Generally going down in capacitance won't cause you issues, but going up can potentially cause headaches, especially if there are a bunch of paralleled outputs. The higher the capacitance, the harder it is to switch the transistor on/off. The effect gets worse as the power level goes up and the...
The speeds in the datasheet are speced at vastly different currents. The KSC2383 is 100Mhz at 200ma. The 2SD666 is 140Mhz at 10ma. It looks like the circuit is running these at about 15ma from the schematic (hot!). At 15ma the 2SD666 is 160Mhz, but the KSC2383 is somewhere down at around 60Mhz...
Given the position of those 2SB646a/2SD666a in the VAS stage and the frequency response of the amp, the KSC2383Y/KSA1013Y might be a sub-optimal choice. They might work, but the KSC3503D/KSA1381E is a much closer match to the original parts. The KSC2383Y/KSA1013Y have significantly lower...
I don't remember the conversation 100%, but I seem to recall talking to the designer at Redgum at Axpona years ago and he mentioned one of the other brands would oscillate when used in place of the Exicon's he would normally use. He was of the opinion there was some difference between them in...
Depending on the device I get a JFET or D(epletion) MOS. I guess its possible the software in your version is doing something screwy. If you test resistance between the outer leads, you should show around 30-50 ohms. What amp is this?
I can't really tell from the picture, but the K170 might be fake. There is the LSK170/LSJ74 that is a direct substitute. As for NOS, there is a seller on ebay "punkydawgs" that sells real NOS stuff (no affiliation). There is still some 2SK163/J44 out there that is real. There are some other SMTs...
IIRC, those are the current limiters. You will likely want to use something with more current handling, like the C945/A733, C1815/A1015, BC636-640 (pinouts are reversed), or A1013/C2383 if you really want to go overboard.
Its probably not worth it replacing the outputs in the good channel. I would spend the extra $20-$30 replacing all of them in the bad channel. With 83v rails, the surviving outputs probably were under a fair bit of duress until power off. You can't be sure they are 100% still. You might use the...
I've seen Carver's with hodgepodges of Onsemi (then Motorola) and Toshiba 1302/3281. You might want to go all On-Semi for the larger SOA.
I would think a MJE part would work for the 2SC2562, given it looks to be just a bias transistor. Kind of an odd part in an amp.
The PCN's on OnSemi's site show what happened. The Fairchild/Samsung parts diffusion was moved to Japan (maybe Sanyo's lines?). The packing shifted to China a while ago, but On-Semi changed suppliers within China - I think it was AUK before. Unfortunately, the new JCET packaging has changed to...
Mine are both different. They may be KMA-200s.There is some voltages listed on the 202B schematic to compare against and a note about checking for bad mosfets. Also make sure that the parts of the PCB that burned up are cleaned up sufficiently so they aren't conductive. You can have shorts...
If the trace is only on one side then you are good. It's the situation where there's traces on both side and the through plating is broken that can cause headaches if only 1 side is solderable.
If you want to shotgun it:
I'd suggest the MJE15034/MJE15035 instead of the 32/33 for the A968/C2238...
This is just a 2 layer board. You may want to invest in a desoldering pump like a Hakko FR-301, it'll make your life a lot simpler and remove some stress on the traces, especially if you get a larger nozzle tip for the TO-3s.
If you've confirmed all the through-holes are indeed intact and...