KR-9600 DC offset

JacTaylor

On the lam
I measured the voltage of the left and right speaker at 170 mv for the left and ~70 for the right. At $.89/each for ZTX694b, I didn't spring for a 100 to find two matched pairs. Instead, I purchased a 100 of the KSC1845. As I was replacing Qe1 and Qe2, I found the hfe measurement of the original Qe2 was half of Qe1 for the right and about 1/5 for the left. Definitely, not matched pairs. After replacing matched Qe1, 2, and Qe3, 6,7 and 8 as suggested by Hopjohn along with caps, I had a measurement of and ~3 mv for the left and 16.8 mv for the right. I checked two or three times after running it for 30-40 min.

Okay, silly question for my first post.
This morning after running it with the FM radio for about an hour, the left was similar at 3.9-4.1 mv but the right was now steady around 7.4 mv. Not that I'm complaining (actually was delighted with 16.8 mv), but is this decrease due to just the longer warm up or is there some kind of magical break in of the new parts?20180327_200805.jpg
 
The original 2sc1345 D or E were probably slated for around 500 hFE. KSC1845F (the higher gain versions E and U have gone bye-bye) are typically around the 450 hFE range, the ZTX694B can easily be found above 800, even above 1000 hFE. That's what you're paying for. The 9600 is worth the extra expense IMO. At Qty 10 they are only .76 cents, .58 cents at Qty 100. .A small investment for some really nice, low noise, high gain devices for anyone doing this regularly, particularly when you consider the rate at which these types of devices are disappearing.
 
Thanks for the confirmation. Actually, I followed your lead on the KSC1845s! Very happy with the new offset readings. I use Sonos and was just streaming Google music and thought they were playing a really low bit rate. They could have been playing inferior quality recorded music but those high offset readings sure weren't helping matters.

As someone new at this, I was sweating bullets about putting those darn transistors in the correct manner. I developed a mnemonic that worked great for me. Testing the transistor leads with the DMM on diode, the NPN were " B&E, (breaking and entering)" for the black lead on the emitter and the positive lead on the base, "R&B", to get a reading of ~680. For the PNP, "Base on balls" for the black lead on the base with the positive on the emitter to get a reading. Still used the DBT, just in case. Took me awhile to figure out the line in the circle, base, and the short line outside the circle as emitter code on the pcb.
 
The original 2sc1345 D or E were probably slated for around 500 hFE. KSC1845F (the higher gain versions E and U have gone bye-bye) are typically around the 450 hFE range, the ZTX694B can easily be found above 800, even above 1000 hFE. That's what you're paying for. The 9600 is worth the extra expense IMO. At Qty 10 they are only .76 cents, .58 cents at Qty 100. .A small investment for some really nice, low noise, high gain devices for anyone doing this regularly, particularly when you consider the rate at which these types of devices are disappearing.
You make a good point. With more confidence in my skills, I probably would have shelled out the 60 bucks. Interesting and disappointing how many of the good parts are disappearing. Out of my 100 KSC1845, I found two around 560 and two matched around 520. The original Qe1/Qe2 were matched as 420/204 and 442/85. At this early stage, I'm just glad it is still running with an improved offset.
 
Back
Top Bottom