Get rid of noise, help please.

[catalin gramada]

Hello
I have a noise problem at input of one stage in my project. It is a buzz which come louder once the treble pot goes up. It is not a huge noise, it is minimal like a floor noise but it is supposed to be a hiss , and not a buzzy noise. I isolated everything. The stages in front was disconected and tubes pulled off. The tone control is completely closed by screen cage grounded to chassis. The connection between tone control and grid was done also with shielded wire, without any difference. The filter cap is the last from the rail and everything in front of this stage was disconected .
Any ideea what to check please ? Thanks.

[catalin gramada]

I just replaced the tone control with a 500k pot connected just in the whiper and the ground. Now is complete quiet almost 90% of the track. To the end of the track it starts to have the same manifestation as the tone control. I just measured the resistance of the pot in position where the noise comes evident and found till 370k into the grid as grid leak the tube is still quiet, no more. Put a litle bit more and start to get noise into input. What the heck ? ...

[Helmholtz]

Reminds me of a previous thread. Is there a feedback path connecting to the first 12AT7 cathode?

[catalin gramada]

Hi. Thanks for reply
Yes it is. Or better said will be. I didn't used in my test cause I preferred to keep everything simple once I found where the noise comes from. I don't know how to diagnose the input noise, but just start to think can be a dirty signal ground and I wonder why? The bus is conected at the power ground in a point bolted to chassis and all the filter caps along it defining the local points. I.m pretty sure there are not conflicts over there. Everything in front of this 12at7 stage was disconected to can do a simple check.
Still I wonder, touching the signal ground bus with finger I get little bit scratch noise. If I touch in the same time the chassis with other hand it didn't scratch. I wonder where this coupling comes from ?

[Helmholtz]

Hi. Thanks for reply
Yes it is. Or better said will be. I didn't used in my test cause I preferred to keep everything simple once I found where the noise comes from. I don't know how to diagnose the input noise, but just start to think can be a dirty signal ground and I wonder why? The bus is conected at the power ground in a point bolted to chassis and all the filter caps along it defining the local points. I.m pretty sure there are not conflicts over there. Everything in front of this 12at7 stage was disconected to can do a simple check.
Still I wonder, touching the signal ground bus with finger I get little bit scratch noise. If I touch in the same time the chassis with other hand it didn't scratch. I wonder where this coupling comes from ?

Ok, so no global NFB yet.
Can you scope the buzz at the PI outpts?
What happens if you reduce the grid leak R to 1M (= max. value for 12AT7)? And if you use a 10K grid stopper at the socket?

touching the signal ground bus with finger

Which exact point do you mean?

If I touch in the same time the chassis with other hand it didn't scratch.

Your body resistance provides a grounding path.