Sansui AU-X517R repair and restoration

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello!
Yes, I'm new here. :)

I am the proud owner of a Sansui AU-X517R that I bought second hand ~10 years ago. It has survived plenty of abuse and has always served me well, but now its starting to show its age. I really like my amp and believe it deserves to be restored to full working order. I have already contacted two people who claim to have the service manual, so hopefully I will be receiving that soon. I am handy with a soldering iron and I have all the equipment I might need (analogue and digital scope, multimeter, etc...), but my knowledge of analogue circuits and especially amplifiers is limited at best. Any help is highly appreciated.

There are several problems that need to be addressed:

1. Input selector
The inputs sometimes start randomly switching around. Sometimes there is sound from only one channel, from a different input than is selected, or even from one input on the left and another on the right... While it is switching I can hear the servo of the weird servo-driven input selector move, so it is somehow connected to that. I expect a purely mechanical problem, though the control electronics could also be involved.

2. Noise
The left channel is noticeably noisy, a lot more than the right. It becomes more noticeable with the volume turned up, even if a disconnected input is selected. Broken caps? Opamp power supply? Something else? This problem has been there for several years.

3. The low frequencies on the left channel sometimes randomly drop out for a few seconds. Sometimes it will do this all day long, sometimes not at all for months. Has been like this for a few months now... It sounds like a high-pass filter, the low end is completely gone. Does not seem to be related to the first problem.

4. The power-LED blinks continuously as long as the amp is on. This has been the case ever since I bought it. The previous owner claimed it has something to do with a broken and bridged overload or short circuit protection (as far as I remember, it's been 10 years). Not quite sure. Some kind of overload protection is still in place, as when confronted with a potentially damaging signal on the input (like the spike when turning on my old computer...) it would mute the output for a few seconds. But SOMETHING isn't working...

If you need any more information just ask. Its still in use at the moment, but I will start dismantling it soon. I can do any measurements or tests you might need to diagnose the problem, just tell me what to do... A list of replacement caps etc. would also be helpful.

Sorry for the long first post, and thanks in advance for any help!

Topy44
 
1. This actually seems to be two distinct problems. Typically you have a stepped attenuator type rotary switch with a resistor network driving an ADC input. Bad contact in the switch means it can't decide between two positions and other such funny things. That one will need a good cleaning. As you mention a channel dropping out, it will appear the actual motor-driven source selector switch may also need some attention. Prepare yourself for some fun as those usually want to be cleaned manually.

2. No clue until I've seen a schematic. Possibly a dodgy semiconductor or bad solder joint somewhere in the preamp.

3. Sounds like an electrolytic with an intermittency problem. Could be the one in the power amp feedback network.

4. Uh-oh. Trace the amp outputs to some relay(s) and inspect the underside of the PCB at these points. You may see some contacts jumpered. Measure DC offset for both power amp outputs, should be a few dozen mV at most. Anything more is reason for concern, anything less indicates a problem in the protection or relay driver circuitry. (Sometimes relay driver transistors develop bad solder joints or die.)
 
Wow, that was a fast and accurate response. Not at all what I am used to from web forums. Thanks! :)

1. Its actually a rotary encoder. This setup is because it has an IR remote, which only switches the internal motor-driven switch, without turning the front panel knob. I believe the problem is actually something else: I think the motor driven stage doesnt have fixed input points, but some kind of contact telling it that it has reached a valid input position. So when one of those contacts opens it thinks its between two inputs and tries to go to the next one. Essentially, I'll have to take apart and manually clean and restore the whole mechanism... Yeah, that'll be fun.

4. To measure the offset I need to measure directly at the output transistors, not at the speaker terminals, correct?
 
Re: 1 -
Mechanical encoders are notorious anyway. Open it up, clean the contact surfaces, and it should be as good as new. Seems to be quite simple in most cases. It appears sometimes they use some sort of silicone for damping in the bearings which then proceeds to spread onto the contacts. Being a pretty good isolator, that does the function no good.

Re: 4 -
As long as the speaker terminals are not disconnected, you can also measure there, of course.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.