acoustic 370 rustle rustle

[ilyaa]

this amps giving me guff again - have to say im getting a little tired of it.

when on (no input signal) the amp makes a low end kind of rustling noise - almost sounds like rats have made a home in the speaker cones. its, unfortunately, an intermittent problem. sometimes the amp will be on for a while and not make the noise and sometimes the noise will get louder and louder as time goes on.

what ive managed to figure out/try so far:

1) i think its coming after the volume, because the volume has no effect on it (it can even be all the way down and the noise still persists)
2) it might come before the main EQ, because the treble/middle/bass knobs seem to have a *slight* impact on it
3) it definitely comes before the graphic EQ, because that has a huge impact on it. if i turn the graphic EQ all the way down the noise *almost* disappears and if i crank it the noise becomes thunderous
4) pretty sure its in the preamp, because if i send the Line Out to another amp the noise comes with it
5) scoping around in there i dont get too much info.
    scope on the speaker out when its happening: what you'd expect, just some low frequency looking     noise
    DC volts on the speaker out when its happening: a little bit, jumping around between like 15 and 70 mV
   scoping around inside the preamp when its happening: i dont get anything at all really UNTIL the collector of Q105, when the flat line on the scope starts to move up and down (not AC-style, but looking like shifting DC) - it moves along with the rustle. as i move further along the preamp the movement gets more significant but stays the same - just that flat line moving up and down.

my thoughts:

poor filtering feeding the preamp transistors an unsteady/fluctuating bias thats leaking into things like a low frequency AC? only thing is ive checked the main big caps and they seem pretty solid - although of course they are old.....

bad/leaky caps in the preamp? particularly C108, C112, C115, C116, C118, C119, C122, C124, C125, and C127.

am i going down the right path? i dont want to disappoint everyone and resort to shotgunning these caps around but in the absence of a clear indicator of the fault im not sure what to do. unless the shifting DC on the scope points to something specific.

on a side note: im curious about the speaker outs on this amp. with the speaker unplugged, the amp shows like 25V on the speaker outs, but once i plug it in that goes down to near 0. that normal?

[Roly]

First we isolate.

Have you cleaned all the pots (sockets,etc.)?

C404 has to charge.  Should really have a pulldown resistor that gives a fairly CR charge on power up, then it's pre-charged when you plug in the speakers - no nasty thump.

[Enzo]

Your speaker out rises to DC without a speaker because there is nothing there to allow the output cap to charge.


I don;t like shotguns either, but I do expect you will want to change out all those small e-caps like C108,112-115, etc.   But I tend to doubt they are your noise.   Noise like that sounds to me like a semiconductor junction breaking down.

No noise on the scope until the collector of Q105?  That sure points to a noisy Q105 then.  But are we looking at the collector?  Or the emitter?  The collector is wired to the power supply like Q103 and 106, so any noise on the collector of Q105 ought to be on those also.

[Enzo]

Try some heat/cold.  Those transistors in the are of Q105, hit each with some freeze spray, any effect?  Take the tip of your solder iron and briefly touch to the side of each.  ANy change?

[Roly]

If you don't have freeze spray you can dab individual components wet with (dry, fresh) metho using a cotton-bud, then blowing gently to evaporate.

[Enzo]

Those "dust-off" spray cans for cleaning computer keyboards can also do it, turn the can upside down.

ANy aerosol spray will chill, maybe not as much as the freon, but enough to shock a part.


I never tried it, but my grocer sells dry ice.  A small hunk of that could be chipped into service.

[ilyaa]

maybe progress, maybe not.

i dont think its the pots or sliders - its not that kind of rustle, its much more of a light and repetitive thumping - not scratchy at all. but ill try and scrub them again.

(yeah, i meant emitter of Q105 - but i think the noise is there before, just much much smaller, and that's where its amplified enough to be noticeable)

hot didnt make any change, but after i sprayed Q105 with some aerosol (from a deoxit container - all i had) the noise went nuts for a second, but who knows, maybe it was the contact cleaner component of it!

anyway, i changed out Q104 and Q105 - no change, noise still there.

C112, 116, 107 i changed, too.

then i went on a DC hunt and found some odd things: i was getting ~7V on the negative end of C125 and C124. there should not be any DC there, right?! switched them out for new caps and no DC there anymore. i was also getting ~1.5V on the negative end of C108 and C115. that should be DC dead-zone, too, right? i switched them out and it looks better but still about 10 mV in that area....

after turning the amp on after that no noise yet, but its an intermittent problem that usually pops up once the amps been on for a while, so who knows. ill report back....

ack i dont like doing this like this - seems so slapdash. the only things im certain off:

its coming from the preamp
it gets noticeable around Q105
it seems like a DC problem

last thing: emitter of Q105 is only at 14V - should be at 20. maybe R125 has drifted low? or R129 low? earlier this amp had a problem where R104 had drifted high and cutoff the very first transistor in the preamp. amp stopped working until i put an ~8M across it and corrected the bias.

[Roly]

i dont think its the pots or sliders - its not that kind of rustle, its much more of a light and repetitive thumping - not scratchy at all

I agree, this sounds more like a duff component, lumpy leakage; first guess is a transistor, 2nd a cap, 3rd a resistor, not forgetting the physical, dry joints and non-connecting connectors, surface contamination...

Q105 has no voltage gain - it's an emitter follower.


Try shorting or opening the signal path upstream, working back from the input socket until it stops, then it may be the stage you just passed (or came to the input of).

Yeah C124/5 should be grounded via a slider and a 3mH choke.  Well found.   :dbtu:

You ARE making progress, you have it surrounded, but this is in the class of fault we call the dreaded intermittent, and these are frustrating and time consuming, but I know no better way than the methodical slog, stage by stage.  This is when "technishing" gets real.  This is a curly one, but you are well on the way to nailing it, and the harder it is the more you will feel like a million bucks when you beat it - "And WHO's King of the    Castle then?"  Gotcha!  Take them technishes baby (as the dummy load and tech takes a smoko).   :dbtu: :lmao:

[Enzo]

More and more it looks like a noisy Q105 then.

[ilyaa]

think it was the caps around the parametric EQ!

changing Q105 had no real effect - but the sound disappeared after replaced C108 and C115. like i said in a previous post, i was getting about ~1.5VDC on the negative end of either of those caps when that area should be a DC dead zone (right?). after i replaced them - no more VDC and no more noise! think the electrolytic caps were my problem.

so far so good - does this fix make sense?

if those caps were leaky, they would have been letting some DC into that EQ circuit, but in this particular case, that DC, seeing as it was the result of a leak, would look like a fluctuating voltage (right?) and be treated as AC by the amp - amplified and turned into the strange rustling noise i heard. it was early in the preamp, so by the time it got to the power amp the tiny leak was audible.

[Roly]

Sounds like you've found it.

The problem with leakage isn't just that current is going where it shouldn't, it doesn't flow evenly, and as you have observed this resolves to a signal that can sound like rustling paper.

You seem to be going well, a few scalps hanging from your belt now.

Once you start nailing jelly to a wall and making it stick you will find that your name gets mentioned when musos have technical problems and people you've never met will start ringing you up to fix stuff - next thing you know, you'll be a tech.  Keep it up champ.   :dbtu: