Noise problem

[catalin gramada]

Hell9.
I have this project on my table. The circuit is absolutely completly quiet...but: For a certain position of P2 (around 2/3 track) it raise a bit of noise. It is just a tiny bit but audible in a quiet environment. Then the noise disappear rotating the whiper to the end of the track. The noise is related to a certain position of P2. Why please? Thanks.

[pdf64]

Have you checked for oscillation?

[catalin gramada]

Yes is absolute stable.

[Greg Robinson]

Source impedance will be highest somewhere around the middle of the resistance range, and that corresponds to around 2/3rds of the way up for a log pot.

[catalin gramada]

Thanks. Could be the amount of ac ripple in heater dc reference a problem in this conditions please ? I forgot to measure it as I found the amp absolutely quiet. But heaters are elevated 80vdc around from a supply point provided and returned directly from/to the first main filter cap. 220k/39k,22u...if matter. Meant it is possible to do the h-k coupling most sensitive (as s/n ratio) for those certain point potentiometer position please ? There is a heater to cathode ac impedance limit for a certain tube at a certain current from this point...maybe...?

please to consider is not a severe issue. The noise is barely audible in a very quiet environment in a 100db speaker...but is here. I may pass the amp as it is but think it deserve an explanation from a building point.

[pdf64]

How about if a lower value pot is used, eg 47k?

[Helmholtz]

Not clear what kind of noise you have.

[catalin gramada]

I.ll try to see if I may detect it. My probes are not good quality ones...

[catalin gramada]

First, found dc grid voltage instability over half turn of the pot which show pdf64 state to reduce its grid leak value to have a sense for me. But still have to disconect first the coupling cap to be absolute sure is a grid leak issue and not a coupling leak...

[Helmholtz]

My probes are not good quality ones...

Shouldn't make much difference at audio frequencies.

All a cheap 1X probe can do is add something like 100p to ground.

[catalin gramada]

And no. It is stable with 220k on the grid without to be conected to previous stage, meant the coupling have a problem...I have to fix that first...

[catalin gramada]

The 1uF output dc offset is all over from 0.4 to -0.4 Vdc swing over a 2M load. I.m over because I cannot test the cap. Maybe the cap is good but due the poor voltage regulation it swing during charging to restore its offset and in this condition may be usable if I will reduce the load to something smaller as pdf64 suggested

[catalin gramada]

The cap is way to large for the next stage load. I did it large to preserve as small impedance for a potential external small impedance input. But maybe I should make two separate output circuits derived from same cathode with two coupling caps for each inputs....then charging transients will be keeped to a decent amounts

[Helmholtz]

Not clear what you mean.
A cap shouldn't leak any DC.
I would avoid an ecap here, though.

[catalin gramada]

It don't but it is referenced to a dc potential to the ground, right ? If this potential to the ground is change it the cap will be restore the reference to the new potential. During this two points a charging current will apear causing a transient in dc offset. At least it looks so...
then the signal will swing on a x axis who swing up and down with a value related with those transient. This it shows to me and did not looks like a damaged cap. A leaky cap will keep constantly a dc to its output not trying to restore in 0v I think
Is just a supposition, I will try to see if anything related by bumping the dc voltage in the cathode btw to confirm.