Hi Max!

Your mods will work fine some notes:

1) I use caps like this all the time. In fact I have used those exact caps. I would not go much higher than 470uf, it would be hard to find any that fit anyway.

2) You are correct, and the value you state will work just fine, it will only lower the corner frequency slightly, not enough to change things much. If you wanted you could reduce C12 to 180pf or so.

3) Yes I have done this too, but not with the mills resistors yet. It is a good mod. It is worth noting that the new current pump amp I am designing which also needs an output resistor like this to set gain will be using Ohmite TA or Mills type non-inductive resistors in this spot.

Cheers!
Russ

Edited by user Wednesday, November 1, 2006 9:40:00 AM(UTC)  | Reason: Not specified

R3 resistor,
I also intend to use some different parts on my second rev c board, when I get around to it. I thought a Mills 0.47ohm non inductive resistor for R3 would be an improvement on the standard inductive ceramic resistor that is supplied, especially since it is in the signal path.
I will soon be building mauro's Evolution amp, where he supplied standard 0.47ohm 5W ceramics in a similar position. I suggested using 12W Mills resistor substititions, but Mauro prefers that I use the 5W ceramics supplied. He says 5W resistors for both the base and full versions for the Evolution amp is ample and 7W resistors are unnecessary. I do remember Mauro writing that the 7W resistors can get hot in the rev c, and the Evolution I would think should be comparable, although it's output wattage is a little higher.

I don't quite know what to make of all this. I thought a "better" replacement component is generally desirable. Maybe the greater excitation of the electrons in a hotter (lower wattage) component results in better sound??

ps.

Hi Max,

You are right - we all differ in our sensitivity to sound, and it is a matter of "try it and see".
Just like your sensitivity to wires (no doubt there are others), I seem to be far too sensitive to changes to my system setup than I would have expected at my age.
I think Mauro's aim in his designs is to achieve the best possible sound using standard quality components. he probably feels the performance of the Evolution amp will not benefit from the use of "better" components. mauro acknowledges though that the performence of the rev c amp is affected by some component choices, notably among them,C9 and C21.
It is worth remembering that a system is only as good as the weakest link, and improving the resolution of one piece of equipment may not necessarily bring an improvement in the sound of the system.

ps.

Hi Max,
I used Rubycon ZA at C9 - seems to be ok. Panasonic FM would be comparable I would think. Pity the Black Gate N series do not fit.
For C21 I used Vishay Roederstein MKP1837,which are often recommended in such a position - also seems to do the job. On my next rev c boards I will try Black gate NX HiQ 0.1uF, 50V as C21. Might also use them to bypass the 220uF rail caps. Martin Colloms recommends them above all others as bypasses. They are inexpensive. I used them to bypass BG FK caps that I used as coupling caps on a MC preamp I was experimenting on, to great effect. They cleaned up the highs beautifully. ( I'm going to rebuild that preamp and use GB Super E- Caps instead).

For the input cable I used Cardas 2x21.5 AWG Twinaxial at $3.50/ft from Parts Connexion. I have no inclination to change it. It needs a very hot (400C)soldering iron to solder the litz wires though.

Regards,
ps.

Hi,
I found time to swap some components after the festivities...
Thanks to my very solid heatsink strategy I could not take off the stock original power module and swap it for my already stuffed new modded module...so I desoldered some caps and did the mod on the original amp.

I swapped C9 for Ruby ZL 220uF/50V and the PS rails caps (whose exact number I forgot) for Ruby ZL 470uF/50V.
R2(33K) was substitued for 10K Riken that I had on hand. I did not address R12 (3.3K) this time.
That way my two pair of amps will differ on C9 (Pana FM v/s Ruby ZL), input cap (Wima MKS10 v/s Epcos) and R12 (stock 3.3K v/s Riken 3.9K). That should be an interesting comparison.
I couldn't swap those 100uF EL caps because Ruby caps had less distance between legs.

Sonics: I don't know...it must be a genetic trait or something but I allways found Pana FC sound good on bass but lean and dry on midrange and extended on highs but have that "razor sharp" quality that I dislike. All this is gone with the Rubycons: they are well balanced on the whole freq. spectrum; deliver rock solid deep bass, with punchy and "thick" midrange; they doesn't lack highs wich are delivered with superb transparency and detail.
The bad thing is that I will end with a bunch of Panasonic caps on my project box! :-D
A pitty that you can't offer them as stock or optional caps. It's very hard to get them.

Exited by these findings I also modded one of my Kookaburra's wich sets volume on my big system. I only get the four 220uf, PS caps, not the chips decoupling caps, but the swap already showed benefit in tems of increased coherence between freqs. ; rounder, punchier mids.
At first I hesitate to do it but I realized that the PGA2311 chip has an internal opamp and that the buffer opamps also must get the best quality power that I can give.

I also considered "snubbing" the rectifier diodes. I will post further findings on the proper Kookaburra page.
Changing some 0.1uF decoupling caps for BG-NX is also tempting as I've done this elsewere (Ruby+BG) with good results.

Bye,
M