Sansui 800 Equilizer board (F-1037A) transistor question.

[jd3rdcoast]

Hi all, new to this forum. Spent most time on Marantz but have recently acquired couple very clean 1968 Sansui receivers. An 800 & 3000A.

This question concerns the 800. I was getting distortion in left channel and was lighting the speaker fuse light as volume approached 11:00 position. Found the L channel Speaker fuse is blown.

While I was looking at the Schematics the F-1037 board shows TR602/TR604 as 2SC281’s. But on my board they are 2SC458’s same as TR601/TR603. Has anyone seen this and is this a SM error or unlogged update?


I was planning to replace all the 458’s due to their reputation for noise with KSC1845FTA’s.

I’ve seen 2SC1815YTA listed as Replacements for the 2SC281?

What is best approach here? All 1845’s or combination 1845 for TR601/TR603 and 1815’s for TR602/TR604?

Thanks;
JD

 

[tnsilver]

I don't know about this unit but I've seen Sansui use two different types of transistors in differential pair arrangements (undocumented too). I used 4 matched KSC1845FTA for this before. Got a reel of 50 and it was easy to find 4 with 432 hfe.

 

[jd3rdcoast]

I don't know about this unit but I've seen Sansui use two different types of transistors in differential pair arrangements (undocumented too). I used 4 matched KSC1845FTA for this before. Got a reel of 50 and it was easy to find 4 with 432 hfe.

Thanks, the schematic shows two different types as mentioned but the board has only the 2SC458’s in this quad set. i got 200 KSC1845’s in February when Mouser got their order in. So I will get a set matched up and get it done.


JD

 

[jd3rdcoast]

I replaced all 4 2SC458's on the 800 Equalizer Board with matched set of KSC1845FTA's with Hfe of 360. (All within 1 value of each other). The ones I pulled were all over the place from 79 up to 130.

With these and the new left speaker fuse all is functioning properly. No clipping on speaker fuse light or distortion on either channel. But, I noticed the right channel is not as full sonically as the left channel, reduced base and mids. I switched speakers and the problem moves with the right side in both A & B outputs. I'm assuming most likely an electrolytic capacitor issue at this point and wondering which board to look closely at first? Equalizer Board (F1137A), Tone Control (F-1094) or Main Amp (F-1100).

Here is pic of the block diagram:

 

[tnsilver]

I have no idea about the 800 common culprits but if it's anything like my 9090XX units, I'd thoroughly deoxit all the rotary switches first, then see what's next...

 

[jd3rdcoast]

I have no idea about the 800 common culprits but if it's anything like my 9090XX units, I'd thoroughly deoxit all the rotary switches first, then see what's next...

Thanks tnsilver. That was 1st thing I did. Sorry didn’t mention that above. They were pretty noisy. Used DeoxIT D5 followed by a bit of CRC 2-26 lube. Not getting any more noise from any switches and pots.

Just now was checking the big filter caps (in circuit) and got the following:

Coupling Caps:
C821 (2000uf 35V)L= 3340uf
C823 (2000uf 35V)R = 3200uf

Rail Voltages are 56.5 (54). I set the DC balancing at the coupling Caps to 1/2 of supply @ 28.2Vdc. Less then 10 mV at speaker outputs.

Power section caps: Ithink need to pull one leg to test properly out of circuit.
C001 (2000uf) = 190uf
C003 (2000uf) = 225uf
C004 (330uf 35V) = 59uf
C005 (470uf 25V) = 95uf
C006 (470uf 25v) = 358uf (end of line)
C007 (1000uf 16V) = 1,050uf

With 8ohm dummy loads and 1V 1KHz input to AUX all is good till sine wave hits about 30VPP then starts to clip. This Model 800 is rated at 22 Watts RMS into 8ohm.

 

[tnsilver]

Ithink need to pull one leg to test properly out of circuit.

Yeah, some of these measurements yield weird results. Not saying it's impossible, but remember to test with a good LCR meter at 120Hz. Higher frequencies of cheap meters give the wrong idea. Also check ESR at 10KHz and verify against ESR charts, then you'll know for sure. Like this, for example:

BTW: Can you by pass tone control like with hooking up a source directly to main-in? If you can, run that test again and you might cancel some of the suspect PCB's.