Sansui 9900A problem and solution

[fernarias]

So I purchased this mint 9900A about 6 years ago from the original owner. The reason for the sale was that the unit was not operational and I got it at a very good price.

After doing some studying, I went into the power supply unit and quickly found the problem, a roasted 100uF capacitor that fell apart in my hands. I replace all the capacitors on that board, since it was already out and got it to function (played thru headphones and speakers and all inputs worked). Happy with the result, I then proceeded to change all the capacitors, replaced the fusistors and replaced the "uniwatt" transistors(as recommended here on the forums). Needless to say it sounded great and it was on my main stereo system (with Cornwall IIs) for six months. I then put it away until yesterday.

 

[fernarias]

When I had played it before I noticed that it got hot but knew that it was the power supply board(I thought it was strange that there are no vents) and not the amplifier and when I check Audiokarma, all the post talked about how hot the 9900A got. I always assumed it was because the power supply was more robust than the 9900, that I also had and didn't think anything about it. So yesterday, when I turned it on again, it got hot just like I remembered and I played a few songs (that sounded very nice) and then proceeded to take a shower. I forgot to turn it off and noticed that it was very hot and it had gone into protection while on. I turned it off and I got hot and a little sweaty.

 

[fernarias]

While waiting for it to cool, I again read some AudioKarma posts but since the 9900A are rare, I didn't find anything useful. After it cooled I turned it on and got the green light however I wasn't satisfied. It's been a long time since I started this journey and I have a lot more equipment to test my gear and also the components inside so the 9900A went back on the bench.

 

[fernarias]

I just completed a Sansui BA-3000 and had already put together a list of transistor and rectifiers compatibles to replace what is no longer available and that Sansui had used (2SA733 is still available as KSA733 but everything else is not). Luckily, I had more than enough parts and proceeded to replace all the transistors and rectifiers on the power supply board (making sure to check twice, thrice and four times since most of the transistors are cbe and I was replacing with ecb). And of course more cleaning and adding fresh thermal compound. After finishing the work I also noticed that the factory had switch the right and left channel on the headphone jack (yes, the speakers were correct) and proceeded to fix the problem.
The unit came on and it went out of protection and it sounded beautiful as I remembered. Only now, the power supply wasn't hot. So success.

 

[fernarias]

So what was the problem. I checked the transistors first, one by one. To my surprise I got good readings although there were a few that had Gain that were 50% or less than the other transistors of the same type but I knew that that wasn't the problem as I have had amps play with Gain that was below specs. Then I started to check the diodes and rectifiers. I got half way through and then one of the rectifiers (a 10D2 at the D601 to D604 position) was shorted. Yup, shorted, this 9900A still played, still sounded good with a shorted rectifier that provides power to the preamp. The rest of the circuit was so well designed that it was able to filter out half a ac wave (although it got hot). Pretty amazing.

 

[markkb]

Cool, thanks for posting. Good write-up.

 

[LBPete]

Glad you found it.

Please post the serial number for the database. Click on the link in my signature block.

-Pete

 

[Overundr1]

Nice

 

[Hyperion]

Well done, not the first time, or the last, that one of those 10D1's has been/will be, the reason for a subtle fault. So good to find the cause in such cases

 

[jheu02]

I guess no more assuming the rectifiers are good if the unit powers up and they don't look stressed.

 

[fernarias]

I guess no more assuming the rectifiers are good if the unit powers up and they don't look stressed.

Well, remember I said that the power supply was running hot. That's a good warning sign but every thread that I read about the 9900A said that it ran hot so I considered it normal. Note, it is warmer than the 9900 but it's not hot (hand feels uncomfortable on the top).
I'm still amazed it ran (and it sounded nice) and it ran for 6 months like this.

 

[fernarias]

Glad you found it.

Please post the serial number for the database. Click on the link in my signature block.

-Pete

Sorry, it sold the day after I fixed it (day I posted the thread). Didn't get to listen to it except to verify that everything worked and didn't get to take any pictures (and it was mint).