Parasitic oscillation in my Vox AC30

knuckleberryfinn

I was just testing my AC30 before trading it, and have discovered that the Vib/Trem channel launches into howling if I go beyond halfway on the volume.  Bugger.  It is dependent on having some input signal, so if I hit the strings hard at halfway, it'll go off on one doing a whale-in-hell impression.  But equally, I can turn it almost up to full and brush the strings very lightly and it won't happen.  This is why I am thinking it is parasitic.  Any suggestions?  I've resoldered the ground on the volume control which was dry, and which would launch it into feedback when I nudged it with the faithful old chop stick.  The volume control is right before the input to the PI, so will it be something around the PI feeding back into an earlier part of the Vib/Trem circuit?  Both the other channels work fine, though there is a faint ticking audible when the volume is very high on the Bright or Normal channels and the Vib/Trem is on.  I've been going through the circuit all day and have pushed wires about and re-soldered anything that looked dodgy.  Oh and I replaced the Vib/Trem bias bypass cap.  So I'm a bit stumped.  Any help much appreciated.

EDIT: I received the amp non-working in a trade, and so I'm not sure how long this problem has been here.

Here's a link to the not very clear schematic: http://www.korguk.com/voxcircuits/circuits/ac30dals.jpg

ICBM

Probably a microphonic valve that's being tripped into self-sustaining feedback from the speakers by a bit of volume or a disturbance. Try changing V1.

knuckleberryfinn

Now why didn't I think of that?  It's the most obvious thing...

...but unfortunately it wasn't any of the valves.  I've now replaced the bias caps and resistors for the vib/trem input and the oscillator, and still it goes into meltdown.  It sounds sort of like a massively distorted theremin going wildly up and down in pitch.  Any further suggestions?

knuckleberryfinn

I'm wondering if it's a bad Vib/Trem volume pot - the one just below the PI on the schematic.  Even after re-soldering the earth at this pot, it still crackles and pops when I nudge the tags.  If the wiper lost reference to ground for an instant, would the signal level going into the PI rise enough to send it into feedback craziness?  Reducing the volume stops the feedback,, so maybe not. This is a poorly educated guess, so ll opinions are welcome before I replace the pot...

NervousJohn

Some questions.

Have you tightened up all the screws in the chassis? You're depending on these for grounding.

Does the theremin sound change with switching the vib trem switch, or the speed knob?

My gut feeling is something in the oscillator circuit is coupling back to the preamp somehow.

knuckleberryfinn

I did tighten the valve screws, some of which were loose.  The earth seems fine now, but I'll go over it again.  

The sound is so loud that I haven't experimented with switching the trem and speed knobs.  I'm worried about it destroying the speakers or causing other problems.  I've had enough chasing it so I've just emailed a local amp tech - hopefully he'll be able to track it down with his fancy diagnostic gear.

Thanks for the input though Mr Nervous.

ICBM

Certainly an odd one that I haven't come across! Not something I can make a good guess at from here, so I think your tech is the best bet.

ecc83

Just a point about the loud noise!

If any of you are contemplating work on amps a load box with a highly attenuated output is a virtual necessity.

Does not have to be "clever" or sound "rock n roll" just allow you to listen to faults in comfort and speaker safety.

Note, you MUST have resistive loads but a 100V line speaker transformer makes a super attenuator, a 3watt jobbie is cheap enough as are the loads (or use a fan heater!)

Dave.

ICBM

ecc83 said:

(or use a fan heater!)

He isn't joking .

A lot of old 2kW-3kW range heating units - heaters, kettles etc - have DC resistances in the 20-30 ohm range, which is ideal as a "16 ohm" dummy load. The exact resistance doesn't matter much, although it's a good idea to check it. Also make sure you disable any off switch, or at least tape it securely - you don't want to accidentally set your dummy load to open circuit.

knuckleberryfinn

I've been meaning to make a dummy load for a while - I ordered in a couple of 50 watt 6.8-ohm resistors.  I've got a metal box somewhere to mount them.  So I could simply use the resistors as a 13.6-ohm (series) or 3.4-ohm (parallel) dummy load, but then I wouldn't be able to hear when the amp went into oscillation.  How do I use the dummy load to attenuate volume?  Do I place it in series between the amp and speaker(s)?

ICBM

Curious as to why 6.8 ohm? That's not a very useful value either in series or parallel. Or singly…

Although in series is not too far away from the average resistance for an 8-ohm speaker.

How much do you want to attenuate? If it's for 'musical purposes' to let you overdrive the amp more, I'd be inclined to put one in series and one in parallel with the speaker, which will actually give you close to a true 8-ohm load with around a tenth of the power going to the speaker (very roughly!), ie around a 10dB reduction.

If you want to use it for testing you're best to use them as a simple dummy load and take a small signal off with a fairly high resistance in series to the speaker so you don't get much volume at all.

knuckleberryfinn

ICBM said:

Curious as to why 6.8 ohm? That's not a very useful value either in series or parallel. Or singly…

[...]

If you want to use it for testing you're best to use them as a simple dummy load and take a small signal off with a fairly high resistance in series to the speaker so you don't get much volume at all.

I ordered them ages ago - probably they were the closest 50W resistors available.

I'd want it to use as a dummy load.  What kind of resistance would I need to the speaker?  Could I use a pot to alter the volume?

ICBM

knuckleberryfinn said:

I'd want it to use as a dummy load.  What kind of resistance would I need to the speaker?  Could I use a pot to alter the volume?

Yes. I'd probably put a 100-ohm resistor in series with a 1K pot, or something like that - to stop it going to straight-through, to both limit the volume and prevent the pot contact from burning out if you get it very close to the end but not quite.

knuckleberryfinn

Okay, thanks.  I'll get on it...eventually.

jpfamps

Is this a HF or LF oscillation?

knuckleberryfinn

It starts off low-ish, and wavers up and down howling - like a theremin - until it reaches very high pitch squealing.  I turn it off or turn the volume down within a couple of seconds, so this all happens pretty quickly.  I'm not even sure if it's parasitic oscillation.  It could be a bad ground I suppose - but I've checked the grounding and it all seems solid.

ICBM

That sounds like it's something to do with the vibrato mixer stage, although it's hard to be sure - sometimes 'normal' stages do things like that when they're unstable.

Have you re-capped it? It could be the 32uF filter cap which feeds both the Vib/Trem channel gain stage and the mixer stage - if it's not decoupling the supply well enough, signal from the mixer could be getting back into the gain stage. If so it has to be partly to do with the main power supply since both these stages are before the volume control - the only thing after it is the phase inverter and power stage, so if it's triggered by turning the volume up it must be something to do with that… I think!