Marshall MG15 CDR Strange Issue

[misterc57]

Had very weak fizzy distorted output. I traced the issue to 4558 IC5A (clean signal at pin 3, bad signal at pin 1).
Replaced IC5 with a new 4558, amp sounded great but after three minutes of playing audio drops and is replaced by a hum.
Pattern... turn amp off, wait, turn amp on. Sounds fine for a few minutes and issue returns.

I have replaced all filter caps, C33, C34, C31, C32, C27, C28. I have a stable V with -14.9 and +14.7 on the rails. However the issue persists, after three minutes of playing the audio drops and is replaced by a hum, while V remains constant.

Simpler scenario... Amp ON, nothing plugged into the input, both volume pots at zero. No sound then after three minutes a hum starts.

Would love suggestions on this one. Schematic I am using is attached. Thank you


Marshall+MG15CDR.pdf

[dmeek]

After sound dies, check for DC voltage at all op-amp output pins. Maybe a missing ground reference somewhere.

[Jon Snell]

The two filter capacitors don't look right ... are they the correct way around?

[misterc57]

The two filter capacitors don't look right ... are they the correct way around?

Now that I look at the schematic it looks like one is backwards,

[misterc57]

So one of the large filter caps was installed backward. Looks like that ruined the cap as it has lost over 50% capacitance so I reinstalled one of the old ones that test good.

However the problem has not gone away.

So I monitored IC5 and was seeing odd V readings. Testing with nothing plugged in I saw V slowly rising at pin 1 and then the hum kicks in after about 3 minutes and then the V at pin 1 was jumping around for a while and eventually V became stable with the following...

rails +/- 17 V

IC5
pin 1. 12.2V
pin 4. -14V
pin 7. 12V
pin 8. 13.3V

[dmeek]

An open wiper on the treble pot could cause this

[Jon Snell]

IC5.
If pin 1 is 12.2v pin 2 will also be as will pin 3. 5a is purely a buffer and is tied to ground through VR6, 5 & 4. Does R13 rise to 12.2v?
Is that what you have?
If pin 1 has 12.2v pin 5 will be close making pin 7 similar as will pin 6

There will be enough information above to find out why the DC level is rising.

[misterc57]

nothing is consistent with IC5

today I have +/- 14.9V on rails and pins 4 and 8, IC1, 3 and 5.

when probing pins 1 and 2 V slowly rise, pin 3 slowly falls, no stable readings after several minutes
when probing pins 5 and 6 V slowly rise, pin 7 slowly rises and is about 5V higher than pins 5 and 6. 5, 6, 7 all different readings

when probing R13 V slowly falls from a few V to under 1V.

IC1 and IC3 have stable low V on all pins with proper supply rails pins 4 and 8

after 10 minutes of probing the rails have fallen to -14.2V and 13.6V, and a minutes later they are back to +/- 14.9.

[g1]

Where are you connecting your black probe, chassis? Is board fully mounted? Sounds like you are not getting correct ground.

[Jon Snell]

If your probe is making the voltage return to zero, then you have a problem with the feed to pin 2 of IC5.
Power it off and measure the resistance between the junction of R13 and R14 to ground. It must be less than 133k.

[misterc57]

If your probe is making the voltage return to zero, then you have a problem with the feed to pin 2 of IC5.
Power it off and measure the resistance between the junction of R13 and R14 to ground. It must be less than 133k.

Board is installed with just the jack nuts holding it on. Black probe is on the chassis frame where a black wire connects the board to ground.

Junction of R13/R14 to ground is over 4M. Reflowed R14 which had a bad joint, now it reads 2.2M (flashing) to ground.

One end of R10, VR5, C15, C16 is at ground.

[x-pro]

See post 6.
Check the contact of the VR6 slider pin with the resistive layer.

[Jon Snell]

You have just found your fault.
There is a 100k resistor R13, a 5k pot, a 20k pot and finally another 5k pot to ground, making a maximum resistance of 130k +-20% to ground..
Which one is open circuit?

[misterc57]

Thank you everyone for guiding me to the issue.

Offender was VR5 A5K Contour pot. I have bypassed for now until I get a replacement and amp is working well.

Probably could have left all those original caps alone.

[Jon Snell]

You are welcome.
That just goes to prove, throwing parts at a project is the wrong thing to do.