AU-20000 Power-Up

Rotoflex4

Active Member
Good Evening,

I'm just about ready to power up my AU-20000 after replacing the output transistors and fixing up the F-2511 (Protector Circuit Board), the F-2508 (Power Supply) and the F-2507's (driver boards). I made a dim bulb tester so I'll be using that. The previous owner told me that the amp hasn't worked in over 20 years, following a couple unsuccessful attempts to fix it (was blowing the main fuse). I'd welcome some advice on a couple items, if someone out there has related experience.

1) Is C07 on F-2511 (1uf 50v) bipolar? I think it is....that's what I've put in there. I'm seeing/reading conflicting info on that and just want to be as sure as possible.

2) Is it a reasonable approach to leave the F-2508 and F-2507 boards out for the initial power-up? Or will that cause unintended problems? I was wondering about focusing on the F-2511 board, making sure everything's OK there before putting the other boards in.

I put extra fuses in the recommended places and will go with lower amp fuses all around. I've put a lot of work into this and want to be as cautious as possible. That's about it. Thanks.
 
Bi-polar are always ok for small value caps, yes to leaving boards out, dbt bulb without driver boards 100w max, shift to higher bulb once you get initial tests complete, good luck.
 
Bi-polar are always ok for small value caps, yes to leaving boards out, dbt bulb without driver boards 100w max, shift to higher bulb once you get initial tests complete, good luck.
Excellent advice. I didn't know that about the small value caps. Thanks.
 
Depends how 'low' you have gone with the rating of the fuses, but just be aware you might get 'nuisance blowing' of those low rated fuses.
Thanks John. Good point. I figured I'd start low and up the values as necessary. If I can get through the evening with no smoke, I'll be happy.
 
So......I powered up the AU-20000 without the driver boards (F-2507) but with everything else installed. DBT (100w bulb) gets bright for a sec or two, then goes very dim. Amp meter lights on. Red power light on. Relay clicks. The meters twitch a few times on startup then settle down. No apparent problems. I measure 57 volts on all 4 lines going to the output transistor boards.

Next.....I installed the driver boards. With these installed the DBT bulb pulsates. The amp meter needles twitch in rhythm with the pulsing bulb. No relay click. If I try to take the same voltage measurements, the DMM readings jump all over the place. No smoke or any other obvious problems. Nothing seems to be getting hot.

I read one thread on AK which suggested that a pulsing bulb was not necessarily an indicator of a problem. Just not enough juice to get things going. I'm tempted to just set the DBT aside and try plugging it right into the wall outlet.

Any suggestions before taking that step?
 
Yes, Make sure the bias controls pots are turned so that bias is at a MINIMUM.

Note that this may not be the same as bias pot resistance at a minimum.

Then try it again on DBT.
 
Last edited:
The manual says to turn the pots counterclockwise to achieve minimum, so I'll do that. They are new Bourns multiturn pieces...presumably this would apply with them as well...
 
Well.....I just blew an output transistor.......I think I'll pour myself a drink.
 
Last edited:
In this particular circuit, bias minimum is achieved with the trimmer at maximum resistance.
 
Last edited:
In this particular circuit, bias minimum is achieved with the trimmer at maximum resistance.
Thanks John. I appreciate your wisdom. That info will come in handy for my next attempt. I think I'll spend the next couple weeks going over all the boards that I've worked on, replace a few more transistors, hopefully stumble upon the cause of the shorts. I've recently read about soldering wires in the pass-throughs on the double-sided heat sink boards. I'll do that while I'm at it.
 
There are several Sansui boards that are counter-intuitive on the trimmer settings. Also the manuals often have errors, so a best practice is to ---
1, get unit on dbt for initial startup
2, attach two dmm's to test points, one for offset and one for bias voltage, set to the correct measuring range
3. preset both trimmers to their midpoint if unsure of correct rotation direction
4. only hook up one output board if possible
5. while powering on watch for rapid voltage rise, if not too far above factory limit adjust trimmer to bring it back down noting direction of turn
6. If relay does not click that usually indicates offset is incorrect. Stop, remove output transistors and fix offset issue first.
7. once offset is correct re-install outputs and remeasure and adjust bias
8. disconnect the first channel, hook up the other channel and repeat
9. Hook up both channels, check offset and reduce bias to 50% of factory value
0. apply mains power and bring the amp to 80% with centered offset and allow to warm up for ten minutes then bring bias to 100% and monitor till you are sure the readings are steady.
Most Sansui amp typologies pre dc servo designs have a fail safe resistor in parellel with the bias trimmer section of the anmp to slow the onset of catastrophic output meltdown due to over bias allowing this method to work well, the later designs are not forgiving at all.

-Lee
 
There are several Sansui boards that are counter-intuitive on the trimmer settings. Also the manuals often have errors, so a best practice is to ---
1, get unit on dbt for initial startup
2, attach two dmm's to test points, one for offset and one for bias voltage, set to the correct measuring range
3. preset both trimmers to their midpoint if unsure of correct rotation direction
4. only hook up one output board if possible
5. while powering on watch for rapid voltage rise, if not too far above factory limit adjust trimmer to bring it back down noting direction of turn
6. If relay does not click that usually indicates offset is incorrect. Stop, remove output transistors and fix offset issue first.
7. once offset is correct re-install outputs and remeasure and adjust bias
8. disconnect the first channel, hook up the other channel and repeat
9. Hook up both channels, check offset and reduce bias to 50% of factory value
0. apply mains power and bring the amp to 80% with centered offset and allow to warm up for ten minutes then bring bias to 100% and monitor till you are sure the readings are steady.
Most Sansui amp typologies pre dc servo designs have a fail safe resistor in parellel with the bias trimmer section of the anmp to slow the onset of catastrophic output meltdown due to over bias allowing this method to work well, the later designs are not forgiving at all.

-Lee

Wow......that info is like gold. Thank you very much. I will refer to it closely on my next startup attempt.
 
I just pulled the two zener diodes off my F-2508 (power supply) board. They look like they have been replaced some time in the past. I have a feeling that they aren't zeners at all. I think they are just regular diodes. I tested them with the diode function of my DMM. I also tested them with 2 x 9 volt batteries in series with a 300ohm resistor, looking for 15v across the diode. Instead of 15v I get .85 volts. This is the same reading I get when (for comparison) I test a regular diode in this fashion.

If they are in fact not zeners then they are likely a remnant of some past repair attempt. My limited understanding is that they play a role in regulating voltage. So....I'm wondering if this may have contributed to my blown output transistor???

A while back someone on AK had cautiously recommended 1N5245BTR as a replacement for the RD-15E zeners. There seemed to be a level of uncertainty however. Anyone know if this is still considered the best replacement? Thanks.
 
Those would work, however I routinely upgrade the 1/2w zeners to 1 watt variants so---

512-1N4744ATR

mouser part number for a Fairchild 1 watt 15v zener diode
 
Well.....I replaced my AU-20000's blown output transistors and soldered thru-wires on the double-sided output boards (F-2497). I'm pretty sure these boards are OK (they are pretty simple and I triple-checked everything I could).

So.......now.......using the dim bulb tester, things power up nicely with the two driver boards (F-2507) removed. Both relays click, nothing smoking, bulb goes dim. When I install the driver boards (either individually or together), I get a bright bulb (this is basically how I damaged my output transistors last time I powered up a couple weeks ago). No damage done today and no fuses blown, but I'm a bit stumped as to my next trouble-shooting step.

If anyone has any tips, that would be wonderful. I'd like to try everything possible before taking it to a tech guy (the nearest reputable shop is about 250 miles away......about 10 hours driving for the round trip....then another day of driving to pick it up when it's done).

Thanks.
 
My last question was a little vague, so I'll try to get more specific. I'm trying to put my detective cap on here and think things through to the best my limited ability. I took voltage readings in various places (using a DBT tester and with the driver boards removed). I found some voltage readings which seem a little off (please see pic). Would this point to a problem with the main rectifier diodes (SG-5TS, SG-5TR) on the protector circuit board (F-2511)? I tested them when rebuilding the board and thought they were OK, but I may have been wrong. I measure 52v ac on both terminals of both of these diodes. I measure +71v dc, -71v dc directly at the terminals of the main filter caps.
 

Attachments

  • diagram.jpg
    diagram.jpg
    96.7 KB · Views: 27
I'm no expert.
I'm guessing no load would make those high (like you observed) in circuit I'd expect them to read about what's on the drawing.
Another thought, line voltage has crept up over the years. I might've been under 120v when unit was built. You could be seeing 123-125 today. That also might account for your difference.
Wait for some more experienced counsel
 
Thanks for your input. Although all the numbers seem to be a little off, the main thing I'm wondering is how come I'm getting Ov where I think I ought to be getting +68v? Maybe it's normal and I just don't understand things, but it seems wrong.
 
Back
Top Bottom