general tips for isolating noisy component

[methodofcontrol]

i've got a sound city 50 plus here. it works fine, but it idles noisy. like a steady white noise. doesn't matter if guitar is plugged in or not. EQ effects noise somewhat, same w/ volume. ultimately, the guitar signal is louder than the noise.
only thing that makes it go away is if i pull V3. that tube tests well, and i tried a known
good replacement as well w/ no change.
so i'm wondering.... should i be checking for DC leaks on the caps on V3? how can i isolate a noisy component w/out shotgun replacing a section?
chopstick poking doesn't reveal anything obvious.
what's the SOP for my next move? thanks!

[Chuck H]

Looking at the schem I see that nightmare phase error circuit is feeding V3. Try grounding the V3 input grids (pins2 and 7) one at a time. If either stops the noise then you have isolated the problem to the circuit feeding V3 and eliminated V3 as a target. I'm thinking it's a bad ground in a circuit feeding one of the V3 grids.

[methodofcontrol]

well... i tried grounding the grids one at a time w/ no effect on the noise.
does that mean that the noise is coming from directly AFTER V3?

[Chuck H]

well... i tried grounding the grids one at a time w/ no effect on the noise.
does that mean that the noise is coming from directly AFTER V3?

Yup. Looking at the schem I would've guessed a bad ground feeding V3 but if grounding the input grids doesn't affect the noise then it must be after V3. But with the volume and tone stack affecting it that implies the noise is before V3. Hmmmm..? Of course if the controls don't affect it much then it could just be the normal noise level your hearing under the problem. If the plate resistors are carbon comp type I might try replacing those with metal film. Including the PI.

[methodofcontrol]

what i hear as "normal" amounts of buzz and hum is what happens when the PI and power tubes are still in play but V3 is removed. granted, my friend who owns the amp swears this quieter version w/ V3 removed is how the amp always sounded before a few weeks ago. i never heard the amp before then, so all i know is the noisier version, (as we hear it w/ V3 in play).
the controls seem to be active in this amp. if all EQ controls are set to zero, the amp does not transfer any guitar signal to the speaker.
i'll try grounding the V3 grids again just in case i had a bad contact or something. is there any other point or component i can lift or ground to further help isolate the issue?
thanks again!

[jrfrond]

Don't know what issues you might be having, but Sound City amps weren't dubbed "Sounds Shitty" for nothing! They are normally noisy as hell. Never liked those amps.

[methodofcontrol]

ok...i went back retried grounding the grids on V3.
what do you know...? grounding pin 7 took out all of the noise. i also grounded the grids one at a time on V1 and V2 with no change.
don't think i have a bad connection, i would assume a bad component. i'm going to measure the resistance and capacitance of relevant components nearby and i'll resolder them as well to be sure.
anyone have any more specific ideas as to where i might be looking?
thanks again!

[Chuck H]

I would check the plate resistor for that stage (if it's a carbon comp resistor it may need replacing) but more so... All the odd tone stack circuits mix at this junction. Three different gain stages are mixed at this grid from different points of a phase correction circuit. All are frequency specific and all are mixed via 470k resistors. The phase circuit itself hosts about six .01uf caps in series with 1M resistor loads in between. I would probably start by replacing all the 470k mixing resistors AND the V3B plate resistor with metal film types. I would also look into the quality of the .01uf caps in the phase network circuit and replace them if I found them to be of poor quality.

JM2C

[methodofcontrol]

i checked R30, R25, R26 and R27 and all measured within spec. same goes for C7, C17 and C20. also, using an alligator clip, i redundantly grounded several ground points to check for bad joints.
it sounds like a bad resistor or cap, or even a bad connection in the socket itself... but i haven't been able to really target exactly where it is.
you think those 470k's are my best bet? what about the 820 ohm? caps are all mustards- would be a shame to replace them w/out necessity.

[Chuck H]

A component can test good and still be noisy. Carbon composition resistors hiss like crazy and the higher the resistance the worse it is. So all those 470k mixing resistors, if they're carbon comp, could be part of the problem. Replacing them with metal film resistors could quiet things down. Also, plate resistors are very prone to thermal hiss because of the high voltage across them. Again, carbon comp resistors are the worst for noise in this application and can become very hissy and even crackly sounding with age. Replace with metal film. Not sure about this one but that phase twisty tone network of .01 caps has me suspicious. I wouldn't be surprised to find out those .01 caps are big ceramic discs. Noisy? not sure but I might replace them and see.

I think this is probably just a noisy circuit that has gotten worse with age. Upgrading to metal film resistors in critical areas and doing away with cheap disc caps may be all it needs. And that's just on the circuits attached to V3 pin 7.

How serious is the hiss? If the noise is unbearable and loud then something is probably actually wrong. If it's just kind of annoying it's probably just a crappy noisy design.

[methodofcontrol]

those 470k's are carbon comp. i replaced R25 and R26 earlier tonight, with no change in noise. it's more than idling noise...something is broken or leaking. every once in a while there's a "pop" noise too...
the caps are all mustards, no discs. but it seems like one of them might be the issue. i couldn't find any DC leaking from any of them. the fact that grounding V3 pin 7 eliminates the noise... there's no reason for me to look into the array of 1m resistors and .001 caps feeding into pin2, right?
is there anything i can ground or jumper on V2 to find out where the noise is coming out of that point?

[Chuck H]

those 470k's are carbon comp. i replaced R25 and R26 earlier tonight, with no change in noise. it's more than idling noise...something is broken or leaking. every once in a while there's a "pop" noise too...

So focus on the plate resistors now. Just replace all of them with metal film. Moisture absorbtion seems to make CC resistor noise problems worse. Best to replace them. But the POP noise could also be a clue if the amp has been exposed to excessive moisture. The plate resistors are the most likely source of noise with CC resistors that have gone funny.

there's no reason for me to look into the array of 1m resistors and .001 caps feeding into pin2, right?

They should be .01 caps according to the schem. And if they're all film caps and not leaking then I would say there is no reason to suspect this circuit... Unless I have neglected to see some phase relationship problem, since this circuit is a phase shifting tonestack. Kind of clever really. Still looking at it.

is there anything i can ground or jumper on V2 to find out where the noise is coming out of that point?

Sure... But before I go into that...

C19 (or C13 can't read the schem well) and C22 (they are the cathode bypass caps for either cathode of V3) are amplifying NOTHING useful in the audio spectrum. I would remove them. They add a 6+ dB boost to frequencies above the useful audio for the amp and could be contributing to hiss. Why are they there? I expect someone thought it was a good idea at the time. It wasn't. Those caps would need to be larger to serve any useful function. Now,

Ground C7, C17 and C20 one at a time at the junction where they meet the 470k mixing resistors and report back.

This is a tricky problem on a mediocre amp. Very tough to do from a remote location.

[methodofcontrol]

i didn't look at the value of those cathode bypass caps on V3 until you pointed it out. wow...those are definitely not useful values. would it make more sense to replace them w/ a bigger value than to remove them entirely?
the 100k plate resistors was something i was looking at. (both on V2 and V3.) i've got replacements, i'll put them in next time i'm in there.
i grounded C7, C17 and C20 one at a time as you suggested at the respective 470k's. grounding any of them took away some noise, but C20 seemed to do a lot more than the others. i had another .01/400v cap laying around, so i replaced it and the accompanying 470k but it didn't seem to help any.

[methodofcontrol]

so, i replaced the 100k plate resistors on V2 and V3. no change.
when i ground pin 7 of V3 all of the noise goes away. when i ground the 470k mixing resistors, each one eliminates a little bit of noise but not enough to leave one area more suspect than another. i'm beginning to think the noise i'm hearing is just the nature of the amp. all of that EQ noise coming into pin 7 of V3 is just, well... noisy, i guess.
correct me if i'm wrong, but the filter caps are in the power section, so by pulling the PI and removing the pre-section from the signal path and thereby removing all of the noise i'm hearing, that eliminates the notion that the noise would have anything to do w/ the filter caps, right?
i did isolate one section of noise that sounds like a bad ground on the back of the pots, i plan to fix that and present the amp to my friend to see if that's what he was talking about this whole time. that would be a quick end to this headache.

[Satamax]

Method, do you have any burgundy CC resistors in there? I've found thoses to be usualy the culprit of the poping noises.

If the bypass caps are out of the audio range, remove theses. IIRC, there's two wires, one yellow and a green around the pi, that if you move theses around cut the noise dramaticaly. On the 120, i've added some local NFB caps between plate and grids. As Mati Adolfsen sugested years back http://www.soundcitysite.com/sc_webpages/maxscmod.jpg

What i see now, is that an easy fix could be done after the three 470k mixing resistors, adding a Π filter between thoses and V3B grid. Cutting out lots , let say above 6 or 8 khz. You'll have to do your own maths tho.