New Posts New Posts RSS Feed - 1970s Design Indulgence
  FAQ FAQ  Forum Search   Events   Register Register  Login Login

Welcome to the Graham Slee Audio Products Owners Forum

 

Open to all owners plus those contemplating the purchase of a Graham Slee HiFi System Components audio product and wishing to use this forum's loaner program: join here (Rules on posting can be found here)

This website along with trade marks Graham Slee and HiFi System Components are owned by Cadman Enterprises Ltd


1970s Design Indulgence

 Post Reply Post Reply Page  <1 132133134135136 345>
Author
BAK View Drop Down
Senior Member
Senior Member
Avatar

Joined: 14 Mar 2010
Location: Kentucky, USA
Status: Offline
Points: 1744
Post Options Post Options   Thanks (1) Thanks(1)   Quote BAK Quote  Post ReplyReply Direct Link To This Post Posted: 18 Jan 2021 at 5:49pm
Originally posted by AlienRendel AlienRendel wrote:

Being honest, a lot of what is in this thread is just technically over my head.  And that's one of the problems with being a layperson investing in audio equipment - I don't have the background to look at circuit designs and know what is better, so I can only trust my ears.  

However, having an equipment maker be this detailed and transparent about their product instills trust, even if I don't comprehend all the particulars.

 And, all should realize that sound quality is Graham's ultimate goal in all his designs.
He does not just use theory or scientific math to prove a system works as intended, he does very critical listening tests in a broadcast studio designed listening room.


Edited by BAK - 18 Jan 2021 at 5:52pm
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!
Back to Top
Graham Slee View Drop Down
Admin Group
Admin Group
Avatar
Retired

Joined: 11 Jan 2008
Location: South Yorkshire
Status: Offline
Points: 16298
Post Options Post Options   Thanks (1) Thanks(1)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 18 Jan 2021 at 9:18pm
I believe that some amplifier designers fix the symptoms rather than the cause. Usually, the cause is a not-so-bright idea.

These 134 pages are a long journey to find and explain these causes. I'm sure it would be an excellent reference to others if edited and appropriately indexed.

I am a bit disappointed by the lack of information on these things in books written by "acclaimed" designers. However, I think I know what their answer would be.
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps
Back to Top
Sylvain View Drop Down
Senior Member
Senior Member


Joined: 18 Jan 2010
Status: Offline
Points: 481
Post Options Post Options   Thanks (1) Thanks(1)   Quote Sylvain Quote  Post ReplyReply Direct Link To This Post Posted: 19 Jan 2021 at 7:52am
...These 134 pages are a long journey to find and explain these causes. I'm sure it would be an excellent reference to others if edited and appropriately indexed....

I ditch my ' music centre'' in 1985 and began the journey of single components so called ''HIFI''. By 1990 something bite me and i sought better and started subscription to many HIFI magazines and began the yearly visit at many dedicated listening room at Audio retail stores. And perhaps that was the biggest mistake and after yearly or very three years change of Amp and CD players and new components , Power supply and DAC .....has not learned until 2015 when I built the Genera and read the Forum. Indeed their is much to make a HIFI SALEMAN ditch his Marketing strategy. My 'Audiophile ' quest began in 2012 when I built 'Genera ' and  2015 the Proprius in the Hifi not rack but behind the speakers.
Back to Top
Sylvain View Drop Down
Senior Member
Senior Member


Joined: 18 Jan 2010
Status: Offline
Points: 481
Post Options Post Options   Thanks (0) Thanks(0)   Quote Sylvain Quote  Post ReplyReply Direct Link To This Post Posted: 01 Feb 2021 at 10:24am
Graham

I read your many Research derive  useful information to design and build natural sounding balance audio.  
You have done more research than I have read since my monthly Audio magazines subscription in 1990 and truly value teh knowledge you impart freely. 

I am seeking to read you on small Isolation Transformers to deal with various circuit circuit interferences and proper use of earthing ...star or other ....

May i beg for a layman language, i am NOT so posssessed in technical but respect Your NOVO headphone, Genera  Phono Pre amp Proprius and still await a better elaborated PSU circuit board with DAK and I small look for a transformer and box.
Back to Top
Graham Slee View Drop Down
Admin Group
Admin Group
Avatar
Retired

Joined: 11 Jan 2008
Location: South Yorkshire
Status: Offline
Points: 16298
Post Options Post Options   Thanks (1) Thanks(1)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 28 Feb 2021 at 10:04am
For amplifier power supplies, peak rectifier diode current is ten times load current but averages at 6.2 times. Here we have the origins of transformer current calculation, which few if any, impart as knowledge today.

Earlier on in this topic, we learned that the phase angle of charging current lasts for only a fraction of the ripple voltage cycle, that being 1/10th, and thus the current during that period must be ten times that of the load current.

However, this is the peak rectifier current, but we see what looks like a single positive going "sine wave" each pulse. Indeed the average must be like the rms? Instead, due to the silicon diode's characteristics, it ends up being 0.62 rather than 0.707.

We can derive two things here. The first relates to transformer power, and the second to capacitor ripple current.

Relating ac to DC, the DC voltage is 1.414ac. Now, here we learn the average current is 0.62. As we can only pull current - it can't be pushed - then working DC to ac, the ratio must be inverted. 1/0.62 = 1.613 (per amp) is the ac current drawn.

A few years ago, I remember being lectured about being wrong about this. It worked to my advantage, and so I accepted the "gift" I was being offered. Neither did I want to undermine the examiner's qualifications, whatever they were. He was, after all, working for a Certified Body that ensures consumer safety.

The problem in assuming the value of 1 is the real-life result is wrong, and on test, the physics rules the day! It is unfortunate then that those tasked with upholding technical decisions made in error do so ignorantly because that is their job. "Jobsworths" plain and simple.

By establishing truth as found through statements of facts (6.2 being the average of 10 in the rectifier in the capacitive input filter), we are empowered to get the results we wish. A case in point was the first transformer custom made for the 5V 3A power supply covered in another topic. Employing the falsehood of Iac = IDC, the output would not hold up.

Should we want 2 x 40 watts into 8 ohms, we multiply 40 by eight and take the square root to find the ac voltage (17.889V). We then divide the voltage by load to obtain current (2.24A) and multiply by two because we have two channels.

We find we require 4.5A DC, which is 4.5 x 1.613 = 7.3A ac. We know from V ac = 0.707 V DC that the nearest preferred value for 70 to 72 volts DC is 50V ac.

We also know that power is V x A, and so 7.3 x 50 = 365 watts. To differentiate between ac and DC, ac watts are called VA, so the transformer rating must be 365VA.

We will only ever use that power on transients. Knowing a transformer will deliver short bursts of additional energy, music and voice reproduction at full rated power will not be used. However, we must be able to test full power, and so to reduce the transformer's VA rating, we need to understand safety limits. Self determined that to be 70%.

70% of 365VA is 255VA, and so a 300VA transformer is quite sufficient. My colleague John has another make of amplifier in for repair. It is called a 4040, which one would expect to do 40WPC. It has an 80VA transformer. It will be easily capable of the 1/5th to 1/10th power of real music at an averaged level. The reservoir capacitors should carry the transients but fails the analyser test at 40 WPC by a long chalk because it isn't continuous.

So, to obtain a continuously rated 40WPC, the transformer must be at least 255VA, and perhaps 225VA (a "preferred value") might also hold up. 300VA will do 2 x 50W into 8 ohms, but is often the size found in 2 x 100W into 8 ohms amplifiers, and usually only 225VA.

Capacitors are often cheaper than transformers, and given plenty of paralleled devices, there might be sufficient energy available to cover the transients. However, there is a genre of music that consumes transient energy availability, often faster than it can be replenished. The genre is rock.

For the sake of rock, which the now older generation of amplifier customers seem partial to (me included), then sizing the capacitors to the correct transformer rating should give the desired result.

Capacitor size (in uF) decides the ripple voltage, and in this amplifier, 4700uF is quite adequate. What we do need to consider, however, is the capacitor's longevity.

It might still not be evident why an amplifier should be left on - powered-up continually - but listen into the third day of on-time, and your ears should let you know. It was also an IEC60950 test that a power supply must work continually past 72 hours because before then, it was not considered to have reached equilibrium.

Capacitor life depends on ripple current rating. Previously we have taken that to be ten times the actual current drawn due to phase angle conduction being 1/10th of the time. We now consider the averaged peak rectifier current because that is where the charge current comes from.

The value of 6.2 must be correct because the physical tests confirm the 1.613 x DC, ac requirement (as above).

As the amplifier will be switched on in its quiescent state (idling), the only current being drawn is the output stages idling current and the voltage amplifier current. This will be in the region of 200mA. When multiplied by 6.2, this will be the constant idle capacitor ripple current, and so it is 1.24 amps.

One choice of capacitor shows a lifetime of 4600 hours (half a year) at 4.3A ripple and a temperature of 105C. If operated at 50C, however, it will last 32 times longer or about 16 years!

If we exceed its ripple current, however, its lifetime is comparatively shortened. In a class A or a much over biased class B amplifier, this will be the case. In our optimally biased class B amplifier, however, it will only be exceeded when reproducing music.

If we draw full output current of 7.3A, the ripple current grows to 45A, vastly exceeding the 4.3A ripple rating. We can often get away with 2.5 times the ripple, but not over ten times, as this indicates. But also remember the operating temperature is much lower than specified.

We should also consider that even with rock music, the transients are short-lived, even though more dominant. There is no hard and fast rule, but perhaps the meters can help?

The meters I have included in the design indicate the quiescent voltage of the output stage such that gm-doubling is avoided. Still, crossover distortion is minimised by the emitter resistors matching the thermal voltage. They also rather conveniently measure the average emitter voltage due to the signal, and as such, indicate the actual average power (given a bit of translational maths).

Even at loud volume and approaching transient clipping, the average indicated power is less than 2 watts. I will spare you the maths, but that equates to less than 1 amp (both channels). If we take the 6.2 multiplier, we can see the current is within 2.5 times the 4.3A ripple rating.

However, the table on page 128 seems to make a mockery of these figures. But let's say we use the factor of 10 rather than 6.2. Our quiescent ripple grows to 2 amps instead of 1.24 amps. The rectifier pulse rate is 100 per second, and the charge time is equivalent to 1kHz. The table indicates a ripple frequency of 3311Hz. Divide 3311 by 1000, gives 3.311, and 3.311 times 2 amps = 6.622 amps. The table indicates 6.46 amps.

The difference here is 4700uF versus 6000uF and 200mA versus 195mA. With the 4700uF capacitor, we have a little more ripple voltage than the 6000uF in the table, which helps account for the difference.

What the table indicates, however, is that ripple current increases with frequency, and so we must choose a capacitor that features more ripple current at a higher frequency. At 20kHz, the selected capacitor has a ripple current rating of 15.8 amps, and so we can see it heading in the right direction.

By taking 6.2 as the corrected multiplying factor instead of 10, having proven it by the rectifier current ratio (0.62 and 1.613 depending on direction), we can see our maths is still correct (or close enough).

The capacitor I have chosen is a Kemet having part number PEH536PDH4470M and can be found here: https://docs.rs-online.com/2d98/0900766b81717268.pdf
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps
Back to Top
Sylvain View Drop Down
Senior Member
Senior Member


Joined: 18 Jan 2010
Status: Offline
Points: 481
Post Options Post Options   Thanks (0) Thanks(0)   Quote Sylvain Quote  Post ReplyReply Direct Link To This Post Posted: 28 Feb 2021 at 5:41pm
We will only ever use that power on transients.......... 
Knowing a transformer will deliver short bursts of additional energy, music and voice reproduction at full rated power will not be used. However, we must be able to test full power, and so to reduce the transformer's VA rating, we need to understand safety limits. Self determined that to be 70%.

Transient dynamics makes the difference between a 'mediocre'', ''good'' and a High Fidelity  Realism and musicality reproduction   

I shall read this bit a few times over the coming week and many many thanks 

Back to Top
Graham Slee View Drop Down
Admin Group
Admin Group
Avatar
Retired

Joined: 11 Jan 2008
Location: South Yorkshire
Status: Offline
Points: 16298
Post Options Post Options   Thanks (0) Thanks(0)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 14 Mar 2021 at 4:58pm
It's been a few weeks since my last update on the power amplifier stage, so here's the circuit of a channel in its latest form.

I am now re-doing the printed circuit board layouts, and hope to be testing this within a couple of weeks.

power amplifier circuit

Right click on image and select view image for a clearer picture


Edited by Graham Slee - 14 Mar 2021 at 7:38pm
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps
Back to Top
 Post Reply Post Reply Page  <1 132133134135136 345>
  Share Topic   

Forum Jump Forum Permissions View Drop Down

Forum Software by Web Wiz Forums® version 12.01
Copyright ©2001-2018 Web Wiz Ltd.

This page was generated in 0.125 seconds.