AU-9900 - cannot get a bias adjusted

ToniCH

Well-Known Member
Hi guys,

Just bought a AU-9900 today. Before testing I checked DC and Bias current.

On one channel DC was ok and on another not ok. So, I adjusted the DC per service manual and now both are ok...ish... One channel is steady and another jumps around about +-10mV but stays quite good within specs.

Next Bias: well, here is the problem. Manual tells me to measure mA and adjust to 50mA. I am getting something like 0.3mA and neither channels reacts to adjustment.

What is the most probable problem and where should I start troubleshooting? Thanks for any suggestions.

BTW. I briefly tested it with junk speakers to see if it works at all and yes, it plays music. I didn't feel anything getting very hot while it was running either.
 
That really is very low - my one was stuck at 25mA and similarly did not respond to adjustment. Mine too, actually played music and sounded ok-ish but was devoid of almost all 'detail' and the imaging was hopeless.

I ended up changing all the driver and pre-driver transistors - you know those grey or green transistors on the driver board. They do not age well and lose their gain dramatically - one of mine had a single figure hfe of '5' if I recall correctly.

Do check that the driver board and the OP transistors are getting power, check that the OP transistors are not faulty - and that the bias trimmer is working, if that checks out it could well be those transistors...

EDIT: And your meter fuse !!! - of course - nice one skippy124 :)

The method of adjustment used here (as per SM) could cause meter fuse failure quite easily.
 
Last edited:
Not sure how difficult it is to get to the output transistor emitter resistors - if they are accessible you could measure across one of these on each channel - and set for approx 6-7mv. This equates to approx 20ma bias current for each output pair - or 40 ma in total, the current measurement method of 50ma includes the driver circuit current as well

Cheers

John

Edit, be careful not to short anything if attempting this measurement
 
Thanks, didn't think of that as the meter seemed to work. But, indeed one of the 2 fuses in my Fluke is indeed blown. I'll get a new one and try again tomorrow. :beerchug:

Good to hear that that is the problem and not something wrong with the amp!

Cheers

John
 
Well, that remains to be seen but looks promising now. :) Lets hope I don't have to return to this issue anymore, thanks Skippy and Hyperion for suggestions!

Its very clean inside and looks like its partially recapped. Brand new looking Nichicon Gold tunes as filter caps for example and some other new stuff here and there.

I plan to recap and renew transistors etc. if necessary when I get some other projects done some time later this year.

2cmtba1.jpg
 
Last edited:
I had my Fluke 8060 gnd pin start to come loose on the internal PCB.. and gave me ALL sorts weird intermittant hair pullin PILES O grief...
 
Had this problem 2 years ago with a "boatanchor" Sansui AU-X1 amp. It was due to shorted VBE multiplier transistors (the ones that set the bias, acting as variable resistors).
 
I'll start by changing the same 8 transistors on my F-2580 board as others have. Hope that sorts it out and then I can continue with general recap and also swap the troublesome transistors etc. elsewhere.

There is some small diodes on the F-2580 board which look just like VD1212s but are different color. I think the sm doesn't name them but only says "diodes". Are they something I should swap out by default or generally ok?
 
I think they are SV-02's - I changed mine for the usual series pair of 1N4148's appropriate for dual junction diode substitution. I also changed the DS430's as well (I think ?), I left the 1S953's alone.
 
Board: F-2580

So, I today changed all el. caps, 12 transistors + couple of resistors, which had glue corrosion. I tested all parts when I took them out and they tested ok, infact I didn't find any broken ones. Even the original ELNA caps were still dead on.

I also measured all resistors on board and failed to find any broken ones. Tested the diodes which say SV-02, they are ok. I also tested one of the 1K Ohm bias trimmers and it moved ok between 125-850 Ohm (is this enough adjustment?).

Then a test run: sadly no change whatsoever. :( But, no smoke either.

I still have 4 can looking transistors and 1S953 diodes to test and DS430 (?) diodes to change but if that doesn't do it I guess I need to look at the ps boards next.

BTW. do you guys have several jumper wires under the board? Wondering if they are factory?
 
Board: F-2580

So, I today changed all el. caps, 12 transistors + couple of resistors, which had glue corrosion. I tested all parts when I took them out and they tested ok, infact I didn't find any broken ones. Even the original ELNA caps were still dead on.

I also measured all resistors on board and failed to find any broken ones. Tested the diodes which say SV-02, they are ok. I also tested one of the 1K Ohm bias trimmers and it moved ok between 125-850 Ohm (is this enough adjustment?).

Then a test run: sadly no change whatsoever. :( But, no smoke either.

I still have 4 can looking transistors and 1S953 diodes to test and DS430 (?) diodes to change but if that doesn't do it I guess I need to look at the ps boards next.

BTW. do you guys have several jumper wires under the board? Wondering if they are factory?

All black wire? - then yes, the jumper wires are factory. (take a good quality pic of these for your records :yes:)

How are the bias transistors TR09 & TR12 = 2SC984 - test good ?
 
Last edited:
are the output transistor supply voltages correct ?
I not at site now so I'll check tomorrow.

Hyperion said:
All black wire? - then yes, the jumper wires are factory. (take a good quality pic of these for your records )

How are the bias transistors TR09 & TR12 = 2SC984 - test good ?
Yes, all black - good.

You mean TR09 & TR10? I changed them but yes, they tested ok.

I changed all other transistors but TR11-TR13 ie. 2SC735 & 2SA562. If these are in any way involved with with bias functions then my problem might lie there?

EDIT: Just realized I mispoke: I didn't change the transistors which are on the heat sink. They tested ok.
 
Last edited:
So just so I'm not being dim about this… the amp works, correct?( Post all the work ) so I assume it draws SOME current, is it just you cannot adjust it at all or your out of range? AND you get a DC swing as you move the bias pot ?
 
So just so I'm not being dim about this… the amp works, correct?( Post all the work ) so I assume it draws SOME current, is it just you cannot adjust it at all or your out of range? AND you get a DC swing as you move the bias pot ?

Yes, the amp seemed to work ie. it played music (didn't try it today after all the work was done). Didn't listen if it sounded good or "lifeless" like others report when their AU-9900 ignores all attempts to adjust bias. From the forum I see this seems to be quite common problem with the 9900s. Usually successful fix has been changing the 8 known problematic transistors on the F-2580 board. And that is what I tried today. I swapped 4 more transistors (TR01-TR04) because they were not the same on L and R channels ie. changed before on one channel with different type, replaced them all with original type.

Very little happens when I turn the bias pots. My bias is almost zero (should be 50mA measured the way its on the manual). Before today meter indicated something like -3mA - -16mA turning pot from one end to another. Today I cannot remember but I think change was even less. Same thing on both channels. Power supply maybe?

I haven't noticed bias adjusting effect DC-reading. But I also didn't measure it at the same time but every time I checked DC it still was in spec.

Pic: what I renewed this far.
 

Attachments

  • F-2580_board_this_far.jpg
    F-2580_board_this_far.jpg
    56 KB · Views: 50
Last edited:
THe Bi polar 56 volts is good coming up to the board?
When twisting bias you see it's effect on the base of TR-15/17?
R-27 / 29 ( 120 ohm ) are good?
 
Back
Top Bottom