Stan Dye via groups.io <standye@...> wrote:
These two observations tell me that your problem is 'to the right' of the capacitor.
That's what I suspected as well.
After spending (too much) time on it, I think I figured out the source of
this buzzing (not the cause, just the source).
First, these 25 MHz oscillations were probably caused by the oscilloscope
probe. They're irrelevant.
There are 100 Hz spikes on inputs of two of four op amps (and, what's
interesting, on outputs of OTHER two op amps).
These spikes are not in sync with mains (in Poland I have 50 Hz mains, so
that was my first suspicion).
What in the QMX generates such spikes? Turns out that CLK0 and CLK1
signals are interrupted for 200 microseconds at this rate (100 Hz).
Why, I have no idea.
I replaced op-amps, replaced C416...C419 (I used 22n, I didn't have 33n,
but it didn't seem to make any difference), finally replaced C406 that you
mentioned (with 2x 1u, the same reason as above). Nothing changed.
So the source is located, and it would even make sense for these spikes to
appear on two of four outputs of IC403 (if the clock signal was always
interrupted in one or two "phases" of the detector -- but I didn't measure
both signals at the same time, so I'm only guessing).
I guess the proper question would be: could you / someone else look at
CLK0 and CLK1 lines (IC403 pins 2 and 14) and see if you see these breaks
there? It's hard to spot on the oscilloscope...
Maybe Hans can confirm that the CPU indeed does such thing? Or deny? Maybe
it's some weird firmware bug?
Have you checked carefully R401/R402/C406? A bad C406 in that biasing
circuit could cause oscillations and disturb the audio.
Yes, now I did (replaced C406). I didn't touch R401 or R402, as the
biasing voltage seems correct.
Are you certain that there is no anomaly with T401?
At this point I'm not sure of anything... but I don't think, given this
new discovery, it's T401. Plus these spikes are not present on its outputs
(pins 7 and 9 of IC403).
Is the buzzing only on 'left' or 'right' audio, or both?
Spikes are on inputs of IC406 (so right) and outputs of IC405 (so left).
Crazy.
And finally, as unlikely as it seems, IC407 has been the most common
failure source of anything related to audio in the QMX.? If none of the
previous apply, I would replace that one.
When I removed op-amps, the buzzing stopped, so I guess IC407 is fine...
btw - you are the third person now recently who I have heard blew the
IC403 with a strong overvoltage burning through the 5V supply, myself
being one of them (I inadvertently turned QMX+ on with a 17.5V supply).?
I don't know why the nearly identical IC402 didn't also blow.? Maybe
it's the doubled bias across the Z/W pins.
It would make sense. Maybe the chip can sustain overvoltage on VCC, but
not on Z and W (their input protection diodes would normally clamp it to
VCC, but with VCC also being too high it didn't happen)...
BTW, I recorded videos (with oscillograms) when I worked on it, partly in
case I needed some more advice, partly to document what I was doing in
case I got confused later. I summarized everything above, but here's the
raw footage anyway (not edited or cut, just merged).
