New Posts New Posts RSS Feed - the most weird problem I have ever had with HiFi
  FAQ FAQ  Forum Search   Events   Register Register  Login Login

Welcome to the Graham Slee Audio Products Owners Club

 

Open to all owners plus those contemplating the purchase of a Graham Slee audio product wishing to use our loaner program: join here
Subscribe to our newsletter here (Rules on posting can be found here)


the most weird problem I have ever had with HiFi

 Post Reply Post Reply
Author
davidismynaim View Drop Down
New Member
New Member


Joined: 31 Mar 2021
Status: Offline
Points: 7
Post Options Post Options   Thanks (0) Thanks(0)   Quote davidismynaim Quote  Post ReplyReply Direct Link To This Post Topic: the most weird problem I have ever had with HiFi
    Posted: 31 Mar 2021 at 11:24am
I have just seen this, and it may explain the most strange problem I have ever had during my 35 years of enjoying HiFi.

My system is a little complex: Mitchell GyroDec with Roksan Tabriz arm and Shure VST MM cartridge. into my (soon to be replaced with a GramAmp2+PSU1) Cambridge Audio based modified phono stage by Jez 'Arkless' into a DBX Venu 360 DSP loudspeaker management system, which handles room correction but most importantly provides a digital crossover (converts the analogue signal to digital, provides sophisticated DSP options and then output via up to 6 DACs to drive 2/3 way speakers. My loudspeakers are Innersound Isis hybrid electrostatic panels and transmission line bass. 

While setting the gain structure on the DBX, I was getting a very bad and loud digital like distortion coming to the speakers when the gain was raised close to around 0db. Loud but not very loud. 

I swapped out the els for a small pair of conventical cone speakers and there was no problem. 

I then realised when the distortion occurred I could see sparks in the electrostatic panels. So they must be arcing!

I also saw the output meters on the dbx showing the high distorted noise. So I concluded the fault conditions must be upstream of the speakers. 

I moved the speakers as far away from the electronics as possible and the fault went away. 

My conclusion, however far fetched it seems, is:

The els must be interfering in air (with the large voltages involved?) with the sensitive phono stage (note it is MM not MC) or now I think maybe the Tabriz arm?

The in air interference from the els to the phono stage might have caused the phono stage to send an ultrasonic signal to the dbx beyond its ADC design, resulting in the ADC to digitally 'crash', leading to very loud distortion through into the speakers. Pushing them into a bad state, resulting in arcing. 

But I wanted to check with you guys. Especially after reading this post by Graham


Originally posted by Graham Slee Graham Slee wrote:

Check the arm tube is actually screening the wires inside it.

This is an arm I had to eventually reject because it relies on the bearings to conduct the ground to the arm tube, and the film of lubricant prevents it, even after nearly 30 years!

arm base conducts ground

As can be seen, connecting arm base to arm base components shows conducting ground.


arm tube not grounded

Shame the arm tube has no protection at all (open circuit)

Any ac mains magnetic field in the vicinity will be picked up as hum by the wires inside the tube.

Any ac mains vibration of the arm tube is fed into the magnetic circuit of the cartridge, and is worse with MM because of the greater number of turns of its fixed coils.

However, there's worse to come, because neither is there any shielding at radio frequencies, and electro-magnetic waves can set-off oscillations at any point in the amplification chain, often emitting a buzz.

So please check for ground continuity. We need to realise we are dealing with 19th century mentality in the 21st century, with its crowded electro-magnetic spectrum.




Back to Top
Graham Slee View Drop Down
Admin Group
Admin Group
Avatar
Telling it as it is

Joined: 11 Jan 2008
Location: South Yorkshire
Status: Offline
Points: 10673
Post Options Post Options   Thanks (1) Thanks(1)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 31 Mar 2021 at 7:35pm
Nothing far fetched about spark transmission - the way radio started - and any high gain stage is bound to amplify what's picked up on its input wiring - especially via the arm tube.

Did I tell you the story about my early experiments aged 9? I'm sure it would bore the pants off everybody, so we'll leave that for another day ;-)

Do however look for possible faults on the ESLs. I don't think they should arc so easily.

I found this page which has some info on arcing/sparking. It's about Quad ESLs but it might help.

Not simple enough for Google-Bot to understand...
Back to Top
davidismynaim View Drop Down
New Member
New Member


Joined: 31 Mar 2021
Status: Offline
Points: 7
Post Options Post Options   Thanks (0) Thanks(0)   Quote davidismynaim Quote  Post ReplyReply Direct Link To This Post Posted: 05 Apr 2021 at 7:55am
I don’t think the speakers have any fault. When I increase gain to the point of loud to very loud the arcing occurs only when the phono is in the circuit. If I play digital the speakers are fine and do not arc. 

When the fault arises it is clear from the dbx meters the distortion is occurring at least from it or upstream and this distortion is what then causes the sparking at the speakers. I assume because it is both loud and contains some very high frequencies. 

I have tried it with your phono stage, an old gram amp 2 in the plastic box and it suffers the same problem. In fact there is a new issue too. A very quiet ‘digital chattering’ sound coming from the speakers. Must be some other interference the arm/phono leads/phono circuit is picking up. I turned off my wifi router, mini pc and it continued. So haven’t isolated the sound. Could be in the mains I guess and bleeding though the psu1?

So here is my question. Anyone know how to better earth the Tabriz arm tube. Graham, looks like you had one. Did you try and meet earthing with it or swap to a different arm? 

Any value in re-housing the venerable gram amp 2 in a metal box to give it better noise rejection? I don’t get the digital chatter with my two other mm phono stage. A loan project phono box 2 and my old Cambridge audio phono stage modified by arkless. 
Back to Top
Graham Slee View Drop Down
Admin Group
Admin Group
Avatar
Telling it as it is

Joined: 11 Jan 2008
Location: South Yorkshire
Status: Offline
Points: 10673
Post Options Post Options   Thanks (1) Thanks(1)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 05 Apr 2021 at 9:40am
Please take the following as me being helpful.

The dbx meters don't measure distortion - they indicate overload.

Distortion measurement is done on a distortion analyser - it might have one in its software (I doubt) - but you say 'meters'.

Therefore, the transients must be hitting the rails - the signal swing at some point in some circuit exceeds the supply voltage of that circuit, but obviously, it cannot exceed it, so it reaches that voltage, 'travels along it' until it comes back down again - that is called clipping.

You do not get clipping in digital audio because it cannot exceed 0dBFS (full scale). The signal might have clipped in the mastering, but that's a sound quality issue and nothing to do with exceeding 0dBFS - which it can't.

Vinyl reproduction is an entirely different matter and is hard to explain even to engineers unless they're versed in it.

A cartridge's output is stated in mV (at 1kHz) and refers to 0dB of a test record. However, the output from vinyl can reach +14dB (5 times) at mid frequencies.

Whereas digital 0dBFS can work within a 5V supply, the output from vinyl on transients would clip all the time at such a supply voltage.

Should a CD be recorded in the same way as vinyl works, you would have a very quiet CD (applies to all digital files). However, that was the original intention.

If we take 0dBVU to indicate 775mV, as it was in professional and broadcast audio, before it became a bedroom plaything, then vinyl transients might reach 3.875 volts.

If digital respected 0dBFS as 775mV, then the transient-less music would be 155mV. Spot the problem?

The digital equipment manufacturers do not have any reference to work with at all! 0dBFS can be 910mV, 1V, 2V, 2.2V - you name it!

But if you have an input rated for 2.2V, it would clip that input circuitry if it exceeded that voltage (which it won't if digital), and give it 0.775V, which it can handle, then on a hot vinyl transient, it hits 3.875V - it clips!

The problem is exacerbated because of the lack of input latitude of most circuits designed with digital audio sources in mind. Rather than placing a level control at the input end, the tendency is to put it at the output end (as I have seen in numerous circuits over my long career). Therefore, regardless of the level control position, the clipping still happens.

Please take the above as the primer, and then we can see where the problem might be.

Not simple enough for Google-Bot to understand...
Back to Top
davidismynaim View Drop Down
New Member
New Member


Joined: 31 Mar 2021
Status: Offline
Points: 7
Post Options Post Options   Thanks (0) Thanks(0)   Quote davidismynaim Quote  Post ReplyReply Direct Link To This Post Posted: 05 Apr 2021 at 12:09pm
I understand much of what you are saying Graham. I probably phrased the issue poorly. I know the DBX meters are not measuring distortion. But they do show the signal jump up when the distorted noise occurs, it is best described as a digital crackle/buzz, which is made worse with the crackle of the arcing sound of the speakers. I obviously don't want to leave this condition on for any amount of time as it could permanently damage the mylar of the speakers, even though the designer, Roger Sanders is proud of the resilience of his speaker panels, claiming they do not fail in the way the old Quad's are known too. 

I do have a short video clip I could email you, though it would be too big for this site. What you can see, hear is as I gradually increase the output stage gain on the dbx, the noise occurs and the meters jump up, reflecting the noise is being generated 'up stream' of the speakers, i.e, it is not that the speakers are being overdriven by a clean signal and then arcing, nor is the volume anywhere near a level where you would expect clipping - even given your point about how analogue differs from digital in this regard. Also, the speakers are capable of very loud volume and do not arc under normal conditions. 

It is this that has led me to deduce (further endorsed with two other electrical engineers / audio designers who I have discussed the issue with) that it is air-borne interference from the speakers (high voltages etc) playing loudly, affecting the sensitive phono stage in some way (possibly the poor earthing characteristics of the tonearm - a roksan Tabritz, or phono cables or phono stage) this then amplifies the interference and gets to the ADC of the DBX, which in turns operates outside of its design parameters resulting in a very abnormal signal to be amplified through to the speakers, which then arc due to the loudness and unusual frequency being amplified. 

I know this seems very odd, but no one has come up with any other explanation. I am wide open to other ideas. 

I was especially interested in knowing if you felt the Rokan Tabriz could be improved with a better earthing arrangement, I guess requiring some new internally wiring (beyond my skills for sure) or if in your opinion its earth scheme is so flawed, one would be better off with a arm with a more 21st century approach to earthing? 
Back to Top
Graham Slee View Drop Down
Admin Group
Admin Group
Avatar
Telling it as it is

Joined: 11 Jan 2008
Location: South Yorkshire
Status: Offline
Points: 10673
Post Options Post Options   Thanks (1) Thanks(1)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 05 Apr 2021 at 1:31pm
Originally posted by davidismynaim davidismynaim wrote:

But they do show the signal jump up when the distorted noise occurs...

I am trying to answer in a non-argumentive way to illustrate the events which take place. It does take me quite some time to type things up in a way I hope can be understood, and I hope, even if you don't accept what I say, it educates somebody along the way (for the time I am taking over this).

A signal jump indicates a regenerative process. It always has and always will be.

You will always see it as downstream because that is the direction of the gain. It can't gain backwards.

I will not apportion blame on our Iranian fellows for them not understanding the importance of arm-tube grounding, simply because it is the "done thing" in dumb-fi, and it would seem they're all at it.

Regeneration is an oscillative function set into motion by going outside a domain, which within works OK.

If your other advisors have studied mathematical modelling, they would know the above to be true.

The instant of regeneration works at the speed of electron flow, which is a constant you can look up.

Your analogue input(s) are specified at +14dB per phase (+28dB balanced) at roughly 1% THD (frequency unspecified).

That is converted to digital and then, somewhere (I am not fully conversant with your setup), returns to analogue.

Should there be a level where clipping occurs, the A-D will not represent it in any meaningful way.

At 1% THD, there will already be clipping. I have tested sufficient circuits in my career to know that is a fact.

Often, it will be found that distortion subsides at roughly -3dB (half power) to a point where the signal does not clip at all.

14 - 3 is 11.

It is within the realms of possibility that the input, receiving an analogue transient, might give rise to the onset of a clip.

The result being overdrive of the ESL principle such that an arcing event takes place.

As I have said before, arcing was the first form of radio transmission.

It is evident that an inductor connected to a gain stage having a gain of, say, 100, can take that event, resulting in a signal vastly greater than the transient from which the event triggered.

The inductor being the cartridge coils.

So, what I am saying, is do not apportion blame until the entire workings are understood, or, as many have discovered, you could be on a hiding to nothing.

Not simple enough for Google-Bot to understand...
Back to Top
 Post Reply Post Reply
  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 0.094 seconds.