FX Loop Buffer Circuit Hum (Not a Ground Loop) (1989 Laney AOR 50 Series II)

[nickb]

C33, C34, and C35 are on the right side of the PCB. They have a + symbol printed to show proper polarity of the cap. They all look correct

Just tack solder (or hold) a 47uF or bigger across R43. That will give you an answer in ten seconds.

[neptoess]

Jon Snell , no access to a scope. Not that I can't afford one. They just take up a lot of space and I've very rarely needed one. The amp thing is a hobby for me.
nickb I like that idea. I've got some 25V rated 100 uF lying around here that would make great test candidates.

[neptoess]

Opened the amp up. Tried some stuff. Took a bunch of measurements while the amp was humming (used the shorting plug in the return jack). All voltages measured relative to chassis ground.

1. No luck holding a 100 uF cap across R43 nickb (with same polarity as C28). The noise remains
2. On the non-grounded side of R43 (the one that's tied to the negative side of C28), I measure 0 V (AC or DC)
3. I measured the -30 VDC supply to the IC at -29.43 VDC with no AC voltage present (measured at D10)
4. I measure -14.5 VDC and 0 VAC at IC3B's output (junction of R44 and C32)
5. I measure -1.2 mVDC and 3 mVAC at the junction of C32 and R45. No clue how this is possible when I measure no AC on the other side of C32. 3 mV is high enough to make me think it's not meter error.
6. I measure -14.6 VDC on the negative side of C29 (this is with or without the shorting plug inserted / hum present)
7. And here's the fun one. I measure 0 V DC or AC at the noninverting input of IC3B (junction of C30 and R41) when the amp is humming.
   1. If I remove the shorting plug, the hum goes away, but I now measure -14.5 VDC at this point
   2. When I plug a guitar into the return, I still measure -14.5 VDC at this point, even when varying the volume knob to the point that the hum goes away. No detectable AC
   3. This means the amp hum is unrelated to the presence/absence of DC on the noninverting input of the op amp (which I guess makes sense, you can't hear DC)

The only thing that jumps out as particularly weird here is C32 (coupling caps do go bad I guess). Does anyone else have any ideas?

[nickb]

Bummer.. I'm gutted I doubt the DCV measurement (2) above as its not consistent with the rest. (5)

At this point i have to ask just how loud the hum is. Put your meter on ACV across the speaker and measure and report with and without the hum.

[Jon Snell]

Stop and think about this a moment.

"2. When I plug a guitar into the return, I still measure -14.5 VDC at this point, even when varying the volume knob to the point that the hum goes away. No detectable AC"

Why is the -14.5volts there?
It is there because there is a DC path, there should be no path. The path is either; C30 or C25.
Remove C25 first then remove C30. Checkout what happens. One at a time!

[pdf64]

...[*]And here's the fun one. I measure 0 V DC or AC at the noninverting input of IC3B (junction of C30 and R41) when the amp is humming....

I query this measurement; surely if it was correct, the opamp would be locked up and unable to pass signal?

[nickb]

I query this measurement; surely if it was correct, the opamp would be locked up and unable to pass signal?

Ah.. but if the output is nailed to the power rail it will follow it which gives the hum.

[pdf64]

Ah.. but if the output is nailed to the power rail it will follow it which gives the hum.

The issue of signal being grossly distorted / cut off hasn't been mentioned though, rather, from what I understand, it's purely a noticeable background hum that manifests when the impedance at the return jack is very low.

[neptoess]

Bummer.. I'm gutted I doubt the DCV measurement (2) above as its not consistent with the rest. (5)

At this point i have to ask just how loud the hum is. Put your meter on ACV across the speaker and measure and report with and without the hum.

Note that I only took this measurement with the shorting plug inserted into the return jack. Didn't feel like popping open a cab just yet, but I did use that Decibel X app to find out that, at about 1 m from the speaker, I measure 59 dBA of hum at about 120 Hz. So that's a discovery, it's not mains frequency hum. I can still open a cab and measure the AC voltage, but I can't imagine this design is inherently this noisy. No one would use it if it raised their noise floor this high. For reference, if I turn every knob on the front of the amp to 10 (including the gain and master volume), pull all the EQ boosts, and put it on the AOR channel, I only measure 57 dBA of 3.4 kHz hiss at about 1 m from the speaker. This would probably be acceptable, because one chord at these settings and you'd be damn near deaf. I've measured this setup at 106 dBA from 5 m without the knobs being at 10.

Stop and think about this a moment.

"2. When I plug a guitar into the return, I still measure -14.5 VDC at this point, even when varying the volume knob to the point that the hum goes away. No detectable AC"

Why is the -14.5volts there?
It is there because there is a DC path, there should be no path. The path is either; C30 or C25.
Remove C25 first then remove C30. Checkout what happens. One at a time!

C25 isn't in the signal path when a plug is in the return jack and nothing is in the send, right? I didn't measure the return jack side of C30 with the guitar inserted, but I can almost guarantee it's 0 with the shorting plug inserted. We do have -30 VDC being introduced at R40, so R40+R41 could be a path. We're also assuming the IC has nothing wrong with it.

I query this measurement; surely if it was correct, the opamp would be locked up and unable to pass signal?

This does bring up an interesting point. If the non-inverting input really is floating at -14.5 VDC, inserting the shorting plug shouldn't provide a path to 0 V. C30 should block that.

So, both pdf64 and Jon Snell have suspicions about C30. I have some 100 nF ceramics here. I could try swapping it. This might also explain why the amp has no noise with either nothing in the FX loop jacks or a simple pass-through cable. C25 blocks any DC coming out of IC3A. However, quick Googling tells me that polyester film caps rarely fail as dead shorts. Also, if it did fail in this way, I should be able to measure a very low DC resistance across it with my DVM. I guess that's my next endeavor.

Anyone else have any ideas? The one thing we keep skimming over is this stupid TL072CP. pdf64 proposed that it could have been hit with ESD. I don't think this is that unlikely of a scenario seeing as it's been in there since the amp was built, and I clearly wasn't the first one inside this chassis when I bought the amp (the previous owner installed external bias test points, and R77 was clearly replaced in the oh so lame fashion of clipping the previous resistor and soldering a new one onto its leads)

[nickb]

I have to give a big thanks to neptoness for doing all measuring, most of it with no prompting.

It's a reshreshing change.

[neptoess]

I have to give a big thanks to neptoness for doing all measuring, most of it with no prompting.

It's a reshreshing change.

I'm an engineer (albeit software / industrial control systems). I don't work with audio circuits, but I've had to do my fair share of troubleshooting.

On that note, I did make a recording of this annoying noise, just in case it helps.
https://www.icloud.com/iclouddrive/0...3L3w#Amp_noise

Starts with the amp on standby and nothing plugged into it to give an idea of the background noise level.
I flip the amp off standby. Nearly no noise coming from it (other than general tube amp hiss)
I strum a few times to give an idea how loud an unamplified guitar is
I plug the cable into the return jack. Signal path is now guitar -> 9V powered Boss SD-1 with vol and gain knobs on 0 -> Amp return jack. There should be effectively no signal, but the hum is clearly audible.
I strum a few more times to contrast unamplified guitar vs hum
I bring the level on the SD-1 up a bit (9 oclock or so) so it's actually sending signal into the return
Play a bit to highlight that the hum is constant
Bring the gain up on the SD-1 and play a bit more (the distortion kind of masks the hum)

[neptoess]

Next time I open the amp, I'm going to:

1. Measure DC resistance across C30 to sanity check that is isn't a dead short
2. Remove IC3 and test it on a breadboard. I haven't touched op amps since my analog circuits lab in college, but a quick search reminded me how dead simple these things are https://circuitdigest.com/forums/ana...#comment-28421

[neptoess]

Also, I may very well just take this to a shop if this next venture doesn't pan out. These guys are only about an hour drive away http://earcosrepair.com/, and their labor is only $50 / hr. I've spent quite a few hours with this stupid thing myself, and my time is certainly worth more than that

[Helmholtz]

Just summarising:

Because of the single -30V supply, the output pin of IC3B should sit at around -15V. In order to achieve this, BOTH opamp inputs need a bias voltage of around -15V.
This -15V bias is produced by the equal value voltage divider resistors R40/R43, filtered by C28 and connected to the positive input via R41.
The negative input gets its DC bias from the output via R44.

I agree that the most likely problem is a leaky/shorted C30 which would pull the positive input towards 0V (actually the output should follow) when the return input is shorted to ground.
Doesn't need to be a dead short. E.g. a 1k leakage resistance would reduce the bias to -0.15VDC (voltage divider 1k/R41).

And indeed some of your measurements don't seem to be consistent.

[pdf64]

Next time I open the amp, I'm going to:[LIST=1]

Try lifting that chassis connection?