Sansui AU-7700 Power supply or Driver Problem

Zekeman

AK Subscriber
Subscriber
Sansui AU-7700 Power Supply Voltage Problem

I have an AU-7700 on the bench that is producing different voltages off the bridge diode configuration that powers the output rails. See pic. The bridge is getting 32 volts in but producing -45 volts on the one side (within spec) and 34 on the other (not within spec).

These diodes are the large metal types on many power supplies. The schematic shows them as SS-5 and SS-5R. Has anyone ever seen these go bad? I have read in other threads that if these were bad the unit would not work at all and I tested the preamp out into a different power amp and the preamp works fine.

See pic and any advice is appreciated!

Thanks,

Zeke
 
Last edited:
I would first see what the diodes do in a no load condition.. The filter cap might be shorted…..BUT based on the numbers I would guess 1 diode is open on the POS rail OR bad connection??
 
Yep, try as the guys have said, and... those rectifier diodes are regarded as almost indestructable.

I have personal experience of this, as I momentarily shorted one of them on my AU-717 - some sparks - but no damage to the rectifier :thmbsp:
 
Last edited:
Yes, although in this case, this might be something you would get if the capacitor is open circuit instead of shorted. If you have AC, AND you have a 50 volts capacitor around of at least 500 uF or so, you can use it to bypass it and see if the DC voltage raises, if you have no other means of measuring it.
 
A scope would make this an easier " call "….
I agree the Diodes are " bullet proof ".
Caps are suspect.
Cold solder joint on the diodes
Something loading the (BLANK) on the power supply…
15+ volts are going SOMEWHERE or NOWHERE!
:lurk:
 
Thanks for the responses. I too have shorted these diodes in other amps and the kept on ticking! I yanked both of these last night and the were not shorted... Both measure ~0.42 on the meter.

So you want me to measure the pos and neg sides with meter set to AC? Will do -but what will this reveal? Also...I do have a scope. I have only used it for measuring output (watts).

Regarding the filter caps...this amp has 10,000 if caps. I do have some salvaged 15,000uf caps so could I use these to test or would I eff it up?
 
Yep, try as the guys have said, and... those rectifier diodes are regarded as almost indestructable.

I agree the Diodes are " bullet proof ".

Just a heads-up; *never* assume a component is bullet-proof. Every component will fail at some point; it's just a question of when.

I'll agree that most decent diodes will take some abuse; but always suspect a component untill you can verify it's not the issue. My old boss lived by the theory of "that part never fails"; there were several times that thinking failed him.

Suspect everything.

this amp has 10,000 if caps. I do have some salvaged 15,000uf caps so could I use these to test or would I eff it up?

That's fine. Since the unit isn't tube rectified; they can handle the extra current draw to charge the caps. They would also work as a replacement. A little extra power-filtering never hurt......

unless you put a 10,000uf cap on a tube rectifier...that'll kill the tube.
 
Thanks for the responses. I too have shorted these diodes in other amps and the kept on ticking! I yanked both of these last night and the were not shorted... Both measure ~0.42 on the meter.

So you want me to measure the pos and neg sides with meter set to AC? Will do -but what will this reveal? Also...I do have a scope. I have only used it for measuring output (watts).

Regarding the filter caps...this amp has 10,000 if caps. I do have some salvaged 15,000uf caps so could I use these to test or would I eff it up?
It will reveal if the capacitor is open or not. If you have a lot of AC, the capacitor is open. Or just parallel one 15000 uF to the suspect 10000uF and measure DC again. If the voltage is okay then, cap is open circuit also. If the 15000 caps fit and are newer than the 10000 ones, you might just put them both in anyway.
 
Zeke
If you measure AC ( where the DC should be ) and get a small amount, that's ripple if there are MANY VOLTS AC a diode has failed.
 
OK gents here's what happened;

I swapped out the filter caps with two good 15Kuf caps into the circuit - Result=same

I measured for AC off of both positive terminals of the filter caps and one measured 0 however the other measured a consistently fluctuating stream of values from zero to other positive and negative values in a consistently circuitous pattern. I noticed the same values flashing by over and over ---> AC!.

Looks like one of the bulletproof diodes failed eh? What do I replace those SS-5 and SS-5R?
 
Last edited:
I can't find any info on these now, although I thought I had looked them up in the past with success. I am guessing they will be 5A, 200V or so.

I think it is obvious you will have to replace them with 'single' axial replacements (replace the other half of the bridge as well).

I'll have a dig around and see what I can find. <digging noises> :smoke:

This one http://uk.mouser.com/ProductDetail/STMicroelectronics/STTH4L06/?qs=WSSg0AwdCh5xkbKry%2bciwg==

10A, 600V - overkill , but it will do the job, and they are not too big physically. (they are also ultra-fast recovery - this 'choice' was forced.)

Stand these replacements up on end, one end into the PCB, and solder them to a stout piece of copper wire formed into a 3 sided 'loop' each end soldered to the PCB, to replace the mounting strap of the old rectifiers - that should do it. ;)
 
Last edited:
Yes your are right, this overkill is for the better, when current goes up the forward voltage raises less fast, they run cooler and having 10000uF or more, 10 amp diodes is less overkill than it looks. In the end, an original one apparently failed.
 
Thanks gents. I'll report back after I rebuild order the diodes and rebuild the bridge.
 
I rebuilt both parts of the diode bridge. Result: Exact same damn problem!!

I posted a video showing how I connected the scope to the filter caps and the display on the scope. When I measure with my multimeter I still show no AC on the left cap but circuitous measurements on the right side cap. Voltages measure the same as what I have in the picture a few posts earlier.

Here is the video.

https://www.youtube.com/watch?v=Ug_yQbEJnQs&feature=youtu.be



To clarify - the scope is connected to the neg terminal on the left side cap and the pos terminal on the right side cap.
 
Last edited:
Im confused is the " A" trace the positive rail and the " B " trace the negative rail with the scope ground on the common rail on the buss bar ( common ) between the 2 caps?

Next step would be to remove the load… but I'm not 100% sure your measuring the correct places...
 
To answer your question; no I had the ground connected to the chassis. I retook the measurement off of the capacitors with the ground on the common buss terminals between the two capacitors and I get proper +39 and -39 V. The spec is 41 V. So this appears to be the correct voltage with the proper connections.

Still no relay engagement however.
 
AH HAH well it's logical to assume that something went POOF down the line due to a FUNK in the power supply .. I would measure the amp output PRE " protect relay" and see what DC offset you have.. small or JUMBO….
Mike
 
Amp output pre relay is:

17 volts DC!!!! Meter negative lead connected to speaker output neg and the positive lead connected to the point on the relay where the amp signal is incoming.
 
Last edited:
Back
Top Bottom