No double sided boards in the 9900. I believe the AU-20000 has them, though.I haven’t been inside an AU-9900 but if it has the double sided PCBs I’d recommend soldering a wire through the pass throughs.
The plated holes tend to deteriorate over time.

My choice for Semiconductors would be the following for F2568:
TR02 - MJE15031G or even better MJE15033G
TR01 - MJE15030G or even better MJE15032G
TR04 - KSA1013YTA or (if you do not mind twisting pins) better ZTX796A (Pinout is EBC !!!)
TR03 - KCS2383YTA or (if you do not mind twisting pins) better ZTX696B (Pinout is EBC !!!)
TR06 - can use KSA992FB or (if you do not mind twisting pins) better ZTX796A (Pinout is EBC !!!)
TR05 - can use KSC1845F or (if you do not mind twisting pins) better ZTX696B (Pinout is EBC !!!)
ZD03 - Vishay BZX55B10-TR 2% zener.
If i understand correctly then ZD01 and ZD02 are 39V 0.5W zeners then use Vishay BZX55B39-TAP 2%.
Now regarding C15 which shunts ZD03 I would use here Hybrid polymer capacitor. As Zenner itself has very low internal resistance I would use here if space permits United Chemi-Con part# HHSE630ELL101MJC5S
as those have very low ESR.
Normally I would say bump up C03 and C04 to 220uF or even 270uF to improve rejection rate however here is ZD01 and ZD02 solving it and in reality the regulator would venture very well even without those caps or should i say loss of capacitance here would not be critical when zener is in place, for the sake of smoothening and shunting and filtering noise I would again use same United Chemi-Con part# HHSE630ELL101MJC5S in place of C04 and C03.
C02, C01 I would bump up to 680 X 80 V or even to 820 X 80 if space permits.
C12,C11 I would bump up to 220 x 63V
Just my 2 cents.
EDIT I made a mistake. I somehow thought that ZD03 is 13V zener.
ZD03 should be BZX55B10-TR
Well that is true that KSC1845F can in many cases replace 2SC1313. However it is not really always 1 to 1 replacement. Here it is part of the darlington transistor and I should say in this particular case 1845F will be OK in terms of collector currents and power dissipation. But I would feel more confident (in terms of collector current and max power dissipation) with ZTX696 or KSC2383Y also with later the rejection rate might slightly drop because hfe of 1313 is almost twice higher. for TR06 and TR05 (error sensing transistors which drive Darlington transistor TR01,TR03 and TR02,TR04) KSA992/KSC1845 are OK as those collector currents will be at the bay no matter what, but again me personally would use here ZTX.Thanks for the suggestions, though not looking to replace everything, only parts that are known to be problematic, which is what I listed on my BOM for F-2568. 2SA726 are known to be bad transistors, and I have found the same in the receivers I've had, so makes sense to replace them. Standard recommendation for replacement is KSA992. Just wondering why you recommend something different for TR04? Standard replacement for 2SC1313 is KSC1845, why something different for TR03?
Well that is true that KSC1845F can in many cases replace 2SC1313. However it is not really always 1 to 1 replacement. Here it is part of the darlington transistor and I should say in this particular case 1845F will be OK in terms of collector currents and power dissipation. But I would feel more confident (in terms of collector current and max power dissipation) with ZTX696 or KSC2383Y also with later the rejection rate might slightly drop because hfe of 1313 is almost twice higher. for TR06 and TR05 (error sensing transistors which drive Darlington transistor TR01,TR03 and TR02,TR04) KSA992/KSC1845 are OK as those collector currents will be at the bay no matter what, but again me personally would use here ZTX.
Yes MJE150xx are overkill for this task but I would say "hot" overkill. and yes KSA1013Y/KSC2383Y ( do not use "O" version as hfe even lower) or ZTX796/ZTX696 overkill as well but again in a good way. In the end remember it is the Power supply of your amp and it should be bullet proof. That is just how i see it.
Oh, yes, you are exactly right, and the amp is currently working and has not had issues as long as I have owned it. However, it's easiest for me to start with the SM for a list of components, then check the unit and make corrections based on what is in there. If I start with the schematic, I fear I will miss something. In either case, need to replace what is actually in there, as you say. Totally agree. That's why I'm taking the time with this unit to check every board against the SM list. Some boards are matching 100% with SM, others it's clear the SM is in error. So, I think I'm following your directive...Regarding those capacitors - you are forgetting the recapping mantra.....
Replace parts so they are the same as what you find in your unit - the unit worked once and probably left the factory like that. The schematic and the parts lists could be anything, and never had to function as a working unit.
![]()
I will check the 10 µF films. I’ve used them before in the Eight and did not have a problem with space.
Any thoughts on C25/26? I’ve not seen caps that look like that before, so am not familiar with what type they are.
Will take a look at options for the filter caps. There are a couple of threads on filter caps for the 9900.
Now that you say that, I meant C23/24, which are shown in the second picture in post #25. I assume those are not electrolytics. I'm just not familiar with them.Hmm I do not see C25/26 on F2570. Are we talking about different board? I see C01 C02 is 0.47uF which probably could be upgraded to polypropylene (need to see size)
Just enter the type number into the AK 'search' box - I just tried this and got 96 hits where people have encountered 2SC1364 and done something about it.What would be the recommended replacement here?