm now tempted to try some dedshete type damping to the inside of the throat below the driver where the largest panel areas could be resonating, even though they're built with 18mm birch ply. Perhaps this feature is an integral part of Scott Lindgren's design?


Partly Maybe although an ABR really need a sealed box to work and the little ports Graham has got by accident put pay to that. Also the moving mass of the ABR would ideally be in the region of (ideally identical to) the MMS of the cone in the main driver. This is a great way to drive a subwoofer though, I've tried it a few times.

The main body of your FHXL is fairly rigid as the panels are all short or narrow and close together, however give it a try, usually people use (as do my kits) felt, that and Scott hates dedsheet....

-------------
UK/EU dealer for Mark Audio Drivers and Manufacturer of speaker kits


Posted By: BackinBlack
Date Posted: 08 Jan 2017 at 4:49pm
Originally posted by KJFAudio KJFAudio wrote:

The main body of your FHXL is fairly rigid as the panels are all shots and close together, however give it a try, usually people use (as do my kits) felt, that and Scott hates dedsheet....


Thanks Stefan,
Felt and contact adhesive it is then, I suppose that dedshete is the lazy option as it's self adhesive.

Ian


-------------
Just listen, if it sounds good to you, enjoy it.


Posted By: Graham Slee
Date Posted: 08 Jan 2017 at 4:54pm
Originally posted by KJFAudio KJFAudio wrote:

Hey Ash, you will just need to check that your amp has a reasonably high damping factor. Are you using one of Graham's? Graham can you chiming here? The small Apliar has got no spider and so very little mechanical damping, especially if you are using it open backed. So you need an amp that has got a good grip on it to act as a brake. So mostly not a little tube SET, Naim (or other RCA) circuits are not a good idea either. In response to your PM, that is fine on this one occasion.

Stefan


If we were to take output impedance at an arbitrary 100Hz, we would see a driving impedance of just 0.36 Ohms, and by existing wisdom this gives a damping factor of 22.

Amplifiers claiming more take into consideration the negative feedback ratio which effectively reduces driving impedance, but when called upon to drive a difficult load the amp runs out of negative feedback, and where it does, it will be just about the same DF.

So I wouldn't put much reliance on damping factor because it can turn round and bite.


-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: Graham Slee
Date Posted: 08 Jan 2017 at 5:03pm
A daft question, but having no 1.5" tube lying around (I thought I had), I do have some 60mm x 60mm section uPVC square downpipe. Could this be used and how?

I have a feeling you'll tell me it could but it would have to be so long that it would not fit the box. But it would be nice if it could be made to work (saving me the inconvenience of yet another B&Q visit).


-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: KJFAudio
Date Posted: 08 Jan 2017 at 7:09pm
I'll model it later, it can be made to work but might have a slightly better idea that is even easier for you, hold on.



-------------
UK/EU dealer for Mark Audio Drivers and Manufacturer of speaker kits


Posted By: KJFAudio
Date Posted: 08 Jan 2017 at 7:15pm
Originally posted by Graham Slee Graham Slee wrote:

A daft question, but having no 1.5" tube lying around (I thought I had), I do have some 60mm x 60mm section uPVC square downpipe. Could this be used and how?

I have a feeling you'll tell me it could but it would have to be so long that it would not fit the box. But it would be nice if it could be made to work (saving me the inconvenience of yet another B&Q visit).



That's more than enough damping to grab this speaker. Thanks for going into a bit of detail Graham. There aren't enough threads about electrical damping, people just think an amp gives the pushing force, they don't think about the stopping. Spider less wideband drivers may turn out to be the next big thing for that added transparency for those with a decent power amp like yours.

Must have a listen to one of your amps more extensively at some point...

-------------
UK/EU dealer for Mark Audio Drivers and Manufacturer of speaker kits


Posted By: Ash
Date Posted: 08 Jan 2017 at 7:32pm
I'll keep the Proprius volume dials very low as the 5, like the other Alpairs, is a high sensitivity driver so hiss will be easy to hear if the dials are too high. The emitting surface is larger than both the HD800 and the K1000; I reckon it will beat both if positioned at a sensible distance in front of the ear. Plenty of potential variables for experimentation. It appears to have to respectably balanced sound at the high test volume so at considerably lower volume, should sound even better.

-------------
We do not see things as they are. We see things as we are.


Posted By: Richardl60
Date Posted: 08 Jan 2017 at 7:55pm
Whilst your tube needs are quite modest, I am aware that the 6" mid range unit in the 'big IMF Transmission line used a cardboard tube stuffed with whatever material. This didn't feature an output from the cabinet but was a means of loading the driver.


Posted By: KJFAudio
Date Posted: 08 Jan 2017 at 9:04pm
Right, to keep the same tuning that we have observed above with your 60x60mm square tubing is just not really doable, the port would need to be 40cm long so it just won't fit Smile

Take a look at the image below, what I am proposing is a narrow but wide slot port at the bottom and rear of the box. It works in exactly the same way as a normal round tube (it doesn't matter a jot what shape the port is BTW). Advantages of this kind of port are, you can bastardise what you've got with a jigsaw and some 'no more nails', it saves dicking about with port length calculators when you can only find ports or tubes at set sizes and generally the port velocity is hugely lower than a pipe so can be used at the front of a box which is better.

In this case we are doing it at the back of the box as that panel is already removable. You'll need a bit of MDF or ply or whatever you got, some of that hateful plasterboard will do for a quick experiment LOL

You need to remove the back panel then take your jigsaw and cut away the bottom bit of the ring brace that the baffle normally bolts to so that is roughly flat with the bottom if the inside of the speaker, use a bit of no more nails and this bot of off cut to fill in the bottom of the lip where the baffle normally sits so that the bottom panel is flat and clear (ish) all the way to the back of the cabinet. Now take your bit of ply/MDF/Plasterboard and cut a piece that is the same width as the inside of the box, check it fits. Then cut the length so that it is 147mm long (165-18mm rear baffle) now glue ( as stick with blutac) this piece so that it is 10mm from the bottom of the box interior and flush with lip at the back. Now put the cut the rear baffle length down so that it is the same length as your new opening. Bolt it back into place. You'll need to seal between the end of the new piece and the baffle. Keep the stuffing above the new opening and make sure there is a free path from the driver to the new port.

This has got the same tuning as your bolt holes.






-------------
UK/EU dealer for Mark Audio Drivers and Manufacturer of speaker kits


Posted By: Graham Slee
Date Posted: 08 Jan 2017 at 9:50pm
Thank you. Details understood. I shall be in woodworking mode tomorrow afternoon, and hopefully playing lots of different music thereafter. I shall report back on progress. As the cabinets stand they're pretty accurate. Can't wait!


-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: Graham Slee
Date Posted: 09 Jan 2017 at 2:23pm
Work now in progress. Just about blunted one jigsaw blade having to chop through the flange of a T-nut as well as the ply. The air in here is blue with smoke from the extremely hot blade.


-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: KJFAudio
Date Posted: 09 Jan 2017 at 2:47pm
I think they are just screw in flanges, assuming I put them in front the side facing you, they just screw out with an Allen key! I rarely use T-nuts, they are hateful things! 

-------------
UK/EU dealer for Mark Audio Drivers and Manufacturer of speaker kits


Posted By: Graham Slee
Date Posted: 09 Jan 2017 at 5:37pm
Some pics of the reflex porting job...

Bass reflex slot port shelf

Bass reflex inside cabinet

Bass reflex slotted port 10mm x 153mm

Just waiting for glues and caulking to dry, and then in with the stuffing and drivers.


-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: Drewan77
Date Posted: 09 Jan 2017 at 6:29pm
Graham... 'So' in that last pic would be more appropriate as 'PG2' (with D.I.Y Wink)


-------------
Older than I once was, younger than I'll be
.............................
Andrew


Posted By: Graham Slee
Date Posted: 09 Jan 2017 at 11:00pm
Originally posted by Drewan77 Drewan77 wrote:

Graham... 'So' in that last pic would be more appropriate as 'PG2' (with D.I.Y Wink)


Sorry Andrew, way above my head. PG2??

And now the results:

This is a bass reflex cabinet in its true light. It does exactly what it says on the can, and for bass lovers, will probably suit them down to the ground...

But not me.

I think the loose rear panel was doing more than being a five mini-port bass reflex. I think the rear panel must have a resonant frequency which when held with just one corner screw, did it for me.

I must be ultra-sensitive to both phase and the smallest amount of response droop. That's possibly how I design my equipment...

No, there is something not quite to my liking and I had to find out what it is, or I hope what I found is correct.

The baffle step cuts around 300 - 350 Hz - it is -3dB (half power) from its lower mid gentle peak in the region of 300 - 350 Hz.

The bass reflex porting comes into play at 68 Hz (driver resonant frequency), producing its typical one-note bass, which I was afraid of.

Although the chart offered by Stefan shows a bump around 125Hz, I couldn't say I could make much of that out, it all seems much deeper, nearer the driver's resonant frequency I guess.

I then took a Solo ULDE and HD250II to the AP and inputted sine waves, adjusting to the frequency I felt bereft of, and after humming it, and adjusting the oscillator to match, found the 125Hz step frequency was as near as damn it.

The problem is the reflex port is outputting a much lower frequency - the room helping it along, and if you drop the amplitude above that, it becomes the equivalent in electronics of a 45 degree phase difference compared to the tuned frequency - and then the baffle-step comes in at 300 - 350 Hz, and the phase reverts back above it.

Being ultra-sensitive to phase throws me, and I can't listen for long after that. The one-note bass slows the tempo, and I always unkindly describe it as wearing lead boots whilst trying to run through mud.

The Pluvia 7 has proven itself as being capable of pleasing me no end, but in a Heath Robinson cabinet. Once "moved" to what the speaker art or science says should be right, it fails miserably for me.

But I hope that Stefan doesn't take all the above badly. I have never been truly satisfied by speakers, and cannot abide production speakers of the type offered in hifi shops - any of them (except the LS35A in a rockwool stuffed stud wall room - read: studio).

The Pluvia 7 can do it. The cabinet design today has let it down.

If it helps, with the resonant "one screw" backs, the drivers hardly moved on strong bass notes, but produced a believable and distributed bass (some call it tight - I call it relaxed). The volume of the bass was I guess some 6dBSPL louder than it was in the sealed cabinet. The sealed cabinet sounded all wrong to me - too bright - all mids and highs; but the loose back simply took all the bass region and made it a flat and relaxed pleasure to listen to.

The bass reflex porting isn't doing that. Therefore it has to be something else.

What I did during my listening was to add to my reference 1 sheet of stuffing; first one sheet parallel to the top of the port shelf - this helped reduce the monotone; then another sheet on top of that - this made the monotone almost bearable but still not right; then I also stuffed a painters tack cloth in each port, which actually made the monotone the same level as the rest of the bass, but sounding like all the bass had been shelved down - it was approaching the sealed cabinet performance.

I then broke the rear panel seal with the port shelf and inserted the single screw again, but because the rim brace had to be cut out the cabinet walls are tight onto the bottom of the back panel, preventing it doing the resonance anymore, so no luck there.

It is either that the porting needs to be much less effective, or the wonderful layered bass I was getting was the resonance of the panel, and little to do with the screw holes. Without knowing the resonant frequency of the back panel suspened from one corner, I will never know.

I don't think a bass resonator will help either: it would need stiff suspension and to be a very massy and hard cone.

I can just imagine Stefan singing Pink Floyd's "What Do You Want From Me" the bit that goes "you're so hard to please", but I know what I want - it's roughly 125 Hz at the same level as the mids and highs, with a bit of extension below it.

Back to the drawing board...


-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: Drewan77
Date Posted: 09 Jan 2017 at 11:10pm
Too cryptic sorry - PGs face is (So) looking out bottom LH side of the last picture but his 2nd album contains D.I.Y. !

-------------
Older than I once was, younger than I'll be
.............................
Andrew


Posted By: Ash
Date Posted: 09 Jan 2017 at 11:12pm
Graham, trying substituting the timber rear panel for a similar thickness cardboard panel. Or try making the whole enclosure out of various conformations of cardboard.

-------------
We do not see things as they are. We see things as we are.


Posted By: Graham Slee
Date Posted: 09 Jan 2017 at 11:52pm
Andrew, it just goes to show how my brain cannot multi-task anymore Wink

Thanks for the idea Ash...

But there might be a silver lining to all this after all Big smile

After a few hammer blows and a well aimed chissel, the port shelfs were removed and the backs reinstated, along with the reference single stuffing sheet parallel to the inside back.

The result is a slot port having depth of the panel (18mm), and using a spread sheet box calculator John C sent me earlier, it was found to resonate at 115Hz. However, it being at the bottom of the box I don't know how to classify the port thickness to its bottom side??

The plodding monotonous bass is gone, and the phase, tempo, flatness, and relaxed sound is now almost the same as the Heath Robinson cabinet, as of yesterday.

It is just missing a certain sharpness now, difficult to explain, but a bit like the difference between hearing a sine wave and a square wave. In fact, it is bass harmonics... I think! Any ideas on that one?

Maybe to block up the slot and drill a round hole instead?



-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: Chris Firth
Date Posted: 10 Jan 2017 at 12:01am
Maybe a support strut in the port - stop the panel above moving?


Posted By: Graham Slee
Date Posted: 10 Jan 2017 at 5:33am
I'm thinking of sealing the back up, replacing the lower batten (brace), and glueing the cut off bit back on the rear panel - this should get it to seal.

Using the "ancient" LCD7 spreadsheet which John C emailed me, for port tuning, I find that two 40mm holes such that the vent length is the material thickness of 18mm (0.75"), the vent tuned frequency for the two holes is 124 Hz. For one hole of that size it is 98 Hz.

Alternatively one hole of 55mm dia will result in the same.

Unfortunately I don't know the attenuation below that frequency, but entering the parameters at http://www.micka.de/en/index.php#ideal" rel="nofollow - http://www.micka.de/en/index.php#ideal - I get -3dB at 92 Hz, after which it rolls off at 24dB/octave, so basically nothing much at 40-50 Hz. Having said that there is room gain, and the harmonics of bass below cut off will still be there.

Where do I drill? Front or rear? I'm wanting to hear my 124 Hz (or thereabouts), where I believe there is currently a dip. If on the rear I will be relying on reflection for audibility, so I guess the holes need to appear in the baffle.

Also considering wrap-around at the wavelength given by the distance between the driver and the holes: this will cause some cancellation - the opposite of baffle-step reinforcement, and probably my chance to deal with annoyance frequencies?

Looking at the chart on page 6 of this topic it looks as if I'm going to end up cutting some of my baffle step gain instead. The offset driver makes a upwards hump around 850 Hz of which I suppose I could sacrifice a little. 850 would set the holes too close to the driver, so it looks as if the frequency I'll lose out on is nearer 500Hz. But thinking about this, the holes by comparison to the baffle size are small, and the difference in frequency from edge to edge of the holes is going to be small, so maybe I don't lose much?

Oh well, faint heart and all that. It's got to be worth trying, and you never know, it might actually work. A search for hole cutters is in order then.



-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: Graham Slee
Date Posted: 10 Jan 2017 at 5:54am
Thinking about this further, electronically speaking double or half the frequency away from your -3dB (or even +3dB) turnover point, and you're within 1dB of flat. This being for a simple first order filter.

So if my port frequency is 125 Hz then at 250 Hz it could be -3dB, and that coincides with the -3dB frequency below the baffle step response, and that should add up to flat. Just conjecture obviously - I have no modelling software. But it makes sufficient sense to me to be in with a chance.


-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: KJFAudio
Date Posted: 10 Jan 2017 at 7:53am
Aaaaarrghhhh!!!!

Stop, don't drill any more holes or cuts anything else up!

I've had an idea!

You'll have to wait till I've got the kids to school though, hang fire till 10ish.

Stefan

-------------
UK/EU dealer for Mark Audio Drivers and Manufacturer of speaker kits


Posted By: Graham Slee
Date Posted: 10 Jan 2017 at 9:00am
OK, but digest my closing comments on page 10.

I shall be out until the afternoon, so will not be doing any "damage" till then Wink


-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: KJFAudio
Date Posted: 10 Jan 2017 at 11:59am
OK, lets step through this. Firstly we need to discount the hypothesis of the rear panel resonating helping in some way. To do that we need to keep everything else the same and try with the panel resonating and not resonating and see what the difference is. My guess and hope ist hat there is no difference. Should be easy to check. This is all assuming we are back to having a sealed box like when you liked the system with just the missing bolts. If you can use some long bits of blutac (you might have to loose the neoprene strips) and stick the rear baffle down so that it can't move then play again.

Asuming that it sounds the same lets move onto the next step. but a bit of waffle first.

The thing my previous models can't take into account really is that the screw holes vary in size. I cut them at 9-10 mm and the inserts would be m8 so does the port act as a 10mm port with a restricted airflow or as a 8mm port with a long flare? Who knows, I just averaged it and called it 9mm

Now inline with Graham's observations, if indeed the the ports are actually acting as 8mm ports that makes a big difference, the f3 would be slightly higher than the slot port I calculated and crucially - and this is what make me think I've got the wrong assumption for screw hole ports - if those ports were indeed acting as smaller ports the excursion would have been much better controlled until further down which seems to be in line with Graham's excursion observation.

Modelling this new slot port (without the shelf) does indeed tie in with Graham's 120ish Hz calcs and listening, however there is a huge amount of port gain here, that aside it does give some gain all the way up to 400Hz so maybe there is a bit more baffle step offset and a bit of extra infill at the 120 ish Hz Graham wants. This is the blue line. The green line is where we were at before my shelf suggestion if my interpretation of Grahams listening are correct and the screw ports are smaller than I originally assumed. It's interesting to note that although Graham's new slot has very low group delay (which would account for it sounding tighter) it does have its group delay centred higher at the port frequency which may account for the loss of focus described. Yellow line is the original sealed box.


 
Anyway onto what to try next Smile

I've tried to devise and incremental way of getting what we want. So firstly put the screws back in the holes, so we are back at having a sealed box. Now take a 10 mm drill bit and drill one hole through the back panel somewhere near the bottom. Everything else should be exactly the same as it was in the Heath Robinson except the screws and the hole. Now assuming my new thoughts about the screw holes and your observations are a bit nearer to what you actually had going on you should have a similar balance but with a bit less bass, add another 10mm hole and you should be pretty close to what you had. Now as you also liked the raised level of the slot (without the shelf) first thing this morning we can now incrementally move toward that and back off at will. Each time we drill another hole, we move the tuning frequency up slightly, the gain of the ports at the tuning frequency increases, the q of the the port pass band narrows and the f3 gets a bit higher. 10 of these ports is as near as damn it to the slot you got this morning. Pictured below are a few steps so you can see what is happening. If you drill one hole to many for your liking just put a bit of blutac in the hole sit back, relax, and listen.



10, 8, 6, 4, 2 and 1 hole you can see 10 lies over my modelling for this mornings slot and 2 lies over my calculations form your observations (for the slot with no shelf), the rest in between.

Here are the port tunings, gain and q



10, 8, 6, 4 have got the best group delay.

Hopefully one of these will get you back to enjoying them more than your LS3/5a again.

When you drill the holes, do it with a backing board so you get minimal break out at the back of the holes other wise we will be modelling irregular flares as well!!

One more interesting thing, I've overlaid the baffle step curve over the top so you have an easy guide to add them together, its a bit rough and ready sorry, remember there is no room gain on here.



Hopefully we can get there now with this incremental approach. If you do I can make a rear baffle with just one hole and a bit of standard pipe with the same tuning (hopefully)

Stefan


-------------
UK/EU dealer for Mark Audio Drivers and Manufacturer of speaker kits


Posted By: Graham Slee
Date Posted: 10 Jan 2017 at 5:53pm
OK, tools at the ready. I'll have a go at the operation doctor. Wink

I thought of the screw holes as 8mm because the rear panel is 8mm clear, which is possibly 8.5mm, and the threaded holes will be 7.5mm on the high points (low points of the screw), so I averaged 8mm.

The 125 Hz was a guess established by listening for the notes, or timbres of the music which I know - which weren't exactly there this morning - and humming until I found the pitch of what was missing, then putting the headphones on and stepping the frequencies to find one with the same pitch.

125 Hz was the closest preset, it being no good trying to home in on the exact frequency with my "cracked voice". It may well be 120Hz, which is the second harmonic of 60 Hz and fourth harmonic of 30 Hz. Whatever the exact frequency, it will be an harmonic of a lower frequency, and that for those who wonder how I know my electronics does the job without a giant floorstander, is how it's done.

Now, regarding the panel not being tight, it might not have been resonance but extra leakage adding to the five open screw holes??

Anyway, off we go with more lovely woodwork LOL


-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: Drewan77
Date Posted: 10 Jan 2017 at 7:18pm
Graham, just a thought - have you considered running a tone generator from your laptop into the Majestic or burning to a CD so you can check specific frequencies by playing them through the speakers? If you had an appropriate mic, you could also check via something like REW or Fuzz Measure & help analyse the impact the new workspace itself is having on the sound.

http://www.wavtones.com/functiongenerator.php" rel="nofollow - http://www.wavtones.com/functiongenerator.php

http://www.szynalski.com/tone-generator/" rel="nofollow - http://www.szynalski.com/tone-generator/

http://onlinetonegenerator.com/subwoofer.html" rel="nofollow - http://onlinetonegenerator.com/subwoofer.html

...I can't get the picture out of my head of you donning headphones, dementedly humming a single note over & over & wielding a chainsaw at those poor speakers Shocked


-------------
Older than I once was, younger than I'll be
.............................
Andrew


Posted By: Graham Slee
Date Posted: 10 Jan 2017 at 10:45pm
I'll try that Andrew, thanks. I do believe I also have various tones in a FLAC file somewhere.

But, the plot thickens...

The cabinets are well and truly sealed, but obviously with no bottom screws to the rear panel, but the other four screws are torqued up tight, so I don't believe there is any resonance. Also, the batten has been replaced at the bottom, and any gaps in the rim "brace" sealed with blutack.

I'm on hole one (no, not the golf course - I don't play!). One solitary 10mm super-clean hole on centre 2" up from the bottom of the rear panel.

Stuffing remains as one loose sheet about 20mm away from the back - like a curtain between the driver and the rear panel.

To be perfectly honest I didn't fully grasp what each hole does, so I wasn't listening for any frequency in particular, just the flow and nature of the music.

I have to say I am very impressed! I haven't played many different tracks yet, but started with Jean-Michel Jarre Areo. Incredibly I found myself listening to nearly all the album - a very long album - and without falling asleep, and that surprised me as I didn't get much sleep the night before (the usual health troubles...).

The thunder storm and the rain effect sounded absolutely real (I actually felt I could get wet!), and there are some muddled voices in places on the album you usually don't hear!

How can one 10 mm (3/8" or 0.375") dia., 18 mm (3/4" or 0.75") deep hole do all this??

I went back and had a re-read, but to be sure, I plugged the dims into the LDC7 spreadsheet, and it gave a port tuning of 32.21 Hz. I'm thinking that can't be right, so I re-read Stefan's post above, and it seems it could be. In fact, the 10 x 10mm holes looks like it gets my 120 Hz, so the one hole must be the lowest frequency of the set.

There is no bloat or drone. Everything is sharp and crisp, and there is bass - tight bass as most call it, but I call it relaxed - you know it's there but it isn't bearing down on you, as the shelf port was doing yesterday. Another thing: the imagery is stable and distributed well left and right, extending to 180 degrees according to my 17 year old son. It also has perspective going back stage around 18" (450mm) further back than the wall.

On Steely Dan's Do It Again, the bass guitar was audible and tuneful as it transits from deep, via a flourish to high, and through another flourish to deep again. This repeating broken only by the chorus. I have had trouble hearing this so far, especially with the sealed box, but it can now be followed as easily as with the LS35A.

What about my 120 Hz? Well, it was there just like all the other frequencies, and in the right measure, so far that is. If I was to be really picky I could say it could do with some further bass volume on the really low notes, and these low notes are stacks lower than the reflex monotone I had yesterday - this being the same driver; the same overall cabinet and the same baffle steps.

But hey, this is a 4" speaker; in fact it's cone area is only 50cm2!

Oh, and the cone hardly moves now, whereas the bass reflex proper saw the cone doing a flap.

I have a lot more listening to do yet, and my conclusions could change, but we'll see.

After that I will drill the next 10mm hole, and repeat the listening tests.



-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: KJFAudio
Date Posted: 10 Jan 2017 at 10:54pm
Thank f**k for that!

I can sleep soundly at last.



-------------
UK/EU dealer for Mark Audio Drivers and Manufacturer of speaker kits


Posted By: Graham Slee
Date Posted: 11 Jan 2017 at 6:23am
Originally posted by KJFAudio KJFAudio wrote:

Thank f**k for that!

I can sleep soundly at last.



If at first you don't succeed, try, try again!

Some people actually follow these topics, post after post, line after line, and when they read about practical, good old down to earth engineering like this, they are able to distinguish between real specialists and "snake-oil salesmen".

The true specialist deserves support, and the people who admire what is written here are the best type of supporters you can get.

They, in turn, get the incredible results not found elsewhere. And you will gain the right type of follower or customer. The type who really appreciates you putting yourself out, because they know they're going to get the best.

Search for history's of different products, and how they were developed, and it's hard to unearth much detail, whereas here people can see multiple pages of it. Most of my developments have been covered by multiple pages: take the big DAC topic; 85 pages!

I was taken aback by how one 10mm hole could make such a difference. It got me thinking why do speakers use such large vents when a 10mm hole can do this?

But if you think about woodwind instruments, the notes on them are made by covering holes with a finger. They are loud enough, and you can hear the notes whether the instrument is being played in front of you or behind your back.

Nearfield monitors are exactly what they say (or should be). They're as close as somebody playing a woodwind instrument would be. If you want to fill an arena with sound from one of these, you obviously need big-rig amplification. Not so in a more intimate setting.

I have seen, and in-fact used, big speakers for sound reinforcement - for scaling up the volume to suit larger venues. Professional quality speakers designed to fill a small theatre often have ports, and are often the size I have seen on so-called mini-monitors. They give a good balance of sound in such a theatre, but simply don't work in smaller rooms - the type we tend to live in and enjoy our records in.

But if you walk into another room you hear the bass! Obviously the reflex porting many of these speakers use is for showing-off, or brag factor. If not, then the humble recorder needs giant holes and very fat fingers! But we here know better.

So that explains why I can hear the bass naturally. I can think of no other explanation. Obviously the average "unwashed" hifi nut is going to laugh at a port smaller than a thimble, but that's his loss.


-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: Graham Slee
Date Posted: 11 Jan 2017 at 6:54am
Help me make sense of where I'm at...

reflex-port tuning chart

Do I have the curves correctly numbered with the correct number of holes?


-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: KJFAudio
Date Posted: 11 Jan 2017 at 7:32am
You do, sorry, should have marked them. Just reread my post as well and there are a lot of typos which may have thrown you. A combination of me typing on a phone, my migraine and dyslexia! I'll correct them later.

The green line is what I've taken a stab at modelling for the screw holes and the blue for your slot vent without my shelf bit.

The tuning of that port is indeed very low and actually doesn't provide that much gain, just enough to force the curve over a bit. I suspect you may eventually prefer a few more holes but this is all interesting stuff. I'd never have arrived here ordinarily.

Yes I hope future surfers come across this thread and find it useful. Thanks for the words of confidence.

Looking forward to the next reports.

Stefan

-------------
UK/EU dealer for Mark Audio Drivers and Manufacturer of speaker kits


Posted By: KJFAudio
Date Posted: 11 Jan 2017 at 9:26am
Fixed my typos, google has got pretty good at catching them these days but dyslexia sometimes means I type words back to front which does confuse it a little!! I do try and sense check everything but you know how time is...

-------------
UK/EU dealer for Mark Audio Drivers and Manufacturer of speaker kits


Posted By: Graham Slee
Date Posted: 12 Jan 2017 at 1:01am
Disexlia lures KO! As we usde to say Wink

Please don't worry about typos, it's what's behind it which matters. I'm useless at typing using a phone - in fact I'm useless at typing, period! Migraines are also a pest, and seem to strike just when you need them least, like when trying to earn a living...

Out of the 10 possible port-holes, only one works (for me). I didn't bother drilling the last two holes - I stuck at 8. There are now seven holes per cabinet stuffed with Blutack!

Today all gang members have been present, and we've played lots of music you'd never hear at a hi-fi show...

Take for example Evanescence! And Anthem by Alex Harvey, and Sheep from Pink Floyd's Animals.

These ruthlessly exposed something unavoidable with full range drivers, and that is intermodulation. Whether that is widely known or not matters not to me, I heard it, and because the "tweeter" is part of the cone, no acoustic "baffle stop" remedy can be applied.

After swapping to the Alpair 7's to find if there was any difference, Evanescence plainly told us there wasn't. We found the solution, or should I say, a solution which works here and for us.

Two ways and three ways often shelve the highs, or if it's an LS35A, the tweeter is surrounded by felt. We decided to shelve the highs using some inductors and a parallel power resistor with them. Calculated to be in the region of 3kHz and with around 5dB attenuation, the highs became considerably sweeter, and it removed a substantial degree of beaming. Because the power resistor is inductively wound (all wound components will be inductive whatever the salesman says), it will roll the highs back up slightly right at the top end.

"Golden eared" Leo was glued to the listening chair for a good hour, which is a good sign. He actually said the result sounded great, which means it sounds great. He doesn't mince his words, and knows his stuff.

He pointed out that few speakers can do the two Evanescence albums we have here justice. He also pointed out that that's why so many go for valve amps with full range drivers - the HF droop valve amps tend to supply, due to output transformer limitations, "cools" the highs. Why should valve amp users have all the fun?

And now a question about the "add-on magnet" on the Alpair 7: is it supposed to ring? I think it might benefit by some gaffer tape being wound round it. By the way, what is the thinking behind it?



-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: Graham Slee
Date Posted: 12 Jan 2017 at 9:31am
Dyslexia joke in bad taste? Not PC? It didn't exist when I was at school but I had plenty of the symptoms. Instead, we got told off or worse - physically punished. I still have trouble spelling common words to this day - highlight the word - right click - choose search Google for "". I did scrape my English exam though. Thank you for listening.


-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: KJFAudio
Date Posted: 12 Jan 2017 at 11:18am
Hey Graham, no not in bad taste, well not to me anyway. I had the same thing at school, they just said i was stupid when I was young and made me stand in the corner all the time! In fact when I had to sit the 11 plus/streaming exams they made me resit them. I got top marks and my class teacher said I must have cheated because I was a bit think. They stood and watched me do them again on my own! Top marks again though. It wasn't until I got to Uni that I realised I might have Dyslexia. I've not got it that bad really and luckily I'm reasonably bright so have found way round it over the years.

Onto the Alpairs with extra magnet. A few reasons which give a bit of choice to the DIYer. With the extra magnet they are a little more sensitive and more linear. Without they are smaller and lighter. Using a neo would make them lighter for the same flux but would make them much more expensive.

Glad you found the amount of holes that worked for you, I had a little gamble with myself that you'd end up at 3, so I've lost, luckily I've only got to pay myself. If we do some sort of Graham Slee kit in the future we could make a rear baffle with all the holes and a set of bungs for tuning Tongue don't think there are any speakers out there like that.

Interesting about your treble solution. This comes back to compromises again doesn't it, of course asking a driver to so so much doesn't make it as linear as narrow band drivers, on the other hand you gain a lot in other areas. I'd not be inclined to filter the driver at all (and don't in any of my full range speakers) I'd much rather have the treble at the level it's supposed to be rather than roll it off to make it cope with a few tracks it doesn't like, as you said yesterday they are quite astounding when you get everything else right. I notice in your collection you have a lot of rock and prog stuff. Have a listen to a bit more acoustic stuff without your filter and hopefully the beguiling fluidity of this kind of setup will shine through much more. I'd be more inclined to use the felt. Some of the better ribbon tweeters like Raal use felt and foam pads and wedges to tune the response in different ways, give it a go.

This is all quite exciting. I'm so pleased that you've got to a speaker you like more than the LS3/5a and still have some tuning to do to wring out the last bit. 

Worth bearing in mind the cost of these compared to a well engineered 2 way, that's a big selling point for wide band drivers.

Stefan


-------------
UK/EU dealer for Mark Audio Drivers and Manufacturer of speaker kits


Posted By: Ash
Date Posted: 12 Jan 2017 at 1:29pm
"Heavy duty music requires a heavy duty driver"- Mr Fenlon

Sounds like the 7 drivers aren't doing too bad for small 4 inch emitters. Evanescence has some heavier tracks probably suited for larger drivers, even for nearfield listening. These MA drivers use carefully controlled driver breakup to extend the frequency response. The more excursion/volume demanded, the more noticeable the breakup becomes (treble beaming and intermodulation distortion etc). It's just like the Proprius and its increased background channel noise with higher dial positions when handling smaller loads. Turn the dials down further and the distortion/noise becomes inaudible. Ultimately, still there but you can't hear it.

You have to use equipment within its limits. You usually have to compromise on something.



-------------
We do not see things as they are. We see things as we are.


Posted By: Graham Slee
Date Posted: 12 Jan 2017 at 3:40pm
The problem in using felt pads and damping is they won't do much at the frequencies where the hard rock music results in distortions. The frequencies are mainly harmonics due to the complexity and number of instruments being played at the same time. The problem frequency range is therefore from around 3,000 c/s (3kHz).

Most musical instruments have fundamental frequencies sub 3,000 c/s, except the piccolo and the highest piano notes, and so acoustic music won't have much harmonic power for the drivers to deal with, and will sound much cleaner.

Rock however, and to a degree, full orchestral works, can contain high level harmonics, which are individual harmonics mixing together to produce further harmonics of varying amplitude. Also, since the introduction of the Marshall and similar amplifiers, much of the tonal structure in modern music is harmonics - distortion in fact. And then we must consider the multiplicity of electronic effects often being driven by keyboards.

I cannot know what my customers listen to, so therefore I must ensure the equipment I design is capable of reproducing whatever is thrown at it. That means whatever I use in listening tests has also to be equally capable.

Also, there is a vast market out there of people who listen to complex music who are honestly turned-off by the hi-fi industry, because they see it as being about special music recorded to show off the equipment, rather than it being about the equipment being able to reproduce their kind of music. As Stereophile put it "We are elitist snobs about our equipment and the music we enjoy" in RIP High End Audio:

http://www.stereophile.com/content/rip-high-end-audio#XTV5o6G0j4ikQDAf.97" rel="nofollow - http://www.stereophile.com/content/rip-high-end-audio#XTV5o6G0j4ikQDAf.97

Anyway, getting back to the problem in hand, the hemispherical wave front of 3k c/s is smaller than the cone size of the speaker, so we can't stick felt anywhere near where it is needed. A bigger driver simply compounds the problem.

Where we cannot reach physically we might be able to rectify electronically, and the simple use of a baffle-step filter to subtly reduce the power of the harmonic rich content can make all the difference between a speaker product being accepted by those wanting to play complex works, and it not being accepted.

When I bought my first "hi-fi" aged 18, I was told I should not play my kind of music on it, but gentler music. I see that still holds, and then those in the industry wonder why they aren't making anything out of it. We could change things if we tried, and I for one, am trying to do that.

Noise is not a compromise Ash, it is a matter of physics which those who design the silicon we all use, know about, and the hi-fi industry is generally ignorant about, and that is why you and many others don't understand the subject very well.

Music consists of rapid changes, but low noise circuitry is slow - it cannot slew the volts required to keep up with the music without distortion, or instability, and sometimes both.

So, to reproduce music with great clarity, and all its nuances, you exchange boastful low-noise specifications for speed as the noise will be hidden by the music. Admittedly you'll hear it if you don't use the equipment for its intended purpose, and instead listen out for the noise, especially if the transducers are right up against your ears.

The thing about the Proprius is that it is intended to drive loudspeakers which are some distance from your ears, and so it isn't designed for the lowest noise like a headphone amplifier is. But then, a headphone amplifier cannot drive such low impedance loads to the power required by loudspeakers whilst having the speed of the Proprius.

-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: Ash
Date Posted: 12 Jan 2017 at 8:49pm
Originally posted by Graham Slee Graham Slee wrote:

Noise is not a compromise Ash


Originally posted by Graham Slee Graham Slee wrote:

(Proprius) isn't designed for the lowest noise like a headphone amplifier is. But then, a headphone amplifier cannot drive such low impedance loads to the power required by loudspeakers whilst having the speed of the Proprius.


So uhhh... it's kind of a compromise? Embarrassed (Just use the equipment within the design constraints it was intended for and problems are avoided). But yeah, I don't pretend to have an in-depth understanding of these topics.


-------------
We do not see things as they are. We see things as we are.


Posted By: Graham Slee
Date Posted: 12 Jan 2017 at 10:36pm
Noise in amplifiers:



The total noise figure is then multiplied by the square root of the bandwidth, and the test gear should see the same.

You can see that by narrowing the bandwidth, the noise is reduced. And by increasing the bandwidth to do the speed required, increases the noise.

You can see that a large resistance has more noise than a small one. Input impedance is mainly set by resistors and input impedance needs to be large enough so as not to load the source - which by being large enough means it will be more noisy than some value that would load the source down.

Slew rate (speed) is often abruptly limited by absolute phase stages. These are often very low noise, but the distortion produced once the threshold is reached is gross. Inverting stages using "virtual earths" do not suffer such an effect, but require larger resistors, which are more noisy.

Therefore appropriate technology should always be used for a particular purpose, and the purpose under discussion is music, which dictates speed over noise.


-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: Ash
Date Posted: 12 Jan 2017 at 10:58pm
Live recordings would contain some "noise" anyway, even if the electronics reproducing the signal are very quiet. So focusing too much on low circuit noise would ultimately be fruitless. Slew rate = Speed = timing, preservation of which is necessary for rhythmic, focused and in-phase sound. I'm guessing slew rate is partly why you're so fond of vinyl over digital. I think you have implied in the past that vinyl has better phase fidelity than digital audio. You have also said that if you limit the bandwidth too much, "you've killed it" (timing/focusing suffers). HD800 is an audible example of this, where the highly resolved low level noise floor has more focus than the music itself?

Your goal was to get the balance/trade-off just right. Not too noisy and not too bland/withdrawn?


-------------
We do not see things as they are. We see things as we are.


Posted By: Graham Slee
Date Posted: 13 Jan 2017 at 5:34am
Phase integrity is the phrase you're after, or even flat phase. Neither are fully achieveable, and so recorded music can never sound exactly the same as it did to the ears listening to it being recorded - but we can get close.

I've also found that if we put too many changes of phase direction into a system - or just one system component - "you've killed it" (even at frequencies beyond our hearing). This surely affects speaker and headphone designers considerably, and in a world free of marketing and budgets (and red tape), where the understanding designer is free to design, just two changes of phase at either end of the spectrum - which are unavoidable - would improve matters greatly. It's one of those cases where the more you do, the worse it gets.

Vinyl is not necessarily better at slew-rate over digital, it's what you do with the emerging analogue signal that makes it what it is, provided the emerging signal is right in the first place.

Vinyl depends on a vibrating system which will never be 100% right, and so requires a lot of care so as not to emphasize its flaws. And here we can swing right back around to the subject of speakers, as what might sound right with a digital source, may not sound right with vinyl as its source.

It is not only speakers obviously, but vinyl can be a good test of speakers, and possibly why a good proportion of what's on offer today, would have never surfaced some 40-50 years ago.

All the gains you can make electronically can be undone by a speaker which does not obey phase flatness as defined above (the two changes of phase at either end), and this is why I'm spending a lot of (too much) time on these speakers. One major factor being vinyl surface noise which can be greatly reduced by good engineering practice in the electronics, only to be undone by the speakers (or headphones), and the result is emphasized surface noise.



-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps


Posted By: Richardl60
Date Posted: 13 Jan 2017 at 7:03am
Presumably Graham until you get your speakers sorted your other R&D works is on ice (e.g. Acession MC)?


Posted By: KJFAudio
Date Posted: 13 Jan 2017 at 4:07pm
I agree with most of what you are saying Graham and of course we must try and design equipment to have the widest possible appeal and that is one of the things I have always tried to do as well, however speakers (or any transducer) can never ever really be that straight wire with gain and are full of so many more compromises to get the to do what they do that the rest of the electronics. It shouldn't just be down to tastes in music though no. A speaker should be a window to the recording as much as possible. However as you have learned during this little journey just changing one thing screws things up. You can never ever know where the user is going to put the speaker and as you now know, even before you embarked on this, a few inches one way of the other changes the balance hugely. Designing and marketing a speaker is largely about trying to put groups of potential use together and then broadly designing a speaker to work for that given space. We now for example have a speaker that works in your space. The Missions I just repaired for a customer however do not work where I have put them today to run in because they require some more boundary loading than where I have just plonked them. My kids said how crappy they were when I played some tunes last night. Of course that isn't fair the speakers that are usually in my room have been specifically designed to be there and represent several thousand pounds worth of components. The other side to that coin is that different speaker topologies have a different acoustic foot pirnt. You mention the two unavoidable phase shifts of loudspeakers, those created by the filters that the cabinet creates at either end of their usable spectrum. Of course in a typical 2 or 3 way speaker there is a phase rotation at the top and bottom of each driver created by the filters (crossover) as a speaker designer you'd obviously try and make these sum to zero rotation, that is never perfect and is only mathematically zero, they are still not linear phase devices. As you said this sort of phase anomaly means (in terms of coherence) that 'You've killed it' . There is also the change in dispersion characteristic, comb filtering as wave fronts combine... Of course with a full compliment of drivers you can solve some of the problems associated with single emitters, one of the major ones being power handling, although this has largely been sorted in the last 5 years or so. inter modulation is another problem. It is a lot to ask of a driver, no getting away from that. 

Point is that a speaker will always be compromised, more so than another piece of the audio reproduction chain. It's impossible to to get an idea of what customers are listening to at home. Some of your users will want to listen to dub-step at 115 dB SPL @18Hz. There is no way of hearing what they are hearing!!

What can be done? Well the felt will help a bit because the small wave front does travel across the baffle as it expands and that causes ripples in the FR. You can go with your method and to that end it may be worth offering customers some sort of pot they can adjust on the back panel. The other thing is damping the cone, these cones are pieces of highly designed and tested metallurgy but at the end of the day it is our mid level driver at this size. The Alpair is a much better cone an the paper is again much better damped although not as accurate FR wise, The MAOP 7 is the best damped although very expensive.

Chose you poison I suppose, no getting away from that, But if the point it to be able to hear component choices and design decisions with well resolved micro and macro detail then I'd have thought a more transparent system with less phase rotations than an equivalent 2 way would be the way to go and just accept that the are one or two really busy tracks that when the wick is turned up loose some integrity and focus. Arguably getting system to reporduce unatural harmonics while them themselves are prone to harmonics is a big ask and there is no real way on knowing what that sounded like. On the other hand we all know what drums, guitars and horns etc really sound like.

Stefan


-------------
UK/EU dealer for Mark Audio Drivers and Manufacturer of speaker kits


Posted By: Ash
Date Posted: 13 Jan 2017 at 9:24pm
Sounds like I'm going to have to buy and sell my way through half the Mark Audio driver range. I envisage you now rubbing your hands together enthusiastically.

-------------
We do not see things as they are. We see things as we are.


Posted By: ICL1P
Date Posted: 13 Jan 2017 at 11:33pm
Originally posted by Ash Ash wrote:

Sounds like I'm going to have to buy and sell my way through half the Mark Audio driver range. I envisage you now rubbing your hands together enthusiastically.

We wouldn't have expected otherwise. It's the way you do things.

-------------
Ifor
=====
Reflex M & ACCESSION M, CuSat50, Majestic DAC, a Proprius pair.


Posted By: KJFAudio
Date Posted: 14 Jan 2017 at 8:16am
Originally posted by ICL1P ICL1P wrote:

Originally posted by Ash Ash wrote:

Sounds like I'm going to have to buy and sell my way through half the Mark Audio driver range. I envisage you now rubbing your hands together enthusiastically.

We wouldn't have expected otherwise. It's the way you do things.


I've always said how much I like Ash.

-------------
UK/EU dealer for Mark Audio Drivers and Manufacturer of speaker kits


Posted By: Ash
Date Posted: 14 Jan 2017 at 10:03am

As you wish, Darth Stefan


-------------
We do not see things as they are. We see things as we are.


Posted By: Graham Slee
Date Posted: 15 Jan 2017 at 7:01am
Originally posted by KJFAudio KJFAudio wrote:

The Missions I just repaired for a customer however do not work where I have put them...


That seems to be the biggest problem here too!

Even the LS35As don't work where I put them. The boom was ridiculous! I've been having great difficulty understanding why. After all, these speakers are designed to work in the back of broadcast vans, as well as at work stations.

However, there is a lot of acoustic tiling in broadcast vans because vehicle metalwork isn't all that audio friendly, so my thoughts turned to the room itself, because if I can't get the LS35A to perform, I don't have much chance with anything else.

Your guess on porting didn't work here - it was too boomy. This isn't because you guessed wrong - it is the room's fault!

I cannot accept it is the shape or size of the room, because I've had far worse yet still got a much better performance from lesser speakers. 20 years ago I had a small flat and a pair of JPW Sonata's, which performed better stood on chairs in my kitchen-diner-living room. That building was built early last century. As a youth I had better performance out of some Marsden Hall speakers stood on the concrete floor of a 1950's built house. I've heard massive JBL bass reflex speakers in the back of the YRN outside broadcast truck, smaller than the room I'm now in - and the sound was spot-on. And the last two locations of the LS35As sounded good enough to pin-point component changes in the electronics nobody else would have dreamt of.

So what is the single difference between this room and the others?

The answer is Gyproc SoundBloc! If only I'd known. It would seem that in making a product to attenuate noise passing through it, it has some highly reflective properties at low frequencies. It would seem they are designed to resonate at low frequencies, such that sound is reflected back into the room instead of being absorbed (passing through to the other side). This obviously gives it its status as a "very good" soundproofing material for builders to comply with part E building regulations.

It also has better soundproofing performance at high frequencies, and here these will be absorbed rather than reflected.

Add these two together and you should hear boom and HF loss. And this is exactly the result I got from the LS35A.

Would combing it on like tiles have improved the situation over dot and dab? I couldn't say. I tried pumping expanding foam into the dot and dab voids in several places - it didn't make any difference.

Having established SoundBloc as being the major culprit, it does make me worried because some high fidelity enthusiasts will find themselves in the same predicament if they move to a part E constructed home.

Like me, they will need to compromise something or other to be able to place their speakers to obtain the best sound balance. I have now moved mine to either side of the exit door, which is similar to a single patio door plus a half width fixed panel, both full height double glazed and uPVC framed.

They are also along the longest wall with the nearest speaker side to a wall being 80 cm away. Spherical waves, which the low frequencies are, will now encounter a mixture of structures to the rear of the speakers (including thin unlined curtains) and might not boom so much. In the firing line the sound waves will encounter racks, component drawers, my desk and workbench (and me), but behind all this is the dreaded SoundBloc.

Possibly the worst track to choose (or the best if you want to hear boom) is Uriah Heep's "Circle of Hands", and is what I chose for testing the room positioning. The piano in the latter part was completely obliterated by the boom in the original speaker positions. Now it is perfectly audible.

Does it sound musical? Jaded ears from all this messing about can't say yet. I will report my findings soon, and if positive, these adventures can start all over again. Yes, it's back to square one by the looks of things.



-------------
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps



Print Page | Close Window

Forum Software by Web Wiz Forums® version 12.01 - http://www.webwizforums.com
Copyright ©2001-2018 Web Wiz Ltd. - https://www.webwiz.net