Boss GE-7 Funkyness

[OuthouseHuddle]

Hello,

I've got a Boss GE-7 that's started acting up ever since I replaced the stock op amps. It sounds very distorted, but intermittently works just fine for very short amounts of time .

Here is what I know:

1. It usually will work right immediately after applying power for a second or two, then it's back to crackling. Jostling it around dosen't make a difference, and it's intermittent as it sits there.   
2. The distortion stems from the first IC. It goes in as a normal signal, but comes out and the bottom of the waveform is clipped off, implying bad bias voltage. I assume that the virtual ground should be set at about half the supply voltage, since R20 and R23 set it, and they are both 33k resistors. Right now, the voltage at that divider is less than a single volt. Schematic Here: https://www.hobby-hour.com/electronics/s/ge7-equalizer.php
3. I tested resistors R20, R23, Caps C18 and C17, and they all seem to be fine.
4. Not of the solder joints look bad, and I've redone most of them

I'm not sure where to go from here. Does anybody have ideas what could be pulling down the bias voltage so much? If there's anything I should test, please let me know.

Thanks

[ElectricDruid]

A overheated cap on that divider might do it. Sometimes if you hold the soldering iron on them too long, you can get the insides to melt through the insulation that makes them a capacitor and they'll short out. So I'd be suspicious of your C18 electrolytic.

That said, there's a lot of other parts of the circuit connected to that bias voltage, and if *any* of those connections have a short to ground *anywhere* then the bias voltage will get dragged down. You'll hear that first in the first op-amp, but that might not be where the problem is.

HTH,
Tom

[OuthouseHuddle]

Same issue with C18 completely removed, and unfortunately I don't see any obvious shorts anywhere.

Would the next logical step be to take up a leg of everything connected to the bias line one by one, checking voltage each time till I narrow down the problem area?

Is there a better way to narrow down where the draw is that I'm not thinking of?

Thanks

[Sooner Boomer]

If it *is* a thermal problem (part heats up after power on and goes bad), a quick spray with cool spray should bake it work again.  If you can't find cool spray, canned air turned upside down so liquid comes out will work.

[Rob Strand]

Same issue with C18 completely removed, and unfortunately I don't see any obvious shorts anywhere.

Would the next logical step be to take up a leg of everything connected to the bias line one by one, checking voltage each time till I narrow down the problem area?

Is there a better way to narrow down where the draw is that I'm not thinking of?

You have done well so far.

I don't know if it is worth lifting each Vcc/2 tap point.  They are all high impedance.
I can't see anything obvious that would pull it down (other than C18).

Some things come to mind:
- Have you checked the +9V rail?   Maybe the whole power rail is dropping.
- Check the DC levels at the opamp outputs and input.  If the DC voltages all track then clearly Vcc/2 is dropping.
- Are you sure its not oscillating?  If you have an oscilloscope it will be obvious if it is or not.
- Some sort of short on the PCB.   That would create a behaviour which is outside what you can see on the schematic.

[OuthouseHuddle]

The +9 rail is stable, and there is DC at the inputs and output of the first opamp. The voltage on the input is higher than at the divider though.

I think it might be thermal. I don't have compressed air or anything to spot cool it down, so I threw it in the freezer for a while.

Here is the video of the scoped voltage of the divider (R20 and R23) while cold as I plug in power.

When it was room temp, it did the same thing, except the voltage dropped down much quicker. Then If I unplugged and replugged the power quickly, it would just start at less than a volt and stay there. However if I wait for a little while and try it again, the voltage will jump up again and then fall, like it does in the video, just quicker.

This would imply some component shorts out as it gets warm, right?

All of this is with C18 and C17 completely removed.

[Rob Strand]

This would imply some component shorts out as it gets warm, right?

It seems so, probably a cap.  Maybe a resistor opening. A bad solder joint could also do that.

In the context of that PCB,  the key is Vcc/2 changing.    There's not much that can affect it on the schematic but we are definitely seeing the effect.

At this point I'd probably make a separate vcc/2 divider from 2x10k's, lift the Vcc/2 side of R29 (470k) then connect only that point to the new divider.   That way the biasing for IC3, IC4 and IC5 is largely separated from the fault.  Currently the interactions between vcc/2 and the fault make it hard to narrow things down.  After that you should be able to measure the DC voltages on the other opamps and around the caps.   I suspect you will find the fault somewhat quicker that way.

Be very careful with the wires on the new divider.  You don't want them to short on something and take out IC5.  You could actually wire the new divider directly to pin 6 of IC5 but there's more risk of the wires shorting.  You wouldn't need to lift the resistor leg with that method.

[PRR]

Cool 'scope!!

[R.G.]

I hate to sound like I'm beating a dead horse, but have you tried measuring the DC voltages on the opamp pins?]

Measuring the DC bias conditions is a very-much-overlooked first step. It may not turn out any useful information EXCEPT to confirm that you don't have a bias problem.

Your description seems to say that the circuit erratically stops working in a distorted manner after a short time, and resets to "good" if it's powered off. That could well be the DC conditions drifting out of the linear operating zone.

So I'd recommend:
1. Get it to fail.
2. Measure DC on all the opamp pins while it's failing, keeping track of which pin measures what voltage.

This will flatly tell you if your bias network is not working or is ineffective - or point to a host of other sins.

Just out of curiousity, why did you replace the opamps?

[Kingmafw]

Stupid question, but which stock OpAmps did you've changed? And what are the alternative OpAmps you've replaced?
In the past I've played with some alternatives, but never had a distorted sound.

[ElectricDruid]

This would imply some component shorts out as it gets warm, right?

Sometimes this kind of thing is seen as a capacitor charges up and shifts something away from where it's supposed to be. That might be a capacitor charging as it's supposed to, or it might not be supposed to.

Dunno if that's possible in your situation, but it's a possibility in many cases.

T.

[ayayay!]

Did you swap out IC5?  That's quite a different chip IIRC.    What did you change, part for part?  Were the originals RC4559's?  Different voltages...