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 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)


1970s Design Indulgence

 Post Reply Post Reply Page  <1 88899091>
Author
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: 8881
Post Options Post Options   Thanks (0) Thanks(0)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 25 Nov 2019 at 9:32am
One of the sins of amplifier design is to willingly add ground impedance. It simply isn't the done thing.

That Dinsdale did so is criticised by Self, who says separating power and signal grounds is better.

Those grounds are not separate as such, as they come from the same place: they join at the power supply 0V.

By taking the signal ground by a separate wire or track, then no ground current can flow where it joins to the other channel's ground, whether that be at the input terminals, or in a kick-can amplifier, in the source.

Except, 0V depends on zero being zero, which when AC current is flowing in the reservoir capacitor ESR, 0V isn't stable.

Therefore there is "ghosting" which produces a conflict where the input grounds join.

Dinsdale's ground resistance is purposefully placed between power and VAS 0V such that it is in the ground where the circuit is subject to global negative feedback. It reduces the effective NFB, which will lead to more distortion, but reduces the distortion of one channel asymmetrically modulating the other.

Dinsdale described that distortion as being "unpleasant... particularly when one channel has a transient such as a cymbal clash".

I have to ask, what is more important: convention or enjoyment. Here, convention seems to imply that what you're hearing is correct, even if it doesn't appear so to you.

The problem I found in employing Dinsdale's ground resistor was it didn't really improve anything, unless it was a very small value of 1 ohm. The improvement was to the bass end but the highs were still arguably strident, and stole attention away from the bass improvement.

It was then noticed that the "RF end" HF frequencies had been upset, and that any increase up to Dinsdale's 10 ohms made it worse still. The RF end wanted the "solid" ground to enable it to behave. The solution? Bypass the resistor with a small capacitor.

It was then possible to play with values in the region of 1 to 10 ohms. I have single resistor values of 1, 3.3, 6.8, and 10 ohms. The one I hadn't tried was 6.8 ohms, so I'm trying it now.
Not simple enough for Google-Bot to understand...
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: 8881
Post Options Post Options   Thanks (0) Thanks(0)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 25 Nov 2019 at 10:35am
Another thing to be aware of in accepting the existence of reservoir ESR, it can be seen to have an influence on gain margin, in that it reduces it considerably. Here again we see impedance at HF we don't want, and all we can do is to bypass the reservoir using the most effective capacitor available, which in my opinion is the axial "Spinguard" type from AVX in X7R dielectric (value: 100n). Note: this will have no effect whatsoever at audio frequencies, where it cannot "short" the problematic ESR.
Not simple enough for Google-Bot to understand...
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: 8881
Post Options Post Options   Thanks (0) Thanks(0)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 25 Nov 2019 at 11:45am
The proof of inter-channel modulation is shown in the maximum output distortion sweep - something not seen on the same scale using separate transformers.
Not simple enough for Google-Bot to understand...
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: 8881
Post Options Post Options   Thanks (0) Thanks(0)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 28 Nov 2019 at 1:12pm
To blame is human, but what if all the development in power amplifier design over the past 50 years turned out to be because the origin of blame was wrong?

In talking about the following I am hoping I don't end up looking the fool - the amplifier still needs time for the dreaded burn-in - it would not be the first time I thought the job was done.

It would seem Dinsdale was a ridiculed lone voice who made an important discovery which everybody else scorned.

Not being able to get the answer, I have to conjecture: Dinsdale was involved in multichannel amplifiers taking their inputs from sensors, with the amplifiers powered from a battery. Their outputs instructing motors to do things needed for the stability of one of the UK's proudest inventions: the Harrier "Jump Jet".

A battery being a unipolar supply, there is little option but to use one of its terminals as the "common" for everything.

One side of all sensors must be commoned to that common. All amplifier outputs are connected to that common too.

All the outputs are doing different things, but if the common is influenced by them, the all-important signals - as he had shown regarding the stereo amplifier - would be subject to asymmetrical interference: even-order harmonic distortion.

Self dismisses this by saying Dinsdale had power-output current flowing in the commons, but Dinsdale had said, quote: "(with individual decoupling)", which makes me think he would have taken the load to power supply common - the individual decoupling capacitors; 2,000uF each.

Indeed, if we view part 3 (https://www.americanradiohistory.com/Archive-Wireless-World/60s/Wireless-World-1965-02.pdf) and fig. 22 on page 75, it can be see that the correct grounding convention is used.

Now, all this might not matter using dual-rail supplies, because the reservoir midpoint might not fluctuate very much due to dielectric absorption, as large electrolytics have a tendency to retain charge for longer than the full cycle of most audio frequencies, and with one stacked above the other might cancel this phenomena out. Here I am only making a wild guess.

But if this were only to apply to unipolar supply stereo amplifiers, and if Dinsdale's answer is ignored, then what to blame for the distortion and poor bass?

All the possible alternative answers are neatly bound in John Linsley Hood's book: VALVE AND TRANSISTOR AUDIO AMPLIFIERS. Not once is stereo loop distortion mentioned. Try finding about it on the Internet...


Edited by Graham Slee - 28 Nov 2019 at 1:13pm
Not simple enough for Google-Bot to understand...
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: 8881
Post Options Post Options   Thanks (0) Thanks(0)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 02 Dec 2019 at 9:29am
Before simulators became cheap, and operating systems, math coprocessors and memory were all large, fast and powerful enough to do the thousands of calculations required; the back-room bod like me had to design power amps by educated guesswork.

Most now coalesce around a singular solution. It's easier that way. In-fact why bother with designers? Amps are a commodity whose only differences are the branding and cabinetry.

To be different might be considered stupid, but I'll do it anyway. Maybe it'll result in a different sound?

Model in every gremlin and you'll know where that carefully planned phase and gain margin went! It's a fight between reality and what you want to achieve, with reality tending to win the argument every time.

Reality is the enemy of power amplifier high fidelity, so you have to work with reality in the hopes of finding the peace which results in the sound you want.

But did it work?
Not simple enough for Google-Bot to understand...
Back to Top
Aussie Mick View Drop Down
Senior Member
Senior Member


Joined: 21 Feb 2014
Location: Bendigo Aust.
Status: Offline
Points: 662
Post Options Post Options   Thanks (0) Thanks(0)   Quote Aussie Mick Quote  Post ReplyReply Direct Link To This Post Posted: 03 Dec 2019 at 5:38am
Graham,
   Hopefully it will indeed result in a different sound! Otherwise, from an artistic viewpoint, why bother? I settled on the Proprius pair, precisely because they were so different that my dissatisfaction with mainstream amplification was finally overcome. 
Being kind to the mainstream, I suppose I could say there is no better or worse, just different. Without the CHOICE, I would never have found my happy place.
Mick.
Rega RP8/Ortofon 2M Black/Rega Apheta 2 - Reflex M/Elevator EXP - C.E.C. TL5 CD Transport - MacBook Pro/Roon - Majestic DAC - Solo ULDE (Focal Elear) - Proprius - PMC Twenty5.22
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: 8881
Post Options Post Options   Thanks (0) Thanks(0)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 03 Dec 2019 at 1:14pm
At the 120 hour mark an "acid-test" - a FLAC of Mike and the Mechanics "Living Years" album - was played.

As Leo once commented "walk round a hi-fi show with that album under your arm and you'll be thrown out".

If anything in the system is wrong "Living Years" will expose it by its strident brightness.

Playing it again now at 140 hours and it's doing OK.

Stereo loop distortion has been proved to be real, and there exist a number of solutions:

1. the Dinsdale resistor
2. the tree trunk symmetrical ground
3. separate transformers

The Teleton used "2".

Expensive Cambridge amps seem to use "3".

Plus, slow reservoir capacitors in a dual rail supply help.

Also, regulated power supplies can help.

Monoblocks simply do not suffer it.


Edited by Graham Slee - 03 Dec 2019 at 1:15pm
Not simple enough for Google-Bot to understand...
Back to Top
 Post Reply Post Reply Page  <1 88899091>
  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 0.086 seconds.