clinic-audio
all on YAMAHA untill 1990
Why don't you use original transistors ?
but not everyone is as fanatical as I am
Pull the output transistors and test them for being shorted. Then you'll need to start a thread of your own in the Sansui forum to pin down the issue(s) causing the 26v. There is no speaker protection relay in the 3900Z. There are many Sansui "experts" in that forum to assist.Thank you so much. I have only tested for voltage at this point. I've got 26V DC coming out of my speaker terminals and I'm trying to pin it down.
The KSA1220AY (PNP) and KSC2690AY (NPN) are virtually perfect substitutions. Both are TO-126 packages.Depending on the circuit, the faster ft may not be an issue at all.According to Motorola the 2SB631 replacement is the 2N4920 and the 2SD600 the 2N4923.

They just made a substitute at the time of production sometime after the schematic was published. You'll find a LOT of brands/models have many in-production changes that are undocumented. It's common. Always go by what is in the stock unit and make substitutions/replacements from there. The (0-1) after the transistor number is the gain group they wanted to use for that particular transistor. Different transistors and makers will use differing numbers, or letters to designate their gain groups. The KSC1845FTA is the preferred substitute in this case. The (F) is the gain group, and the (TA) means TApe packaging. You should gain match the input pair to help equalize DC Offset.The SM shows 2SC1000 (0-1), the actual parts on board are C1681; and I assume it is better to use substitute for what is currently installed on board? And that would be KSC1845? I am confused.
Not to open a can of worms here.... Ideally, you want to match both hFE and Vbe. If your goal is to minimize DC offset, then I think Vbe will be more important. hFE will affect offset even with perfectly matched Vbe's, but it is a secondary effect created by the base currents in the base resistors.You should gain match the input pair to help equalize DC Offset.