Well, there is no surprise here, Andrew, as you and I ALWAYS seem to be commenting on different aspects!
You seem interested in how calculations and technical aspects might contribute to performance. On the other hand, I care only about what contributes to better sound. I value your input up to the point where you seem to ignore how things actually sound.
Please note that there were many comparisons between caps of equal value, and these exhibited considerable differences in SOUND. At least for these experiments, filter values were identical, yet caps exhibited substantial differences in performance. Believe me, I regret that the experiment is not equally valid at all frequencies for the different values of caps. If I had the funds, I would have chosen all caps of the same value. I think the tests are still valid for all comparisons of midrange and treble.
Can you tell us how the AC across the 'lytic cap would affect the sound? Would the skewed "RC ratios" swamp all the other sonic attributes of every cap? The only reason I chose 1.5uF values (other than the supplied Obbligato) for the recent round of tests was your previous statement that such a value would not allow significant AC across the 'lytic cap. Now I am beginning to question whether that even matters, and that perhaps 2.2uF is, in fact, the best value as suggested by the designer of the amp. Have you performed any listening tests to determine the effect of this AC generation? I'm really trying to comprehend this technical aspect and relate it to the sensual aspect.
Perhaps Dario is onto something with his tests of the interaction of various components.
Has anyone else tested different values of the same cap in this or a similar circuit?
Peace,
Tom E
You seem interested in how calculations and technical aspects might contribute to performance. On the other hand, I care only about what contributes to better sound. I value your input up to the point where you seem to ignore how things actually sound.
Please note that there were many comparisons between caps of equal value, and these exhibited considerable differences in SOUND. At least for these experiments, filter values were identical, yet caps exhibited substantial differences in performance. Believe me, I regret that the experiment is not equally valid at all frequencies for the different values of caps. If I had the funds, I would have chosen all caps of the same value. I think the tests are still valid for all comparisons of midrange and treble.
Can you tell us how the AC across the 'lytic cap would affect the sound? Would the skewed "RC ratios" swamp all the other sonic attributes of every cap? The only reason I chose 1.5uF values (other than the supplied Obbligato) for the recent round of tests was your previous statement that such a value would not allow significant AC across the 'lytic cap. Now I am beginning to question whether that even matters, and that perhaps 2.2uF is, in fact, the best value as suggested by the designer of the amp. Have you performed any listening tests to determine the effect of this AC generation? I'm really trying to comprehend this technical aspect and relate it to the sensual aspect.
Perhaps Dario is onto something with his tests of the interaction of various components.
Has anyone else tested different values of the same cap in this or a similar circuit?
Peace,
Tom E
I use a formula that I found in this Forum which relates the input RC to the NFB RC to specifically limit the AC voltage across the NFB cap. The reason why we might want to do this is to minimise the distortion that the electrolytic will pass to the -IN amplifier terminal. Self and Cordell and many others tell us that the NFB cap will distort the amplifier output if it is made too small, i.e. develops significant AC voltage across it.
The formula that gives some meaning to Self and Cordell is
NFB cap >= sqrt(2) * Input cap * Rin / Rfeedback.
For the MyRefC 1uF would require >=362uF for Nfb cap
1u5F requires >=540uF and 2u2F requires >=725uF.
MyRefC has 220uF installed.
That's the theory behind the RC ratio and why
I have been using input RC~90ms and NFB RC>=140ms for some years now and I have found it to sound good to my ears. It's nice when theory explains why it sounds good. It's even nicer when Theory allows one to predict the component values that might be close to optimum for good sound.
Now back to the input filters.
If you want the 0u68F to give extended bass response then increase Rin to 130k. RC = 88ms.
NFB cap >= 1.414 * 0.68 *100k / 390 >=247uF 220uF is probably close enough to satisfy the no AC voltage criteria for low distortion.
If you want to be sure that the NFB is not creating avoidable distortion then try increasing it to 330uF, or maybe even as high as 470uF
As an aside, I intend trying NFBcap = 220u//220u as an easy to carry out experiment on the effect it has on sound.
The formula that gives some meaning to Self and Cordell is
NFB cap >= sqrt(2) * Input cap * Rin / Rfeedback.
For the MyRefC 1uF would require >=362uF for Nfb cap
1u5F requires >=540uF and 2u2F requires >=725uF.
MyRefC has 220uF installed.
That's the theory behind the RC ratio and why
.
I have been using input RC~90ms and NFB RC>=140ms for some years now and I have found it to sound good to my ears. It's nice when theory explains why it sounds good. It's even nicer when Theory allows one to predict the component values that might be close to optimum for good sound.
Now back to the input filters.
If you want the 0u68F to give extended bass response then increase Rin to 130k. RC = 88ms.
NFB cap >= 1.414 * 0.68 *100k / 390 >=247uF 220uF is probably close enough to satisfy the no AC voltage criteria for low distortion.
If you want to be sure that the NFB is not creating avoidable distortion then try increasing it to 330uF, or maybe even as high as 470uF
As an aside, I intend trying NFBcap = 220u//220u as an easy to carry out experiment on the effect it has on sound.
Last edited:
It certainly is nice when theory and results agree. Thanks for that explanation. I just wish someone could explain how this problem manifests itself in terms of sound.
So this means a larger cap is not the direction I want to go. Perhaps 1.0uF is ideal, as Dario was kind enough to point out was the size originally selected by Mauro. Darn. I'm really getting tired of trying caps, and it has been quite expensive.
Dario, I'm using the resistors from the GB, the ones I suggested: Vishay/Dale metal film. I don't have the document in front of me, but I think they are CPF21K... I know you suggested some other type, but I can't remember which ones or why. Do you really believe these are a big factor? I mean, they're resistors in the power supply. I chose these for tight tolerance and low noise. What influence can a different type of resistor have here?
Peace,
Tom E
So this means a larger cap is not the direction I want to go. Perhaps 1.0uF is ideal, as Dario was kind enough to point out was the size originally selected by Mauro. Darn. I'm really getting tired of trying caps, and it has been quite expensive.
Dario, I'm using the resistors from the GB, the ones I suggested: Vishay/Dale metal film. I don't have the document in front of me, but I think they are CPF21K... I know you suggested some other type, but I can't remember which ones or why. Do you really believe these are a big factor? I mean, they're resistors in the power supply. I chose these for tight tolerance and low noise. What influence can a different type of resistor have here?
Peace,
Tom E
R1, R4 KOA SPR3 3W 1K (660-SPR3CT631R102J)
R1, R4 (Alternate Metal Film) KOA MOS3 3W 1K (660-MOS3CT631R102J)
Yes, Tom, they make a clearly audible difference. Both KOAs are way better than those Vishay.
I don't know why, maybe coating or dimension...
The difference is like the one between the AVX BQs and Wimas.
Much more harmonics, sound stage is bigger and deeper and a little bit more bass with KOAs Metal films (MOS).
KOAs Carbon films (SPR aka Kiwame) are like MOS but more, there is more flesh, harmonics, soundstage and bass.
Kiwame's marketing pretend that those carbon film are less noisy than metal film.
Power supply has a big influence in sound, the DCB1 with Salas shunt PS is an example...
Also in DCB1's PS carbon films are recommended in the BOM according to extensive testing done in the first GB's run.
From the MyRef I've learned that every component seems to influence sound...
Dario, thanks again. I guess I'll try some of those carbon film resistors. Why 3 watt? I thought 2 watt was big enough. I was not planning to change any other parts, but your recommendation and the cost is enough to warrant more trials. I'm tearing them apart again anyway to permanently install the Sonicaps and CL60's.
To all: I did some more careful listening, and the bass is just fine. Despite using two bass traps, my room produces a big drop off in bass between 75 and 45 Hz, and that's what I was hearing earlier. Between 40 and 30 Hz I have a slight peak, and that seems somewhat diminished with the 1.5uF cap. That's why I have a subwoofer (which was turned off for all of these listening experiments) with a small parametric equalizer built into its plate amp.
I am not the least bit interested in going back to a smaller cap, nor am I about to change all kinds of parts to try to meet the criteria of a formula which no one can say with certainty is important to this amp. I don't hear any distortion caused by bigger caps. In fact, this is the clearest, most pure sound I've ever gotten from my system.
Peace,
Tom E
To all: I did some more careful listening, and the bass is just fine. Despite using two bass traps, my room produces a big drop off in bass between 75 and 45 Hz, and that's what I was hearing earlier. Between 40 and 30 Hz I have a slight peak, and that seems somewhat diminished with the 1.5uF cap. That's why I have a subwoofer (which was turned off for all of these listening experiments) with a small parametric equalizer built into its plate amp.
I am not the least bit interested in going back to a smaller cap, nor am I about to change all kinds of parts to try to meet the criteria of a formula which no one can say with certainty is important to this amp. I don't hear any distortion caused by bigger caps. In fact, this is the clearest, most pure sound I've ever gotten from my system.
Peace,
Tom E
In this moment I'm again using the 10nF FKP2.
...
If someone wants to play a bit with C7 values it would be welcome
Dario, thanks for your tip on C7. I tried C7 = 10nF (Rifa PHE840) with:
C13 shorted
C9: 220uF, 4V Black Gate PK
C6, C11: 220uF, 25V Panasonic FC
C1, C2: 330uF, 50V Panasonic FC
C4: 47nF Rifa PHE840 polypropylene
The mids are nearly perfect, the bass has excellent slam and attack without losing texture, but highs are very slightly bright. I'm attributing this to insufficient burn-in on both the Black Gate PK and Panasonic FCs, as well as the magnetic leads on the FCs (which will increase the lead inductance).
I have 22nF, 33nF, 39nF and 47nF polypropylenes, but not all will fit in the location for C7 - I'll try one or more of them in the next build. I'll also try Nippon Chemi-con KY 330uF/50V for C1, C2 - it appears to have better specs (ESR, ripple current, endurance, ...) than the FCs. I'll check if the leads of the KY are magnetic - if they're non-magnetic, it looks like an overall win-win over the FC, and the KY is of 10mm dia., which will fit on the Twisted Pear version 1.2 boards.
Of course, Nichicon Muse KZ 220uF/50V is an even better candidate for C1, C2, but it won't fit on the Twisted Pear version 1.2 boards.
Tom - thanks for your careful experimentation. Just to consolidate the information at one location for future reference, if it's not too much trouble - could you please list the values and types of C1, C2, C4, C6, C7, C9, C11, C10, C12, C30, C34 that you're using with the Sonicap 1.5 uF at C13 (if they differ from the Ultimate BoM)?
Tom, you're welcome
I've used 3W resistors since the first GB. TP BOM includes 3W resistors since builders reported excessive heat.
Those power film resistors have a lifetime related to temperature, like electrolythics, so less heat more lifetime.
I think you wont regret trying SPR3s...
In the meanwhile I've ordered locally the Sonicaps to try.
LinuxGuru, you're welcome.
If you hear harsness probably you can blame both FCs and Rifas.
In this same thread last year I was recommending PHE426 as cheap alternative for C13 but a german forumer tried and found them harsher than a cheap Audyn Cap MKP QS. He was right.
Every single time I've tried FCs I've found them unacceptably harsh, much more than FMs
If you want to cure harshness I suggest to replace FCs in C6 and C11 with KZs or better with Silmics and C7 with a Wima FKP2.
Also C1 and C2 are good candidates for better caps but, as you already pointed out, the space don't allow the use of 16mm caps, as all good same value audio caps are: BGs, Silmics, Cerafine, KZs...
In the 12mm limit FMs are quite good but I've had reports from Suburra about Cerafines ang BGs std that are better and both removes the last bit of harsness (he forced the 16mm caps using C17, C18 holes)
I'll wait for your feedback about different value C7s
Link to separate thread on a modification to the RevC, which I'm calling the RevD:
http://www.diyaudio.com/forums/chip-amps/175937-myref-revd-successor-revc.html#post2342433
Please post comments, suggestions, etc. related to that modification on that thread, not here - TIA.
http://www.diyaudio.com/forums/chip-amps/175937-myref-revd-successor-revc.html#post2342433
Please post comments, suggestions, etc. related to that modification on that thread, not here - TIA.
In the build I'm currently testing (which is more than a year old!), all caps are the value designated on the schematic.
C1, C2, C6, C11 are all Panny FM's. I suggest you dump all the FC's. Dario likes Silmic's; I prefer the FM's. In my next build, I'll try the Silmics supplied in the GB.
C4, C7, and C30 are Wima MKP2
C9 is Blackgate Std 50V
C10, C12, and C34 are ceramic Kemet Golden Max
Dario, the 3W resistors will fit and cost no more, so that's what I'll try. I'm glad to learn you could source the Sonicaps locally at a decent price.
Peace,
Tom E
C1, C2, C6, C11 are all Panny FM's. I suggest you dump all the FC's. Dario likes Silmic's; I prefer the FM's. In my next build, I'll try the Silmics supplied in the GB.
C4, C7, and C30 are Wima MKP2
C9 is Blackgate Std 50V
C10, C12, and C34 are ceramic Kemet Golden Max
Dario, the 3W resistors will fit and cost no more, so that's what I'll try. I'm glad to learn you could source the Sonicaps locally at a decent price.
Peace,
Tom E
Time to resurrect this old thread with some news.
I think I'm finally done with these amps I started building more than a year ago. They are very similar to the most recent group buy with "The Ultimate BOM".
At Dario's suggestion, I recently replaced Dale metal film R1 and R4 with KOA SPR3 carbon film. He is correct about them improving the sound. They reduced brightness in the treble, took away a sort of metallic edge, while still allowing very clear and detailed highs with beautiful depth and sweetness. If your amps suffer at all from overly lit highs, I suggest this simple and inexpensive change.
The Sonicaps have thoroughly broken in, and they sound even better than they did initially. These caps provide very realistic tonality and spaciousness, with no spotlighting of any frequency. They really are a superb bargain, giving up nothing to and even surpassing some far more expensive caps.
Peace,
Tom E
I think I'm finally done with these amps I started building more than a year ago. They are very similar to the most recent group buy with "The Ultimate BOM".
At Dario's suggestion, I recently replaced Dale metal film R1 and R4 with KOA SPR3 carbon film. He is correct about them improving the sound. They reduced brightness in the treble, took away a sort of metallic edge, while still allowing very clear and detailed highs with beautiful depth and sweetness. If your amps suffer at all from overly lit highs, I suggest this simple and inexpensive change.
The Sonicaps have thoroughly broken in, and they sound even better than they did initially. These caps provide very realistic tonality and spaciousness, with no spotlighting of any frequency. They really are a superb bargain, giving up nothing to and even surpassing some far more expensive caps.
Peace,
Tom E
I've my sonicaps still in the bag... I've had no time to do any test recently....
If you want to cure harshness I suggest to replace FCs in C6 and C11 with KZs or better with Silmics and C7 with a Wima FKP2.
...
Also C1 and C2 are good candidates for better caps but, as you already pointed out, the space don't allow the use of 16mm caps, as all good same value audio caps are: BGs, Silmics, Cerafine, KZs...
...
I'll wait for your feedback about different value C7s
For the Rev C version 1.2 PCB:
I replaced C6,C11 with 100uF/25V Silmic IIs; C1,C2 with 100uF/50V MUSE KZs. Both are definite improvements over FCs at these locations even with just a couple of hours break-in. Less edgy, smoother, more airy.
C7 is still experimental: I tried an AVX Skycap ceramic 47nF/50V/5%/C0G instead of a 100nF/63V Wima MKS2 here. First audible impression: no regression, and probably some improvement in neutralizing some of the edginess and sibilance in the upper mids/highs. I'll let it stay and bed in for a while, but it's good enough to repeat it again in the next build.
Just a hunch: a low-Q, moderate ESR, linear (C0G) wideband ceramic bypass at C7 is probably better than a high-Q, low-ESR film cap at taming some of the metallicity/resonance from lead/trace inductances near the LM318. The leads of C7 must be non-magnetic, for sure.
As expected...
I've tried C0Gs as well and they're quite good and better than some film caps but a Wima FKP2 is FAR better.
If it fits, metallized paper is worth trying out for C7 - it has some of the same properties as C0G ceramic - moderate ESR, relatively low-Q, high linearity. Also a fair capacitance-to-volume ratio, probably comparable to or better than MKP/FKP.
The issue with FKP2 for C7 is that the highest value that I can find that will fit is 10nF/63V. It's adequate for now, but I'm giving a 7.5mm pitch option and a bit more width for C7 in the next PCB re-spin.
Hi Jon,
not yet, I hope next week...
Happy New Year to you too and same for the rest of our brave builders!
Hi Jon,
not yet, I hope next week...
Happy New Year to you too and same for the rest of our brave builders!
Thanks! I'm hoping to order some of these in the next few days...
Updated BOM?
My new years resolution is to try my hand at building an amplifier, and Uriah suggested the ultimate BOM My Ref "Rev C". After skimming the thread, it looks like there have been a significant number of improvements on the choice of parts since ClaveFremen last posted a BOM (post #1064). Would it be possible for someone to update the ultimate BOM with the new ultimate parts? It would be greatly appreciated by those of us who are still at the beginning of their DIY adventure.
Linuxguru - I sent you a pm regarding purchasing boards. I am hoping that you still have some Rev "C" boards available, or that there is some other way that I can obtain them.
My new years resolution is to try my hand at building an amplifier, and Uriah suggested the ultimate BOM My Ref "Rev C". After skimming the thread, it looks like there have been a significant number of improvements on the choice of parts since ClaveFremen last posted a BOM (post #1064). Would it be possible for someone to update the ultimate BOM with the new ultimate parts? It would be greatly appreciated by those of us who are still at the beginning of their DIY adventure.
Linuxguru - I sent you a pm regarding purchasing boards. I am hoping that you still have some Rev "C" boards available, or that there is some other way that I can obtain them.