I have not yet got around to replacing the films with the lytics as you suggest only because the hiss was not really objectionable as previously mentioned. The gain is around 10db not what I wrongly reported earlier. This BA3 balanced pre with Salas BIIB boards is really special in all ways except that I find a bit too much brightness which gives the effect of violins sounding edgy. You mention that "Lytics can round the edges". Not sure what you mean. Do you feel it could help with this edginess? Or perhaps there is something else you suggest. Thanks.
nash
It could be third harmonic, it could be the hiss, it could be EMI interference, it could be little instability. Is there a non fuzzy straight line when scoping the rails with the probe AC coupled to the scope set at a sensitive vertical scale? If not, does any evident rail confusion on the screen go away with the reg outputs converted in two wire mode? If all looks well anyway, trying Nichicon FG 100-470uF can assure the rail hiss will drop vs small value film cap and this may make the sound smoother.
I adjusted P3 to increase second harmonic and while the character changed a bit as expected it did not really help with the brightness.
Unfortunately I do not possess a scope yet. I know you have repeatedly asked us to get one and I will.
I think I want to try the Nichicon FG as you suggest first and note the difference in sound quality. Mouser has 35V (my reg output is 24V) ones in 100uf and 220uf with 10mm diameter. The 330 and 470uf are 12.5mm diameter and I am not sure whether they can squeeze in, you know better
I will buy whatever you suggest?
Thanks again.
nash
Just following up and posting my observations:
Tried 2 wire mode and there was no difference to the hiss.
Tried the 220uf Nichicon FG electrolytic caps and again no difference to the hiss level. A slight bit rounded at the top but mid range clarity and naturalness deteriorated so removed the electros and put the Wimas back in.
Regarding the glare/brightness I had mentioned, I replaced the AC power wire from the power inlet to the transformers from a silver plated copper twisted pair to a solid copper twisted pair. This reduced the glare by over 50%. I also made a new shunt attenuator replacing the KOA Speer MF resistors with Takman REY 1/4W resistors and TX2575 series and this has done the trick.
Please see my pics in http://www.diyaudio.com/forums/pass...-your-diy-pass-amplifier-368.html#post4808806
Thank you.
nash
Not having even moderately positive result in hiss with the FGs which have a huge difference in value to the Wimas means the hiss is not originating in the Vref. Having changed to two wire mode and no difference in hiss again further removes the idea that there was some interference or oscillation on the rails. At least until you will have an oscilloscope to make sure. Most probably it originates in the audio circuitry for some reason. Too much current bias through some sensitive part maybe. Or its just the gain revealing some hiss from the source if the gain structure of the whole system is prone to. I mean I don't know, just guessing. Your metal work is very sturdy. The sectioning strengthens it further. Built like a tank. Congratulations.
Thank you Salas.
Frankly, the hiss does not really concern me since it cant be heard more than 9in from the speakers. Perhaps the issue is with too much gain- about 10.5db from the preamp feeding a power amp with 27db gain.
But it sure sounds good!
the glare/brightness that changes points to a PSU ringing problem that is altering the amplifier performance.
Andrew, I do not have a scope so could you describe what a PSU ringing sounds like. I hope I am not misleading you by my use of the words brightness/glare.
Let me try and describe what I heard before. A high note of the violin when peaking would sound somewhat edgy, higher piano notes would also sound a bit sharper than natural and so with instruments like a clarinet when the high note is ending sounded a bit too piercing.
Thanks, nash
exaggeration of HF leads to an apparent brighter sound.
It's the harmonics being increased relative to their fundamentals that we hear as not quite right.
This performance shortcoming can be directly due to overshoot of fast signals when the stability margins are too low.
It can also be brought about when the amplifiers PSRR (Power Supply Rejection Ratio) allows perturbations of the supply rail to momentarily affect the signal. It's at HF that this effect is worst. We hear extra HF detail that is not present in thereal signal.
Those supply rail variations can be ringing of the supply when we present it with a fast changing current demand.
That is a likely symptom of transformer+rectifier ringing.
It's the harmonics being increased relative to their fundamentals that we hear as not quite right.
This performance shortcoming can be directly due to overshoot of fast signals when the stability margins are too low.
It can also be brought about when the amplifiers PSRR (Power Supply Rejection Ratio) allows perturbations of the supply rail to momentarily affect the signal. It's at HF that this effect is worst. We hear extra HF detail that is not present in thereal signal.
Those supply rail variations can be ringing of the supply when we present it with a fast changing current demand.
That is a likely symptom of transformer+rectifier ringing.
Thanks Andrew for furthering my understanding of the effect of harmonics on brightness.
nash