No, there is neither Zobel (Boucherot cell) compensation, neither Thiele network on the output.
It had been a point from the original designer, Mauro Penasa. As You might have learned recently, from our discussion in HERE, the output stage is a high output impedance current pump. As such, the actual output voltage will always be a direct function of loudspeaker passive (impedance) /active (EMF) contribution.
The speaker is an integral part of feedback, in this amp.
For this reason Mauro did not want to alter even minimally the amp - speaker link characteristics.
For this reason he had omitted these extra compensation/ filter components at this sensitive point.
Though. I have since long contemplated the need for a Boucherot cell.. When the output is open (relay inactivated) the feedback loop sees a strong alteration, and in standby mode the core amp section could even oscillate. No, generally it does not happen, but in marginal cases it could. A Boucherot cell is invented for this reason...
To further blur the lines, in the even more original Tim de Paravicini MF A370 scheme - there IS a Zobel / Boucherot included on the output.. much probably for the reason above, when no speaker is connected.. (or there is excess stray inductance in the speaker setup)
Ciao, George
It had been a point from the original designer, Mauro Penasa. As You might have learned recently, from our discussion in HERE, the output stage is a high output impedance current pump. As such, the actual output voltage will always be a direct function of loudspeaker passive (impedance) /active (EMF) contribution.
The speaker is an integral part of feedback, in this amp.
For this reason Mauro did not want to alter even minimally the amp - speaker link characteristics.
For this reason he had omitted these extra compensation/ filter components at this sensitive point.
Though. I have since long contemplated the need for a Boucherot cell.. When the output is open (relay inactivated) the feedback loop sees a strong alteration, and in standby mode the core amp section could even oscillate. No, generally it does not happen, but in marginal cases it could. A Boucherot cell is invented for this reason...
To further blur the lines, in the even more original Tim de Paravicini MF A370 scheme - there IS a Zobel / Boucherot included on the output.. much probably for the reason above, when no speaker is connected.. (or there is excess stray inductance in the speaker setup)
Ciao, George
Last edited:
Best would be a duplicate relay action: lifting the load, but in the same time connecting a Zobel network in place of it..
Actually the relay IS doing already part of it: we just should implement the Zobel parts to the presently 'empty' relay contact.
I think I will play with it..
Again: it does not change anything with respect the present operating conditions.. It only changes the inner balances during standby mode. Or during startup, the only moment with real importance..
Actually the relay IS doing already part of it: we just should implement the Zobel parts to the presently 'empty' relay contact.
I think I will play with it..
Again: it does not change anything with respect the present operating conditions.. It only changes the inner balances during standby mode. Or during startup, the only moment with real importance..
Last edited:
Now, to be honest, we /(as in someone, some day..) should look up Mauro's original thinking about these things. There is a lot of intervention of his, either here but even more in the italian forum..
So who am I to pronounce the 'truth' in place of him!?
So the usual pinch (bucket!) of salt to be taken into account.
Furthermore, Mauro had been a rigorous engineer, first of all.. So very unlikely would he be thinking in audiophile terms like 'detrimental contribution' of some parts..
So I think we can only assume that he preferred to have a 'cleaner' situation, while he had been looking at it, both in terms of feedback loop, and in terms of output impedance.. (we know that he had been concentratig on these)
And 'cleaner' in sense of 'less complex', rather than 'less dirty'...
So who am I to pronounce the 'truth' in place of him!?
So the usual pinch (bucket!) of salt to be taken into account.
Furthermore, Mauro had been a rigorous engineer, first of all.. So very unlikely would he be thinking in audiophile terms like 'detrimental contribution' of some parts..
So I think we can only assume that he preferred to have a 'cleaner' situation, while he had been looking at it, both in terms of feedback loop, and in terms of output impedance.. (we know that he had been concentratig on these)
And 'cleaner' in sense of 'less complex', rather than 'less dirty'...
I am not sure if anyone is able to comment, but having built a MyRef_C (Russ White Red Board 1.2 2006) I am finding the sound really lacking much HF content. I have used all the correct values for the on-board components except for the PS caps. These are suggested to be 10,000uf. I had some nice Panasonic 18,000uf caps left over from another project and so I used these. Could these have too much capacitance?
By the way Dario has been away from posting on here lately. I hope that he is well.
By the way Dario has been away from posting on here lately. I hope that he is well.
In general, I can only say that there is no topological reason for less resolution. Proof is how it does sound when assembled with parts selected for resolution..
Then, in my experience, each unit will have it's own sound, depending on the parts selection.
Now I am not sure what was the base of your BOM. The 'contemporary', original suggested set for the Russ board, or You did follow rather one of the BOM versions of Dario..
What I know is that the BOM of Dario seem to be directed for a 'balanced' sound, rather the warm side, as in contrast with sheer resolution. I appreciate that kind of sound, even if I had deviated from it since a while.
As you start to raise the target, and head for the more specific, higher level BOMs, all the changes are pointing to the higher resolution side..
Then, in my experience, each unit will have it's own sound, depending on the parts selection.
Now I am not sure what was the base of your BOM. The 'contemporary', original suggested set for the Russ board, or You did follow rather one of the BOM versions of Dario..
What I know is that the BOM of Dario seem to be directed for a 'balanced' sound, rather the warm side, as in contrast with sheer resolution. I appreciate that kind of sound, even if I had deviated from it since a while.
As you start to raise the target, and head for the more specific, higher level BOMs, all the changes are pointing to the higher resolution side..
Last edited:
Like, for more concrete examples: the Wima FKP capacitors are not properly a feat of high definition.
When you swap for the suggested polystyrene, or Armtrans types, that will bring a lot in that direction.
The feedback capacitor, C9, should be controlled too.
The diodes, (MUR series..) the resistances in the feedback loop (dale in place of, for example z-foils) are again, more on the 'natural' side.
If you have the standard Caddock in R3 position, then that is another aspect..
When you swap for the suggested polystyrene, or Armtrans types, that will bring a lot in that direction.
The feedback capacitor, C9, should be controlled too.
The diodes, (MUR series..) the resistances in the feedback loop (dale in place of, for example z-foils) are again, more on the 'natural' side.
If you have the standard Caddock in R3 position, then that is another aspect..
Puffin: As Joseph K has already stated, Dario’s suggested BOM, even using his suggested "audiophile" components upgrade, will lead to a "balanced & warm" sounding amp. The use of naked Z-foils and Amtran’s AMCHs will bring more delicacy to the upper end but having not heard a Russ White build I can’t comment on the differences between builds.
For informations sake, you could go look at this thread:
“Fremen Edition My_Ref Amplifier – Completed Projects and Photos” to compare various builds and impressions on SQ.
Cheers
For informations sake, you could go look at this thread:
“Fremen Edition My_Ref Amplifier – Completed Projects and Photos” to compare various builds and impressions on SQ.
Cheers
Thanks for the replies guys. I used good quality components to populate the board. I was not aware that the presentation of this version of a GC was designed to be smoother than the 2 caps and half a dozen resistors circuit a la 47 Labs. I may try reducing the PS capacitance. I also understand that C9 is a candidate for experimentation to tailor the sound.
Smaller PS caps is likely to change the sound, but I would be surprised if it has much effect on HF content. Bigger caps often have more ESR which can effect the ability of the cap's ability to respond to large current demands. But HF content usually doesn't require much in the way of PS current. Still, it is worth a try. The "audiophile" choice of Mundorf caps is there because they have lower ESR and you can hear it in improved bass definition. But I don't hear any difference in the HF content.
While George and Pete are correct in characterizing the smooth, warm sound of the FE amp, I wonder if you are describing something bigger that might be indicating a problem? Although smooth, I have never felt I was missing HF content.
But if you are talking character choices, then yes, components do make a difference, and it seems more than than other amps because of the current pump architecture, at least IMHO. And if you find yourself playing with C9, I suggest that C13 is the most influential on sound character and I think that C13 and C9 interact. I have tried combinations where the HF content became too much and everything became etched and a little harsh.
Jac
While George and Pete are correct in characterizing the smooth, warm sound of the FE amp, I wonder if you are describing something bigger that might be indicating a problem? Although smooth, I have never felt I was missing HF content.
But if you are talking character choices, then yes, components do make a difference, and it seems more than than other amps because of the current pump architecture, at least IMHO. And if you find yourself playing with C9, I suggest that C13 is the most influential on sound character and I think that C13 and C9 interact. I have tried combinations where the HF content became too much and everything became etched and a little harsh.
Jac
Thanks for your reply Jac. I will see what more exotic caps can do at C9 and C13. Rob.
There is something wrong, the amp has no such thing as lack of hf content...
The PS caps value should not be a problem but could have imparted a different sound signature than the inteded one but unless they're awful I would exclude them as the source fo the problem.
I'm fine, simply at work night & day, 11PM here and I'm patching several customers Exchange servers for a zero day vunerability...
No, it's not, more later
In any case, yes, I would could confirm that a well executed Myref amp is leaning to more the natural balance then the 'classic SS' sound.
There had been a lot of work & attention spent on 'polishing' the rough edges..
Mmmh... Guys where did you took this from?
'warm' sound or "'polishing' the rough edges" are never been goals for the voicing of the amp...
Up to version 1.2 the component selection was made upon best perceived sound from several components and here the 'polishing' thing may have some merit but mostly was what sounded better to my (and others, as BCMBOB) ears.
From version 1.5 a radical different approach has been taken with the goal of the maximum possible neutrality and maximum detail, the amp has slight hint of a warm signature due to the RN55 resistors that has been selected for other desirable traits like details and noise floor and Wima FKP2 caps.
Must be taken in account that the real deal is the BOM with audiophile parts, which is pretty neutral.
Puffin can you share the exact BOM you used so that I can help you identify the possible culprits?
Anyway if the HF lack is serious I would suspect an error on the LM318 compensation network.
If not C13 and C9 have a big impact on sound.
Last edited:
It’s what we hear Dario. This is meant as a compliment. I love the sound of my Evo A build with the audiophile components. The warmth envelopes you while the upper end is beautifully articulated and clean. I would never describe it as "bright" however, not with a LM318 in place. My other build, with ADA4627 in place, shifts the music towards the upper end.
My speakers, my DAC, my listening environment, will give me a different experience.
With respect,
Pete
Glad to hear that you are well. CV-19 has made the world a very different place.
This is the BOM I used :-
Number Desc 1 Desc 2 Desc 3 Source Part Number per count total
D1-D3 1N4001 1A - 50V Diodes On Semi Digikey 1N4001RLOSCT-ND 0.224 3 0.672
Q2-Q3 BC546 NPN Bipolar Transistor On Semi Digikey BC546BOS-ND 0.164 2 0.328
Q1 BC639 NPN Bipolar Transistor Fairchild Mouser 512-BC639 0.18 1 0.18
ZD1-ZD4 BZX85-XX 12V 1.3W Zener Diode Fairchild Mouser 512-BZX85C12 0.08 4 0.32
C7, C23 100nF MKT (box) 50V Capacitor Panasonic Digikey P4593-ND 0.108 2 0.216
C10, C25 100pF MKT (box) 50V Capacitor Panasonic Digikey P4570-ND 0.072 2 0.144
C12, C27 220pF MKT (box) 50V Capacitor Panasonic Digikey P4574-ND 0.072 2 0.144
C4, C5, C19, C22 100nF MKT (box) 100V Capacitor Panasonic Digikey P4656-ND 0.159 4 0.636
C13, C29 1uF MKP (box) 50V Capacitor Panasonic Digikey E1105-ND 0.741 2 1.482
C9, C14, C22 220uF Radial 50V EL Capacitor Sanyo AX Bdent 63MV220AX 0.4 3 1.2
C1-2, C17-18 220uF Radial 50V lowESR EL Capacitor Sanyo FZ Bdent 50MV220FZ 1.04 4 4.16
C3, C8, C20, C28 10000uF 50V EL Caps Sanyo DAC Bdent 50PL10000DAC 4.29 4 17.16
C15 22uF Radial 25V EL Capacitor Sanyo SAX Bdent 25MV22SAX 0.18 1 0.18
C6, C11, C16, C21, C26 100uF Radial 25V EL Capacitor Sanyo AX Bdent 25MV100AX 0.18 5 0.9
U1-2 LM319N DIL Op Amp National Semi Digikey LM318N-ND 1.017 2 2.034
IC1-2 LM3886 Amp National Semi Digikey LM3886TF-ND 4.16 2 8.32
D1, D5 8A 400V Bridge Rect. Diodes GBT804 Mouser 621-GBJ804 1.52 2 3.04
R5, R8, R28, R31 22K0 .25W .1% IRC .1% RC55 Mouser 66-RC55-D-22.1K 0.96 4 3.84
R6, R9, R29, R32 47K0 .25W .1% IRC .1% RC55 Mouser 66-RC55-D-47.5K 0.96 4 3.84
R2, R25 33K .25W 1% KOA Spear Mouser 660-MF1/4CL3322F 0.12 2 0.24
R7, R30 12K0 .25W 1% KOA Spear Mouser 660-MF1/4CL1202F 0.12 2 0.24
R10, R33 390 .25W 1% KOA Spear Mouser 660-MF1/4CL3920F 0.12 2 0.24
R11, R35 1 .25W 1% KOA Spear Mouser 660-MF1/4CL1R00F 0.17 2 0.34
R12, R34 3K32 .25W 1% KOA Spear Mouser 660-MF1/4CL3321F 0.12 2 0.24
R13, R36 100K .25W 1% KOA Spear Mouser 660-MF1/4CL1003F 0.12 2 0.24
R15-18 75K .25W 1% KOA Spear Mouser 660-MF1/4CL7502F 0.12 4 0.48
R19 10K .25W 1% KOA Spear Mouser 660-MF1/4CL1002F 0.12 1 0.12
R20 47K .25W 1% KOA Spear Mouser 660-MF1/4CL4702F 0.12 1 0.12
R21 220K .25W 1% KOA Spear Mouser 660-MF1/4CL2213F 0.12 1 0.12
R22 8K2 .25W 1% KOA Spear Mouser 660-MF1/4CL8251F 0.12 1 0.12
R23 6K8 .5W 5% KOA Spear Mouser 660-MF1/2CL6811F 0.14 1 0.14
R1, R4, R24, R27 1K 1W 5% Vishay-Dale PRO1 Mouser 594-5073NW1K000J 0.16 4 0.64
R14 470 1W 5% Vishay-Dale PRO1 Mouser 594-5073NW470R0J 0.16 1 0.16
R3, R26 0.47 5W WW Huntington Elec Digikey ALSR5F.50-ND 1.35 2 2.7
RELAY1 24VDC 2sc 8A-250V 24v 8A Omeron Mouser 653-G2RL-24B-DC24 4.1 1 4.1
PL1-8 Faston Connectors (PCB Mount) Digikey A24742-ND 0.055 8 0.44
However, not all the components were of the same manufacture as the BOM, but the values were the same.