So I built this KSA-50 (Pinkmouse) years ago. Sounded better than any amp I had built before... but always had this slight, but annoying hum....drove me a bit batty...
The power supply rails were around +/- 35vdc, and I had a 56kuf - 1ohm - 56kuf on each rail of the power supply.
A while back I changed the resistors to an inductors (2.5mh, added to each rail) and the amp got quite a bit quieter, but somehow lost some "magic"....don't know if I am imagining it or not, but have not used it much since I did the change...can anyone point out a reason for this?
Getting ready to just change it back....
The power supply rails were around +/- 35vdc, and I had a 56kuf - 1ohm - 56kuf on each rail of the power supply.
A while back I changed the resistors to an inductors (2.5mh, added to each rail) and the amp got quite a bit quieter, but somehow lost some "magic"....don't know if I am imagining it or not, but have not used it much since I did the change...can anyone point out a reason for this?
Getting ready to just change it back....
What was the change in ripple on the rails? What is the DCR of the 2.5mH coils (and what voltage drop across it)?
Do you have other options than 56mF for dividing up the cap banks?
Do you have other options than 56mF for dividing up the cap banks?
The ripple dropped considerably on the Duncan PSU Designer Calc, and was confirmed when I put the coils in - hum dropped quite a bit....but I never measured the remaining ripple and voltage drop accross the coil. I can measure when I have a minute this weekend....
I have more 63v caps, and I can do a CRCRC with added 10kuf and resistors between the 56kuf caps....
I have more 63v caps, and I can do a CRCRC with added 10kuf and resistors between the 56kuf caps....
The 2.5mH would give 1.9 ohms at 120Hz, not including the resistance of the coil, as Andersonix is pointing out. Perhaps you'd get some reduction, but maybe more at somewhat higher frequencies. One would wonder what steps you'd already taken to ensure noise at those frequencies is managed, if this is having a significant effect. Also, have you tried some RF suppression along with the coil?
It can be a mixed result swaping inductors for resistors. .sounds good in theory. Dosent always equate to an improvement in overall sound.
Hmmm, I really think the 1 ohm resistor instead of the 2.5mh coil sounded "livelier"....
But a real Krell power supply (at least from what I remember from my KSA-250 I had years ago) didn't have a CRC or CLC...it was just two monster 47kuf/100v caps connected together, per rail....CC
But a real Krell power supply (at least from what I remember from my KSA-250 I had years ago) didn't have a CRC or CLC...it was just two monster 47kuf/100v caps connected together, per rail....CC
Audible hum on a home-built with 56kuF caps and a decent circuit is probably layout, signal OR ground, not PS filtering.
OK, put it back to CRC and I am happier again...
But one more question...I have 56kuf - 1ohm - 56kuf on each rail feeding both channels - would I be better off having each channel on their own 56uf cap per rail? The Transformer has dual secondaries, so I can do it...
The original KSA50 power supply had a "Dual Mono" set up this way, while later KSA-80 on had a single, doubled up power supply feeding both channels...which is preferred?
EDIT - PRR - this is a heavily biased Class A amp - I can hear the hum increase as I increased the bias....so I believe it to be power supply related...
But one more question...I have 56kuf - 1ohm - 56kuf on each rail feeding both channels - would I be better off having each channel on their own 56uf cap per rail? The Transformer has dual secondaries, so I can do it...
The original KSA50 power supply had a "Dual Mono" set up this way, while later KSA-80 on had a single, doubled up power supply feeding both channels...which is preferred?
EDIT - PRR - this is a heavily biased Class A amp - I can hear the hum increase as I increased the bias....so I believe it to be power supply related...
Attachments
Give both a try. Each will have its own sound signature.what you prefere is up to you.thats part of the journey in diy.
Obviously. More current more ripple. But that does not show that poor filtering is the reason. More likely cap current corrupting the "ground" path.