< sorry , just catching-up here... will have a think.. >
& BTW that's a great little scope to have found, Mat: Hitachi 'scopes like that are both tediously-reliable, and have a really useful/ simple /intuitive control interface. Hope you enjoy it!
(Now - to be simple-stupid - do you get the same results if you re-run the basic tests, and just swap the 'scope channels doing the testing..?)
& BTW that's a great little scope to have found, Mat: Hitachi 'scopes like that are both tediously-reliable, and have a really useful/ simple /intuitive control interface. Hope you enjoy it!
(Now - to be simple-stupid - do you get the same results if you re-run the basic tests, and just swap the 'scope channels doing the testing..?)
The signal doesn't appear to change regardless of which end I put the probe (load or no load).
No load:
Load:
A lot of scopes on ebay are used and abused, with no guarantee they work properly. This one was serviced (capacitors replaced, calibrated etc), and came with two brand-new probes, so I thought it was worth the £120 that I offered. Tektronix seem to be the most popular analogue scopes, but they cost quite a bit more.
Yes, it is surprisingly simple to use. To the untrained eye, they look quite complicated. I can see why people suggest getting familiar with an analogue scope before going digital, though. All those menus look like a nightmare.
The tests above were done with the other channel of the scope. I'll do some proper A/B comparisons later...
No, they are not the same. The no-load signal isn't clipped. The loaded signal is. This bears out the T8/T10 theory.
NB There is no point in looking at ten or more cycles at once. Change the horizontal so you can only see one or two cycles.
NB There is no point in looking at ten or more cycles at once. Change the horizontal so you can only see one or two cycles.
I was playing with the sim and a high resistance coil actually came fairly close to the fault symptoms and has massively different signals on each end of the coil as the bridge tries to correct a non correctable error.
I think we have to try them if only for elimination purposes. The amp should maintain an output of up to 4 volts peak into 8 ohms without the lower NPN output even fitted.
Also I'm acutely aware this issue was initially caused by work being done on the amp and logically the area to concentrate on is where problems have actually been found, the bridged solder joints, the coil actually being open and so on.
Whatever is going on here is I think going to be something simple, its just finding what has happened.
T7 has say 80V available through 1120 ohms. That's < 80mA, and across 8 ohms that is well under a volt.
The load/no load suggestion was yours, and a good one, but you seem to be ignoring what its visible failure in the form of negative clipping clearly implies.
Correct. However the amp does maintain 4 volts peak into 8 ohms without the dumpers fitted. This shows the amp at the point of negative clipping. There is 34ma in the 1120 ohm and 26ma in TR8 under quiescent conditions. Under drive conditions TR8 is supplying up to 500ma without the dumpers fitted.
Its an interesting fault so so difficult to diagnose at the end of a keyboard.
An idea.... its a good fault is this 😉
Is R19 (0.1 ohm) OK and not open?
With the amp on (no signal) can you measure the DC voltage across D8 and D7. D7 looks like it should have around 650mv across it, D8 much less.
With the amp OFF can you measure on ohms and diode meter ranges across D8 and D7. Make sure to account for meter lead polarities if reading initially in circuit. Compare with the other channel.
Also can you read across T9 base/emitter on ohms and diode ranges.
Are any of those readings low on the bad channel compared to the good channel?
Also with the amp on and no signal applied what DC voltage do have across R20?
Is R19 (0.1 ohm) OK and not open?
With the amp on (no signal) can you measure the DC voltage across D8 and D7. D7 looks like it should have around 650mv across it, D8 much less.
With the amp OFF can you measure on ohms and diode meter ranges across D8 and D7. Make sure to account for meter lead polarities if reading initially in circuit. Compare with the other channel.
Also can you read across T9 base/emitter on ohms and diode ranges.
Are any of those readings low on the bad channel compared to the good channel?
Also with the amp on and no signal applied what DC voltage do have across R20?
Sorry for the delay, I caught the bug from my son and have only just recovered 🤢
The only thing that stands out is the voltage reading across R20.
R20 Left (bad channel) = 0v
R20 Right = 0.5v
Everything else looks fine apart from D8 is slightly different. Left (bad channel) 105mv - Right channel 75mv
The only thing that stands out is the voltage reading across R20.
R20 Left (bad channel) = 0v
R20 Right = 0.5v
Everything else looks fine apart from D8 is slightly different. Left (bad channel) 105mv - Right channel 75mv
The R20 voltage has to be a clue. I think ejp mentioned replacing T8. Did you try that?
D7 might be worth a look. Did that one have around 600mv across it? If very leaky it does seem to give a similar symptom to what you have.
I just feel I'm missing something obvious with this but I can't see it.
D7 might be worth a look. Did that one have around 600mv across it? If very leaky it does seem to give a similar symptom to what you have.
I just feel I'm missing something obvious with this but I can't see it.
No, I haven't replaced T8 - yet...
I removed D7 and tested it with the peak atlas DCA75. If it was faulty, wouldn't the DCA75 say so? I also put the multimeter on it in resistance mode, in both directions and couldn't get a reading.
Edit: Yes, D7 measured around 600mv
I removed D7 and tested it with the peak atlas DCA75. If it was faulty, wouldn't the DCA75 say so? I also put the multimeter on it in resistance mode, in both directions and couldn't get a reading.
Edit: Yes, D7 measured around 600mv
Just swapped T8 over and the problem moved to the other channel. It looks like we've found the culprit 🙂
I'll order a new one and report back....
I'll order a new one and report back....
Well that seems pretty conclusive which is good.
Very strange what has happened though, the history of the fault and all that 🙂 Oh well.
Very strange what has happened though, the history of the fault and all that 🙂 Oh well.
I've replaced the offending transistor, and the clipping is gone 🙂
The right channel sine wave is a slightly bigger than the left. I'm guessing this could be the volume pot? Well, I suppose it could be anything. Does it matter?
The right channel sine wave is a slightly bigger than the left. I'm guessing this could be the volume pot? Well, I suppose it could be anything. Does it matter?
The difference in amplitude might look a lot but in practice its miniscule. Also be sure to check its not the scope. Use one channel and make the measurement, then transfer the probe to the other channel.
As you say, most likely a pot matching issue... but its actually very close.
Channel 1 lets say is 6.3 volts peak to peak and channel 2 lets say is 6.7 volts peak to peak. That is 0.5db matching error which is close.
As you say, most likely a pot matching issue... but its actually very close.
Channel 1 lets say is 6.3 volts peak to peak and channel 2 lets say is 6.7 volts peak to peak. That is 0.5db matching error which is close.