Squeal from a Prototyped Preamp into an Existing Power Amp

[Dino Boreanaz]

On the advice and encouragement of members here, in an attempt to get a more guitar-like tone from my bass amp, I have prototyped a Marshall 5002 (Lead 20) preamp on a breadboard and run this into the power amp of my Marshall 3505 (Micro Bass).  The result is everything I hoped it would be ... brighter tone, clearer mids, and a really nice overdriven distortion at higher gain settings.  Due to parts availability, I used slightly different pot values from those shown on the schematics.  I used 25k instead of 22k pots for the gain (VR1) and mid (VR4), and a 250k instead of a 220k pot for the treble (VR3).  I've attached the schematics for reference - the 5002 is the top portion of the hand-written schematic showing multiple amps.

There is just one issue that I need to fix before I make this a permanent modification.  There is a high-pitch squeal when I have the gain, treble, and mid all at maximum (10).  The squeal changes slightly in pitch and loudness as these three controls are slightly lowered, and is eliminated at gain settings below about 8.  By changing the capacitor in the second-stage feedback loop (C2) from 220pF to a much higher value like 0.01uF (10,000pF), the squeal is also eliminated, but all the treble is gone and the amp sounds very muddy.  I've tried adding capacitors in parallel to gradually increase this value in hopes of finding a compromise between eliminating the squeal and retaining some brightness, but it seems that any values that are high enough to eliminate the squeal also take too much treble out of the tone.

I'm assuming a stock Marshall 5002 does not squeal at any combination of gain, treble, and mid settings.  Is there some interaction between the 5002 preamp and the 3505 power amp that is different than the stock 5002 which would cause this preamp to behave differently in front of the 3505 power amp as compared to its intended combination.

Are there circuit changes (different value components, additional components, etc.) that I could make to eliminate the squeal while retaining the current clear, bright sound at the highest gain settings?

One final question, possibly unrelated to the squeal issue.  I measured the supply voltage to the op amp and found that it measures -13.9V and +10.2V.  These values seem quite different than the +/- 15V specified on the 3505 schematic.  Should I be concerned about this even if it is not related to the squeal?

[g1]

Breadboard builds are more prone to oscillation because of all the wires etc. and layout issues.
I'm assuming you were testing values for changes you were going to make to the 3505.
If so, just go ahead and make the changes to the 3505 circuit board.  You may find the squeal is no longer there because of better layout/lead dress.

The difference between the + and - low voltage supplies seems a bit much.  Does it have the zeners that are shown on the 3505 schematic?  Have you checked the 2K7 resistors (R17 & R20) and the 15V zeners?  I see there is a bit of an error in the 3505 drawing as there are 2 zeners marked ZD3, one is a 15V and one is a 9.1V.
Make sure they are in the right spots.

[Dino Boreanaz]

I am planning on putting the 5002 preamp into the 3505 chassis, but it will require a new board since the circuit is quite different.  So I can't put the new components into the existing 3505 board and there will be some additional length of wire to connect between boards and from the new preamp board to the pots.  I'll have to try moving things around to see whether the squeal changes as I move the long jumpers that I currently have running between the breadboard and the existing 3505 board.

As for the op amp supply voltages, the zeners are in place.  I've tested the resistors in-circuit with a multimeter and the 2.7k resistors seem to measure right around 2.7k.  With the power on, the two 15V zeners measure 13.9V and 10.2V and the other 9.1V zener measures 8.6V.  It does seem like one of the 15V zeners is faulty; what effect would this have on the sound of the amp?

[Jazz P Bass]

That + Vdc supply is not working correctly.

I would suggest that you measure the voltage on both supplies .
+Ve & -Ve.
They should be more or less the same except for polarity.
Flip your meter to read Volts ac & measure them again.
You want a low amount of ripple.

If this test passes, measure after the 2.7K resistors.
Same thing.
Volts dc and then Volts ac.
The Volts ac ripple should ideally be zero.

The zener diodes will 'dump' the voltage that is higher than the zeners voltage rating to ground.

If you do not have enough voltage going in to the diode, it cannot make the difference up.

[Dino Boreanaz]

I measured the DC voltage before the 2.7k resistors and both + and - were nearly identical at about 22.7V to ground, 45.3V between them.

I could not measure a stable AC voltage before or after the 2.7k resistors.  The readings continuously ranged from a few hundred volts down to less than 0.1V.

The DC voltage after the 2.7k resistors measured the same 13.9V & 10.2V as I had previously measured, but this is in-circuit so I don't know whether it's due to the zener "dumping" too much voltage to ground or the resistor dropping too much voltage before it gets to the zener.

So I also measured the voltage drop across the 2.7k resistors and this accounts for the difference.  The voltage drop across one is 8.8V and the other is 12.5V.  These add up as you'd expect: 13.9V + 8.8V = 22.7V and 10.2V + 12.5V = 22.7V.

Then I measured the resistance of the two 2.7k resistors again and found that the resistance of the one on the side with 13.9V started significantly lower than 2.7k, but slowly rose to over 2.6k.  The resistance of the resistor on the side with 10.2V read a very stable 2.63k immediately and its reading did not change.  Does this indicate that the capacitor on the 10.2V side is bad?  Since these measurements are all made in-circuit, I would assume that the measured resistance of the resistor should change as the meter charges any connected capacitors.  Does this sound right?

[Jazz P Bass]

" I  could not measure a stable AC voltage before or after the 2.7k resistors.  The readings continuously ranged from a few hundred volts down to less than 0.1V."

What kind of meter are you using?

If the Vac ripple is jumping around like that you either have a bad load downstream or the caps are bad.

[Dino Boreanaz]

Yeah, I wasn't sure what to make of the Vac readings either.  I'm using a digital multimeter, nothing fancy, but it has always given reliable measurements.  I can try re-measuring some of these values.

Based on the difference in the resistor readings, I'm suspecting bad caps.  I've probably gone as far as I can without taking some components out of the circuit, so this may have to wait a little while before I can devote some more time to it.