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 8788899091 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 (0) Thanks(0)   Quote BAK Quote  Post ReplyReply Direct Link To This Post Posted: 18 Nov 2019 at 3:17pm
When Graham said: "Dinsdale noted the distortion producing stereo loop when he added another channel to the 1961 Tobey & Dinsdale mono amplifier to convert it to stereo." This is the diagram.


Do you have the power supply grounds for left and right connected together?
Or are the speaker grounds for both channels?

 To eliminate the stereo loop, the left and right channel grounds should only be directly connected together at the chassis signal input... and the rest of the power supply and power amp grounds are separate with each channel's PSU floating (0v not connected to the chassis) and each channel's power returning to its own PSU filter 0v. Having a separate PSU for each channel enables having separate "power grounds".


Edited by BAK - 18 Nov 2019 at 3:39pm
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 (0) Thanks(0)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 20 Nov 2019 at 7:28am
"Although solid-state amplifiers have been around for some 40 years, it would be a great mistake to assume that everything possible is known about them." Doug Self, 2009

I've been brainstorming this stereo-loop distortion with Bruce (by Skype) for the last two evenings, trying to define it mathematically, and we got as far as stating it's an inverse function.

Inserting any impedance at points Z2 and/or Z3 might cause amplifier designers to laugh out loud, but so far it would seem to be addressing the "decaying" sound quality problem - decaying to the point where newly added components reach their optimum or "burnt-in" state.

Using a two transformer supply it looks like Z1 increased, and say if using two truly independent supplies (which would have to be hefty 72 volt batteries!), Z1 would approach infinity.

As Z1 approaches zero Z2 and Z3 are required to preserve the sound quality, as per Dinsdale's discovery, and even with the "correct" grounding Doug Self says circumvents the phenomena - it might in a high NFB design, but even then, this is outside the properties of an individual amplifier.

The stereo-loop phenomena only happens on connecting the input commons together, which is inevitable where there is more than one channel.

Bruce's suggestion of a ladder ground (or even a ground plane) might collapse the loop, and at least one fine sounding early Japanese amp ran a single ground down the middle like a tree trunk; but in this design L and R amps are either end of the case.

Provided the assumption is right (there is always time for it to trick me again) then we might have discovered (or re-discovered) another reason why progress took the dual rail power supply option. Equilibrium seems to cancel this effect.

So, to my mind, I think we might have discovered the solution, which might be written down as

Z2 (Z3) >> 1/Z1 (where Z1 is a real number and not infinity or 0)

The exact ratio depicted by >> is as yet unknown, but >>> sounds worse than >> and so does >.

And also, we still have no idea of the value of Z1.









That none should be able to buy or sell without a smartphone and the knowledge in how to use apps
Back to Top
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 (0) Thanks(0)   Quote BAK Quote  Post ReplyReply Direct Link To This Post Posted: 20 Nov 2019 at 3:37pm
Z1 being in the AC return current path "connecting" via the transformer windings between channels...
We can only approximate the value of Z1 through measuring the resulting difference between:
 1 channel used in mono (without the 2nd channel opperating)
and 1 channel with a signal applied while the 2nd channel is quiet and/or has a different signal.


Edited by BAK - 20 Nov 2019 at 3:42pm
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
BackinBlack View Drop Down
Senior Member
Senior Member
Avatar

Joined: 05 Feb 2012
Location: Hinton, N'hants
Status: Online
Points: 2020
Post Options Post Options   Thanks (0) Thanks(0)   Quote BackinBlack Quote  Post ReplyReply Direct Link To This Post Posted: 20 Nov 2019 at 4:21pm
Just wondering, surely this "stereo loop distortion" must also, to a lesser degree, be evident in all stereo pair pre-amplifiers and even phono stage amplifiers? 

Ian
Just listen, if it sounds good to you, enjoy it.
Back to Top
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 (0) Thanks(0)   Quote BAK Quote  Post ReplyReply Direct Link To This Post Posted: 21 Nov 2019 at 2:01am
Originally posted by BackinBlack BackinBlack wrote:

Just wondering, surely this "stereo loop distortion" must also, to a lesser degree, be evident in all stereo pair pre-amplifiers and even phono stage amplifiers? 

Ian
To a much, much lesser degree... possibly in lesser degree than the noise floor.

 This "stereo loop distortion" has been heard in power amplifiers.

 High-power current, if not controlled properly, can generate noise and distortion enough to be noticed... 
amounting to be above the noise floor, if only a little.
 
The loop that is of concern is the high power current returning to the PSU (and the mains) from the varying output of a power amplifier.

 When high-power current(s) are circulating through a circuit, they can modulate any lower power currents that are sharing the same path(s).
 The high-power current(s) can modulate another high-power current (as in a stereo) that shares the same path.


Edited by BAK - 21 Nov 2019 at 2:02am
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 (0) Thanks(0)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 21 Nov 2019 at 7:03am
Originally posted by BackinBlack BackinBlack wrote:

Just wondering, surely this "stereo loop distortion" must also, to a lesser degree, be evident in all stereo pair pre-amplifiers and even phono stage amplifiers?

Ian


Yes, to a (much) lesser degree, but knowing that somebody with acute hearing has tried "dual-mono'ing" two of my phono preamps and testifies it sounds more "open", investigating this might be a logical future step.

I'll just restate what Doug Self wrote: "Although solid-state amplifiers have been around for some 40 years, it would be a great mistake to assume that everything possible is known about them."

But here, in the power amplifier we have current pulses which might reach 2.2 amps prior to clipping, along with printed circuit traces and wires, which although are suitably rated, have resistance.

At the point where the power supply common branches to the power take-off point of left and right amplifiers there is an impedance that is non-zero, but is believed to be so.

It is (in a single rail supply) the negative end of a capacitor, or paralleled group of capacitors, and they have equivalent series resistance (ESR). These are the reservoirs of current (charge), and when supplying current it is via this ESR.

Ohms law can be rearranged to state that as current increases across a fixed value of resistance, the voltage across it increases.

The ESR I have for the 6,000uF of reservoir capacitance calculates as 0.01 ohms at 200Hz, and a current pulse of 2.2 amps results in a 22mV voltage difference across the capacitor.

This indicates the "solid 0V" afforded by star "earthing" is not solid.

The 22mV difference applies around the loop, and clearly something has to give. It isn't DC so PCB track and wire inductance comes into play even though it's low (200Hz for the sake of this argument) frequency.

Even if it isn't exactly 22mV - in normal listening it would be much less - the illustration shows some "fight" or "tussle" within the loop.

The sensitivity is around 800mV for full power, and at 800mV this 22mV exists. It is a percentage of nearly 3%.

At 10% signal: 80mV, 2.2mV exists, same percentage.

As far as the "wiring" goes either the dog wags the tail or the tail wags the dog, or both, but it is a ground current which is asymmetrical so becomes part of the signal as even harmonic distortion.

I only picked 200Hz because it's a bass frequency twice that of the power supply switching rate and it is bass loss that inflicts itself on this amplifier when it shares the same power supply. 200Hz is "out of the way".

The bass loss isn't really bass loss as when measured the frequency response is flat, but the even harmonics place ear-misleading 400Hz, 800Hz, 1600Hz etc, on the signal due to the stereo loop - audio analysers are not equipped to measure this.

The other reason for choosing 200Hz is that with a dual secondary power supply, there should be sufficient "dwell" on each diode of the bridge rectifiers which is in the charge path at the time, to have little impedance.

The dual secondaries are, one would think, independent of one another, but it is looking like their starts and ends have little impedance between them (high resistance at DC - low impedance at AC). I am a transformer ignoramus!

When using two transformers the coupling is at the mains switch, and I guess there is much more impedance between left and right power take-off zero volts, than the dual secondaries?

Listening tests reveal little difference between single supply and dual secondary supply, but a marked difference (as in better) using two transformers.

Although I tried to convince myself this was an interference thing due to toroid’s being more tightly coupled, I might have been barking up the wrong tree.

What tends to prove this "phenomena" to my mind, is that by adding impedance at Z2 and Z3, the bass improves to be extremely similar to what it is using two transformers.

The actual value which seems to sound best is 1 ohm. And if the "error voltage" is 22mV, the current in this resistor is 22mA, and the error which would be introduced to distort common (non stereo difference) signals is countered by the resistor's position which is within the negative feedback loop of the amplifier.

The above tends to fit in with what Dinsdale was saying. Whether or not Doug Self's interpretation of the fault considered that Dinsdale could have grounded the amplifier correctly, I am not sure. I think Dinsdale had grounded it properly but failed to see the capacitive ESR. In all my extensive and exhaustive research I have failed to find any description of it except for this one mention by Dinsdale in this single article in Wireless World.

So, getting back to the preamp question, if it exists it must have some effect, but to what extent and will that extent be audible?

In a preamp the output current is often in the low 10s or 1s of milliamps, and often in the region of 5mA. The error voltage is 50uV using the same values as above, but for only 1,000uF it increases to 300uV. If the input is only 300uV as in a MC cartridge output, then at 200Hz it is 120uV, so that stereo loop distortion would swamp the signal. But our hearing tells us that is not the case. The answer therefore seems to be both the power supply regulation, and the PCB track or wiring impedance being much lower versus the flow of current.

Regulation of the active type (a voltage regulator) is able to detect uV differences and change its output to make the voltage correct at all times (up to around 10MHz) such that an ESR induced difference of 300uV is corrected back to where it should be, and therefore operates a form of negative feedback on the ESR.

This might also explain John Linsley Hood's observation that a regulated power amplifier power supply "improves bass".
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps
Back to Top
BackinBlack View Drop Down
Senior Member
Senior Member
Avatar

Joined: 05 Feb 2012
Location: Hinton, N'hants
Status: Online
Points: 2020
Post Options Post Options   Thanks (0) Thanks(0)   Quote BackinBlack Quote  Post ReplyReply Direct Link To This Post Posted: 21 Nov 2019 at 11:43am
Thank you Bruce and to Graham for such a detailed explanation. It seemed logical that the problem would exist in any stereo amplifier pair, what I couldn't deduce was the impact it might have at lower power levels.
As always our ears tell us the real story in that it hasn't been a noticeable problem with well designed phono and pre-amps, at least to my ears. 

Ian
Just listen, if it sounds good to you, enjoy it.
Back to Top
 Post Reply Post Reply Page  <1 8788899091 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 0.109 seconds.