I am looking to replace the 4 power transistors in a 1970’s Kenwood KA-6004. There are two which are NPN and two PNP. The part numbers in the schematic and on the transistors are:
2SA744 and 2SC1402.
One chat room has suggested the MJ21193G and MJ21194G as replacements, but I am hesitant to just accept the recommendations of these sites not being certain if due care has been taken to fully vette the items.
I apologize for asking for your time to do this, but I have not been able to locate any cross-reference for these transistors.
The main reason I am looking to replace them is the strange voltages I get, and, ultimately, the resistance readings as follows; Q3’s E-C and B-C really concern me:
Taken between the Emitter, Base and Collector - (reversing the leads on the PNPs) and when I reversed the normal leads on each transistor only Q3 showed a reading between Base and Collector of 213k, all others showed overload. Q1/Q2 = npn, Q3/Q4 = pnp
Q1 E-B 2.215M E-C O.L. B-C 2.037M
Q3 E-B 2.670M E-C 2.74M B-C 98.5k
Q2 E-B 2.289M E-C O.L. B-C 2.118M
Q4 E-B 2.750M E-C O.L. B-C 2.107M
I want to replace the originals with something which has comparable, or better specs, and something which should last like these have.
Thank you.
I’d advise against fixing such things unless they’re definitively broken. Discrete transistors don’t have much for wear mechanisms under normal conditions, and unless one’s familiar with the adjustment procedures common to such equipment there’s a fair chance of ending up with less than what one started with.
As for resistance measurements on a transistor, they don’t mean much. One has little to no idea what sort of test voltage the meter applies, and the nonlinear I-V characteristic of a PN junction means that small changes in applied voltage translate into large differences in resulting current flow. One can use the diode test function of a typical meter to check whether the junctions are intact, but more sophisticated tools are needed to do much beyond that.
As for those suggested replacements, they’re similar in many respects, though by no means the same thing. Could they work? Maybe. Component tolerances of the era being what they were, workable designs necessarily were tolerant of a fair degree of variation. But again, unless one’s familiar with the applicable adjustment procedures, swapping transistors (even with the original part number) might well leave a person with a device that performs more poorly than the one a person started with.
Be advised that the realm of “audio” attracts an inordinate proportion of, well, questionable theories… Don’t take things with a grain of salt, bring the entire shaker. I’d suggest Douglas Self’s books on the topic as worthwhile resources for study.