That's good news, that even with the offset trouble the change has been a success. I'm curious in what ways you found the sound better?
They are real. Hopefully it is a result from installation. I solder my own. But been watching this issue before tackling the project. It is good to hear that the sonics improve. My stock Rev A is sounding very nice as is. Well kind of stock.
I have been very reluctant to check the design. But if opamps and a few other changes improve sound I need to dig into it.
The largest improvement, in my opinion, is for those who use it as an integrated amplifier. Original My_ref FE is not so good that way, a bit "muddy" and "coarse": you really miss a preamplifier.
But using ADA4627 things change, it seems as you added a small pre inside the chassis! More refined sound, more details, more dynamics, more bass, better instruments definition and separation. Only timbre is not up to using a good tubes preamplifier.
As a pure power amplifier there is a good improvement, but it's less substantial because original FE is already very good in that utilisation.
I'd say more dynamics, a little more open, great bass (it was very good with LM318 anyway, but now it's more controlled), better coherence of voices. Maybe a bit faster too. Still lacks a little bit to reach hiend amplifiers, but the distance is shorter. A real bargain for the price, even if compared to used gear!
I tested it with very good high sensitivity monitors (around 92db), to be honest. I couldn't test it with lower db speakers, maybe it would perform worse.
N.B. Tonight, peak offset 26mV and stable well below 10mV, reached in half an hour. Quite near to normal FE levels. I can't understand why, but offset is improving even if i do nothing... It seems like opamps need burn in to work properly...
Last edited:
Not to break the string of interesting opamp discussion going on here, but I just want to emphasize one aspect of construction of this amp.
After a couple years, I finally got around to replacing the Caddock MK132 at R10 with a naked Vishay Z-foil. The change of sound has thoroughly convinced me that this resistor is the only choice for obtaining the best sound from this amp. I have, over the years and in several revisions, tried Vishay/Dale and other metal film, Shinkoh tantatlum, and Riken, but the TX2575 is by far the best. All sense of stridency is gone. The Caddock provided great detail, but there were times when high strings made we wince at the squeakiness and shrillness. The Z-foil Vishay is smoother but allows similar detail, with an opening of images and depth of soundstage that no other resistor provided. I have read that too many of the naked Vishay's can introduce sterility and a coldness, but I have not yet reached that point. Every place in this amp and other components that I've tried it has provided only improvement. The sound I get now is open, smooth, and detailed without any sense of harshness. I have also read comments by "experts" that a single part, especially a resistor, cannot possibly make a significant contribution to the voice of a component. Those experts are wrong.
Peace,
Tom E
After a couple years, I finally got around to replacing the Caddock MK132 at R10 with a naked Vishay Z-foil. The change of sound has thoroughly convinced me that this resistor is the only choice for obtaining the best sound from this amp. I have, over the years and in several revisions, tried Vishay/Dale and other metal film, Shinkoh tantatlum, and Riken, but the TX2575 is by far the best. All sense of stridency is gone. The Caddock provided great detail, but there were times when high strings made we wince at the squeakiness and shrillness. The Z-foil Vishay is smoother but allows similar detail, with an opening of images and depth of soundstage that no other resistor provided. I have read that too many of the naked Vishay's can introduce sterility and a coldness, but I have not yet reached that point. Every place in this amp and other components that I've tried it has provided only improvement. The sound I get now is open, smooth, and detailed without any sense of harshness. I have also read comments by "experts" that a single part, especially a resistor, cannot possibly make a significant contribution to the voice of a component. Those experts are wrong.
Peace,
Tom E
Tom: in a word: Yes.
And as far as I understood, also Dan could say something similar?
......
Ok, would like to ask, pardon me if I don't remember: did You do already the same for R7 (12kohm). That is maybe an even more sensitive place.
And what surprised me recently (or rather not, Dario kept telling) is the input filter series resistor . And the influence of the filter caps..
Ciao, george
And as far as I understood, also Dan could say something similar?
......
Ok, would like to ask, pardon me if I don't remember: did You do already the same for R7 (12kohm). That is maybe an even more sensitive place.
And what surprised me recently (or rather not, Dario kept telling) is the input filter series resistor . And the influence of the filter caps..
Ciao, george
Last edited:
Only timbre is not up to using a good tubes preamplifier....
N.B. Tonight, peak offset 26mV and stable well below 10mV, reached in half an hour. Quite near to normal FE levels. I can't understand why, but offset is improving even if i do nothing... It seems like opamps need burn in to work properly...
Luca, very intersting note about the pre. We had experienced something very similar, with our friends C.J. pre. It puts a luxurious shine to the sound. The Evo 'integrated' with the Noble Black pot directly on the input -- seems like less spectacular, and more 'closed'.
But with a hint at being more correct, like that grey aspect rigorous secretary in administration..
And it's addictive, now I can not listen to my DCB1, at home.
Ciao, George
Ps. Nice to hear about the offset. Hm, it makes me turn back to the surface leak version.. that is something that could change like this. Don't forget that you washed the board, too. Without immediate effect, but still..
Ciao, George
Last edited:
While in general you would be right in this case there is nothing to be sorted on the circuit... the My_Ref works correctly as is from 2004/2005 when Mauro Penasa introduced it in this forum.
Others changed the opamp to experiment and no one else had problems with offset.
The most likely cause is flux residue or, less likely, faulty opamps.
While I prefer too the FE with a preamp I would not call it, used without pre, muddy or coarse...never...
I suspect that the ADA4627 better PSRR is masking the silly smoothing caps configuration you have, those long leads should 'slow' your PS, hence your 'muddy' impression, IMHO.
BTW Audyn True Copper are no longer in BOM for a reason... guess why?
N.B. Tonight, peak offset 26mV and stable well below 10mV, reached in half an hour. Quite near to normal FE levels. I can't understand why, but offset is improving even if i do nothing... It seems like opamps need burn in to work properly...
It's not burn in in this case, it's flux residue that dries up.
Do another thorough cleaning session with IPA and cotton.
Thanks for the description. I also found it improves most aspects of the sound, a fundamental step forward in resolution rather than helping on just one or two areas.
Tom: in a word: Yes.
And as far as I understood, also Dan could say something similar?
Absolutely - I was expecting perhaps more detail, but what I heard was a nicer tonality and a more full bodied sound, as if a source of dissonant distortion had been removed.
I'm not talking about slowness, but not perfect definition of sounds and less dynamics than using a pre.
Maybe i can better define it as sound not easily flowing outside the speakers, like if the amp was "lazy" in pushing out sound. Using a pre this doesn't happen, even using LM318.
Also other users noticed that when used as integrated, FE's SQ is lower than when used as a pure power amp. That's quite normal to me, as that can be said about many power amps.
Don't misunderstand me: i'm not saying My_ref FE is not a very good amp, just that it's (much) better as a power amp than as an integrated - at least using LM318s. Using ADA4627 the gap is much smaller.
Anyway, offboard smoothing caps is a configuration used in many power amps that have a completely opposite behaviour, like my Circlotron OTL monoblocks, that are indeed very fast and "flowing"! So i wouldn't call it "silly".
How much could a < 10cm silver and copper wire slow down a 1A (or higher) current? Maybe there aren't copper traces connecting those caps onboard, too? How long are they? 2, 3, 4 cm instead of 10? And why should a <10cm wire slow down the PS only when using the amp as integrated?
I used the same configuration on other amps (chipamps too) and never had this "muddy"effect when switching from onboard to offboard caps position.
But in life you never know for sure, so i will mount onboard Mundorf AGs on next My_ref i'm building, then i will tell the difference.
BTW Audyn True Copper are no longer in BOM for a reason... guess why?
Again, i used TCs as coupling caps in many hifi components and none of them showed any "slowness" or "muddiness", indeed the exact opposite!
I always found them better than Mundorf Supremes and many of my audiophile friends agree.
So i really don't understand why they have been removed from the BOM - except that i understand that different tastes are normal. Nor i can guess why they should have this bad effect only when using My_ref as integrated... and why they should cease this effect when replacing LM318s with ADA4627s...
To be clearer: if i replace a component in an amp and it sounds better, i'm going to think that the SQ bottleneck was in that component, not in others components that after replacement (and/or in other configurations) sound perfectly good. That's so simple. I understand it's a bit simplistic, but really am i wrong?
Obviously it's not burn in, i was kidding.
I'll do what you suggest, but i'm sure i made a very thorough cleaning last time, it took more than an hour, using alcohol, deflux and compressed air.
Last edited:
Anyway, offboard smoothing caps is a configuration used in many power amps that have a completely opposite behaviour, like my Circlotron OTL monoblocks, that are indeed very fast and "flowing"! So i wouldn't call it "silly".
(...)
I used the same configuration on other amps (chipamps too) and never had this "muddy"effect when switching from onboard to offboard caps position.
You're right, that configuration is used widely on class A amps but when you have dedicated space on boards I still find silly to use outboard caps, strictly IMHO.
Maybe I'm wrong but I think that having those caps on board and so near to their consumer (the LM3886) is a plus, at least for RF rejection.
If you will do it I'll read with interest your findings about it.
Probably it wasn't clear, I was not referring to muddiness but to hardness that at the moment I've associated with your 'coarse'.
In my previous system the original blue True Copper were pretty neutral and open, very fast but they exacerbated hardness from other parts.
It was probably a problem in my system but later Audyn changed the design so that they were way bigger (more RF pickup and impossible to accomodate on board) and lost their neutral character for a warmer/mellower one.
At that point they were no longer better than Supremes to my ears or an option.
I'll do what you suggest, but i'm sure i made a very thorough cleaning last time, it took more than an hour, using alcohol, deflux and compressed air.
According your description it seems so but a second cleaning session should not hurt either.
This is the approach that made My_ref FE so special!
It's very difficult to express my impressions in italian yet, doing it in english is an "impossible mission" for me... Choosing words is a trap...
I meant "coarse" as unrefined, not fully defined at 360°...
I never listened to old TCs. But i find current TCs are a little bit too strong on highs, that can be slightly disturbing in some systems; but this is not the case. I agree they are a bit on the mellow side too, but just a bit when compared to other caps.
Anyway, their flow, definition and spatiality is so good that i nonetheless prefer them to Mundorf Supremes (that to my ears tend to have still a slightly warm sound, but less richness and definition).
Personal tastes, of course!
Fitting them onboard is not so straight, but still possible using some little tricks. Not the same for TC MAXs, unfortunately.
According your description it seems so but a second cleaning session should not hurt either.
I'll do and let you know about the results.
Last edited:
I've tried VSR which is better (more detail, more neutral) than RN55 but not as neutral as naked Vishays.
The S102 I've tried were used so I'm not so sure about them but they were not on par with naked Vishays or even VSR.
Never tried RCK02 but they're pretty cheap for a metal foil resistor.
I never tried them in My_ref, but i had exceptional results using Rhopoint Econistors: they are near z-foil level for detail and neutrality, but a little bit less "hard". I found them even better than z-foils in some applications (not in all cases, though), specifically as i/v. They are nearly 20% cheaper than Texas z-foils too.
The problem is they are very difficult to find, Newark sells them at reasonable prices but not in all values.
Rhopoint sells them directly and often has some lower tolerance rating resistors at discounted prices (tolerance rating is very high for these resistors anyway!), you can contact them by email and ask to their commercial division. But i warn you: considering shipping costs, only buying a certain quantity is cost effective, for few pieces you won't find a significant saving in respect to z-foils.
MyRef sounds good with almost any resistor at R10 and R12. For the parts I've tried, my impressions follow. PRP's are cheap and detailed but a bit metallic as opposed to musical. Vishay/Dale are cheaper, with a flat response but not much sparkle or "air". The only resistor I didn't like at all was Rikens, which made a juicy midrange but muffled the highs to the point of sounding dull. Shinkoh in this amp and other apps, and non-magnetic Audio Note tantalum in other apps brought the music to life but were a bit fizzy and over-emphasized on top. Takman MF did not make a memorable impression (perhaps they were a bit bright?), but maybe that's good, and they're relatively cheap.
If you don't want to spend the money for Z-foil, and that's perfectly understandable, I suggest Vishay/Dale for the most "even" sound. You don't give up much fidelity and you save a lot. None of the "boutique" parts priced between V/D and Z-foil were worth the extra money. The Z-foil is worth the extra money if you have it to spend, but it's not essential to get decent sound.
Peace,
Tom E
I like using the Dale RN-55E series. Very smooth. Not sure they are still manufactured. I have a few, need to locate the values for the feedback positions. I do not have the correct values in metal foil.
Looked over George's amp and he had somehow attached through hole resistors to the pads for surface mount. It was impressive.
Looked over George's amp and he had somehow attached through hole resistors to the pads for surface mount. It was impressive.