The plan - make the nicest AU-9500 I can!

w1jim

I can fix it but good...
Subscriber
It's Sansui integrated amp time around here.
Earlier in the month I restored and upgraded a sweet AU-999 - all the great mods from the AK Sansui Gurus.

http://audiokarma.org/forums/index.php?threads/knock-knock-its-my-ntm-au-999.701143/

Then, at our New England get-together this past weekend there was a pretty AU-9500 that was calling out to me - another beauty for the stable.

I also picked up a nice TU-517 tuner that I recapped and will tie into what ever system I end up with here.

All the parts are in - here's my game plan for the AU-9500:

Complete recap - including doubling the size of the main power supply caps.
Rebuild the phono preamp board with 1% metal film caps. Replacing the carbon 10% resistors should make for a quieter and more accurate phono stage.
Refresh the thermal grease on the output transistors.

Here's a common problem - the lens (or jewel) at the power indicator light is gone.
I was thinking of enlarging the hole and mounting an appropriate LED - but once you drill the faceplate, no matter how nice it looks it's been butchered.
Solution is a 1/8" light pipe lens from Mouser.

http://www.mouser.com/ProductDetail/Bivar/PLP1-625/?qs=sGAEpiMZZMuu2qEGwSNRoI7NaoVtmgDBGQrRHsadHNQ=

When I removed the faceplate I found the back part of the original lens with the mounting clip - which I'll use with a dab of silicon RTV to secure it.

More pictures to follow.
 
New caps and especially resistors isnt what's the important thing here - you need to replace basically every single transistor in this unit - they all go bad, and I mean REALLY bad.
 
Oh joy - I did repair an AU-9500 for an AK'er a few years back that had a noisy / intermittent transistor.
I'll review the schematics and past threads for what's involved - at least these are very easy to work on.
 
...noisy / intermittent transistor. I'll review the schematics and past threads for what's involved - at least these are very easy to work on.

One of the main culprits is the early production (i.e., BCE pinout) 2SA726 transistor, the newer ECB 2SA726 transistor is often OK. Similar situation for the early production 21SC1313 transistors; the remaining transistors are often fine. Here's a link to an AK AU-9500 restoration thread that might prove useful: http://audiokarma.org/forums/index....-upgrade-of-scribby31s-sansui-au-9500.654341/
 
Thanks all for the good pointers to relevant threads.

Here are the pictures of the power on indicator fixed with the light pipe - no holes drilled, looks clean and simple.


image.jpeg

image.jpeg

image.jpeg
 
One of the main culprits is the early production (i.e., BCE pinout) 2SA726 transistor, the newer ECB 2SA726 transistor is often OK. Similar situation for the early production 21SC1313 transistors; the remaining transistors are often fine. Here's a link to an AK AU-9500 restoration thread that might prove useful: http://audiokarma.org/forums/index....-upgrade-of-scribby31s-sansui-au-9500.654341/

Lee - thank you for the advice, mine has the earlier versions.
Fortunately I have a deep stock of KSA992 and KSC1845 on hand to swap out those nasty troublemakers. They're only about 7 cents each when you buy them in 100 lot.
Are there any other known problem parts I should be especially wary of?
 
Amp driver boards completed.
My boards were different from the ones in Leestereo's thread - missing R880 and the series capacitor associated with it. Also no thermal fuse.
I replaced the 4-electrolytics, one resistor, the two 2SA726 transistors and put in a multi turn trim pot for VR801 and still need to replace VR803.

image.jpeg
 
Preamp/tone board completed.
The 2 tantalums at the input are replaced with film caps, all signal path caps replaced with Nichicon bipolar Muse's, 3 power supply cap values doubled and al 12 signal transistors replaced with modern Fairchilds (matched).

On to the big caps!


image.jpeg
 
Just about done - power supply all done with stock values, the 5 main caps have been doubled in size - despite their diminutive size.

Time to tweak the amp, do a little more cleaning and then on to the listening test.

image.jpeg
 
Last edited:
Ugh - annoyance time.

When I turn the volume up past 9:00 the unit goes into protection.
I've narrowed it down to the right channel.
When I swap driver boards it stays in the right channel.
When I crisscross the jumper it stays in the right channel so it's nothing before that.
Swap the speakers - no change.
Put in new outputs - no change.
New mica insulators - no change.
Tested the 0.47 ohm emitter resistors - they test good.
The heat sinks on the right channel also get hot.

I'm thinking perhaps the transistors go into oscillation or a problem on the protection board.

Thoughts?
 
Is there a Zobel network/Boucherot cell (resistor & capacitor) on the output?.... you could check them?

From all that you have done to find/isolate the problem, I am thinking it must be something about the right channel feed to the protection circuit? Often there are just a few components that are not common to the protection circuit proper, a wrong value capacitor here could cause problems.
 
The frustration continues.

I found that the right output transistors get hot even if no speaker is selected - that means I can use the temperature to determine if it's malfunctioning.

I disconnected the output of the right amp stage from where it exits the transistors (prior to the protection circuitry) and found that the Right outputs were still getting hot.
This means the protection circuit isn't the culprit.

I loosened the R heat sinks and didn't see anything amiss.
I also checked the small transistor clamped to one of the outputs - it's OK of course.

Now I suspect that the replacement output transistors are inappropriate.

The old transistors were a mismatch (not the same device in each location).
I replaced them with the Onsemi MJ15024G/25G.

Next test will be to swap transistors from left and right channels - when I return from a short business trip.
 
Ah, sweet success!
Sometimes it's good to walk away for a bit and come back with a fresh vision.
Ends up one of the driver boards had a weak 2SC1124 transistor - replaced that and everything is working great.
DC offset and bias current are both spot on.
Since I replaced the output transistors on one channel I decided to do the other channel just so it would match up.
Hopefully I can start with my listening tests tomorrow.
 
Comparison time.

I'm in the midst of comparing my AU-9500 with my AU-999.
Both are fully recapped and have some upgraded transistors and strengthened power supplies.
I used my A/B comparator that's discussed in this thread:

http://audiokarma.org/forums/index....-out-youre-invited.343405/page-3#post-4330096

I'm running them through my McIntosh XRT-20 speaker system (24 tweeters, 1-8" midrange and 2-12" woofers per channel).

Initial thought is that they sound very similar - they're both quite terrific.
But the more I listen the more I think that the AU-9500 is a bit stronger on the low end whereas the AU-999 tends to be a bit more detailed.
That being said, on some music I can't tell any difference and other music there is a discernible difference.

I need to listen more.

image.jpeg
 
Back
Top Bottom