Yamaha DX7 Keyboard overheating. +29V going into 7815

[K3mHtH]

Hey everyone,

I have a US Model Yamaha DX7 keyboard that is overheating very quickly. The power supply is outputting all the correct voltages, but checking the input voltages I'm finding them to be very high. +26V going into the +15V 7815 regulator. -29V going into the 7915, and +12V going into the 7805 +5V reg.

Schematic is on Page 5 here: https://homepages.abdn.ac.uk/mth192/pages/dx7/manuals/dx7-circuit.pdf


Is there any component failure that can happen that would cause the increase in voltages like that? Power transformer, rectifiers? I'm a bit stumped on this one.

[nctnico]

Did you check your mains voltage? It wouldn't be the first time mains voltage is too high due to a wiring problem.

[K3mHtH]

Did you check your mains voltage? It wouldn't be the first time mains voltage is too high due to a wiring problem.

Mains was 119 VAC when testing, sorry forgot to mention.

[nctnico]

OK. Did you check the DC voltages with an oscilloscope? My guess would be that the capacitors have dried out and are no longer doing their job so you have excessive ripple on the inputs of the regulators and perhaps also ripple on the supply rails. You can also check using a multimeter in the AC range.

BTW: how old is that DX7? I think these where introduced over 30 years ago?

[K3mHtH]

OK. Did you check the DC voltages with an oscilloscope? My guess would be that the capacitors have dried out and are no longer doing their job so you have excessive ripple on the inputs of the regulators and perhaps also ripple on the supply rails. You can also check using a multimeter in the AC range.

BTW: how old is that DX7? I think these where introduced over 30 years ago?

I never checked the voltages with a scope, but I did check them with my multimeter for both DC and AC. The AC component of the voltages going into the regulators were small enough that I didn't write them down.. maybe .4V. But I will go back and test with a multimeter to actually see what the ripple is. And yeah, this thing is probably approaching 30 years.

Interestingly, the cap that's across the mains (I think it's an RFI suppression cap?) has some visible cracks. All other caps visually look OK.

[wraper]

OK. Did you check the DC voltages with an oscilloscope? My guess would be that the capacitors have dried out and are no longer doing their job so you have excessive ripple on the inputs of the regulators and perhaps also ripple on the supply rails. You can also check using a multimeter in the AC range.

Wait, how bad capacitors can cause DC voltage measured being higher than normal? It can be only the opposite way.
What is meant by overheating, does it stop functioning normally? How big is temperature. Such voltage on Vreg inputs is not something exceptional. If the current is not too high, this can be completely normal. If mains voltage drops to say 100V (at which it must operate, and some tolerance, because it can be even lower than that), then voltage drop on 7815 will be only 6V while 3V drop is a bare minimum allowed, the there must be some margin too.

[nctnico]

OK. Did you check the DC voltages with an oscilloscope? My guess would be that the capacitors have dried out and are no longer doing their job so you have excessive ripple on the inputs of the regulators and perhaps also ripple on the supply rails. You can also check using a multimeter in the AC range.

Wait, how bad capacitors can cause DC voltage measured being higher than normal? It can be only the opposite way.

You are right! Ofcourse the capacitors get charged to the AC peak voltage; there could be some increase if the capacitors are totally shot but not that significant.

[K3mHtH]

If the capacitors are getting shitty and allowing a higher ripple, could that be putting extra strain on the regulators?

[fubar.gr]

If the capacitors are getting shitty and allowing a higher ripple, could that be putting extra strain on the regulators?

Not really, actually quite the opposite will happen.

As the capacitors go bad, mean input voltage will drop, therefore the mean voltage difference between input and output will drop and the regs will run cooler

[Anks]

Maybe measure the current being drawn from the rails. But to be honest this could be normal unless you have noticed a definite change in temperature.