A mystery of guitar amp hum to be solved...and I need clues

[oc disorder]

Wasn't going to butt in but I have !!
What about C71 and the possibility of line and neutral swapped?
Surely C71 is not needed although rated at 250v?????

Just thought worth mentioning.

[Dave H]

Your measurement methods seem correct. Steve's explanation about the meter's capabilities may be the reason that you are getting the high AC readings but they still seem way high to me even for a meter that cannot accurately measure AC in the presence of DC. Try putting a capacitor, say 0.1uF /100VDC rating or higher in series with your red probe when you measure the AC. What does that do to the reading?

It's the usual cheap meter problem. It's half wave rectified calibrated RMS for a sine wave and if there is any DC present it reads either 2.22 x the DC or zero depending which way the leads are connected. (38.5 x 2.22 = 85.47). I have tried a series capacitor and it didn't give a sensible reading.

Edit: Red can you get hold of a better DMM to measure the AC ripple?

[TexasRed1974]

It's the usual cheap meter problem. It's half wave rectified calibrated RMS for a sine wave and if there is any DC present it reads either 2.22 x the DC or zero depending which way the leads are connected. (38.5 x 2.22 = 85.47). I have tried a series capacitor and it didn't give a sensible reading.

Edit: Red can you get hold of a better DMM to measure the AC ripple?

So, I borrowed a better meter from work, and retested the power supplies: Now TP14 and TP15 both show around 150mVpp (If I'm reading the meter correctly. An image is attached if it helps). On the other side of the 5w resistors (R95, R96), I also get about 150mVpp or so (this is with no input in to the amp).

Wasn't going to butt in but I have !!
What about C71 and the possibility of line and neutral swapped?
Surely C71 is not needed although rated at 250v?????

Just thought worth mentioning.


[ATTACH=CONFIG]39329[/ATTACH]

Are you saying that I might have wires from the transformer switched to the wrong junctions?

[TexasRed1974]

So, finally getting back to figuring this out.

Measuring AC at the speaker output (red on J5 and black on J15), I get 2.4 Vacpp at ~60 Hz (scope attached). But it's only the positive half of the wave (is this what is considered AC ripple?)


I assume this means that there is something (transistor or capacitor) on the positive half of the amp that is not properly smoothing the signal. Am I at least on the right path?

My plan is to trace this back from the output towards the input until I find the point it goes away or at least changes.

[Jazz P Bass]

Man, that is one funky ass scope trace.

That is not AC Ripple.
AC Ripple looks like a sine wave.

[TexasRed1974]

Man, that is one funky ass scope trace.

That is not AC Ripple.
AC Ripple looks like a sine wave.

Sweet! I'm glad I'm dealing with funky ass stuff.

I'm tracing it back across the schematic. That trace doesn't exist at Q14, so I'm trying to narrow down where it begins to appear between Q14 and J5.

What could cause a trace like that?

[nickb]

Sweet! I'm glad I'm dealing with funky ass stuff.

I'm tracing it back across the schematic. That trace doesn't exist at Q14, so I'm trying to narrow down where it begins to appear between Q14 and J5.

What could cause a trace like that?

The signal at Q14 would only be about 24mV so you might miss it. You could also try shorting the base of Q14 to ground to kill any signal coming in.

It reminds me of the current pulses that charge the main power supply smoothing caps EXCEPT it's 60Hz not 120Hz. Was the waveform on the power supply caps 60Hz or 120Hz?

Another place to check for power supply ripple would be the at the junctions of R66 / C5, and then R69 /C51.

It might be informative to repeat the output signal test but with a 1KHz sine wave injected (no need for a speaker) to check the linearity of the power amp. You can get signal generator apps for phone or use a PC. Very the amplitude to see the result. Anything other than a sinewave out is bad. My thinking it that the power amp might not be working correctly and it distorting the ripple waveform to give what you see.

[TexasRed1974]

So, I am still working on this and I'm continuing to learn things. I wanted to hop back in here and report what I've found.

*The hum coming off the speaker is a 60Hz hum (600mVACpp @ 60hz)
*There's a 40mVDC offset at the speaker.
*There's a 145mVACpp @120Hz ripple coming off the rectifier, at TPs 14 and 15, and of course everywhere there's a 40v marking.
*There's no ripple on the 16v side.

I'm wondering a couple of things:

1) is 145 mVACpp enough to cause the hum?
2) would replacing the rectifier or transformer get rid of the ripple?


I'm searching for the answers to these theories elsewhere, but any thoughts from here would also be appreciated.

[nickb]

So, I am still working on this and I'm continuing to learn things. I wanted to hop back in here and report what I've found.

*The hum coming off the speaker is a 60Hz hum.
*There's a 40mVDC offset at the speaker.
*There's a 145mVACpp @120Hz ripple coming off the rectifier, at TPs 14 and 15, and of course everywhere there's a 40v marking.
*There's no ripple on the 16v side.

I'm wondering a couple of things:

1) is 145 mVACpp enough to cause the hum?
2) would replacing the rectifier or transformer get rid of the ripple?


I'm searching for the answers to these theories elsewhere, but any thoughts from here would also be appreciated.

I don't think any of those are a concern.


That 'funky ass' waveform is - that makes we wonder about linearity. Did you try running a sine wave through the insert jack with a few different Vpp swings at the output and checking the output with the scope?

[TexasRed1974]

I don't think any of those are a concern.


That 'funky ass' waveform is - that makes we wonder about linearity. Did you try running a sine wave through the insert jack with a few different Vpp swings at the output and checking the output with the scope?

Thanks for the reply...I really appreciate your patience, even though I am still not very knowledgeable about this.

So, I did run a 1kHz sine wave through both the input and insert jacks, although I have to admit I had to google what you meant by linearity. I think I understand the general concept of it, but am not sure how exactly it fits in with my hum.

Regardless, I have some observations:

1) When the hum is present, there is a 600 mVACpp @ 60 Hz sine wave at the speakers...this is with no input.

2) Adding the 1kHz sine wave to the insert and/or input jack while the hum is present hardly does anything to the waveform in point #1, although I can hear the distorted tone. The frequency on the output stays around 60 Hz, and the amplitude stays in the half a volt range. But if I turn the sine wave as loud as it will go (either at the generator or on the amp), I can finally get the frequency to increase to about 125Hz and the waveform is pretty erratic and bumpy.

3) A change in the behavior of the symptom since I began this thread is that the hum will come and go intermittently. It'll be present when I power the amp on (sometimes), and stay present for several minutes...but then it will disappear (sometimes suddenly and sometimes gradually). If I keep the amp on long enough after that, the hum will return (again, sometimes suddenly and sometimes gradually). Also, sometimes plugging something in/out of the input/insert jack will make the hum come/go.

4) When the hum is not present, there is no AC present on the output and there is no distortion. When I add the 1kHz tone to the input and/or insert jack, the output waveform is a nice-looking sine wave at 1000Hz. Increasing the volume of the amp or tone signal will smoothly cause the amplitude of the waveform to increase as you would expect. Not erratic at all.

5) There is 40mVDC present on the output whether the hum is there or not.

6) I have to admit ignorance on this last observation. This whole time I have been measuring the output using J5 and J15. But today, I connected the 4ohms worth of speakers to J5 and J15 instead, and measured the output at the speaker terminal. I have been assuming that the presence of a load wouldn't matter, but clearly it does. All that to say, the funky ass waveform is only there if the speakers are NOT connected (and the hum is present). If I measure the output with the speakers connected, the funky ass waveform is not there (regardless of whether the hum is present)...and I see the very nice-looking 600 mVACpp @ 60 Hz from point #1.

[nickb]

Good observations

The bumpy waveform is just your 1Khz sine added to the the hum.

What jumps out at me is (1) the wanted signal path is affected at the same time the hum appears, (2) the hum is 60Hz (3) it comes and goes and (4) the DC offset on the output is not affected. These clues suggest something between the insert jack and the base of Q14 is awry, quite possibly a bad connection. Also it's not 120Hz coming from the DC side of the power supply but more like hum pickup due to a floating ground.

Q28 and Q16 form a power-on soft ( i.e. gently ) mute circuit and a fault there could certainly kill the signal path and may affect the hum. I think there are two possible strategies. Scope AC & DC measurements around that area but especially on the drain & gate of Q16 with and w/o the hum, you might find that by gently pushing on associated components in that area affects the result, or lift the PCB and inspect that area for bad solder joints.

On the 'FA' waveform at no load it may be crossover distortion as was suggested earlier by DH. Let's put that to one side for now.

[TexasRed1974]

Q28 and Q16 form a power-on soft ( i.e. gently ) mute circuit and a fault there could certainly kill the signal path and may affect the hum. I think there are two possible strategies. Scope AC & DC measurements around that area but especially on the drain & gate of Q16 with and w/o the hum, you might find that by gently pushing on associated components in that area affects the result, or lift the PCB and inspect that area for bad solder joints.

I scoped the area around Q28 and Q16. For the most part, any given point I measured would register either as 700mVACpp @ 120HZ or as 0 VAC. I've marked the measurements here: .

I haven't been able to get the hum to go away...I guess I've moved the amp or PCB enough to make the intermittentness disappear.

I've spent some time pushing on components in the area, but nothing seemed to change the symptom. I've also looked for bad solder joints...haven't seen anything yet that is obviously bad. But admittedly, that area of the board has some less-than-pretty joints on it, so I'm going to take a closer look.


I was reading that a soft mute shouldn't have any DC present...could that 10 V (or 40 mV) be causing an issue?

[nickb]

Intermittent intermittents...they're the worst

If there really were a 40ms offset at that point the output offset would huge. It's just a measurement accuracy issue. It would not cause the hum in any case.

Sometimes freezer spray shows up bad connections.

[TexasRed1974]

I've made a right mess of the solder joints and need to clean that up before going forward. Thanks again for your insight!

[TexasRed1974]

Finally found the culprit. With the help of a magnifying glass, I found that one of the leads of R23 looked a little different from the joints around it. So, I resoldered the component, and viola...the hum disappeared!

Thanks again for everyone's help. I'm glad to be done with it