Realistic STA-78 with high voltage on left channel output

Greetings, fellow stereo curators!

I need some assistance from you today regarding high DC voltage on the left channel in our beloved STA-78 receiver. Dad bought it new in the late 70s and passed it on to me, but when he tried using it this summer in our garage it let out the magic smoke after a loud bang. Since I studied EE in college and love working on old electronics, I investigated and found TR613 (power amp) completely shorted and R623 roasted alive.

I ordered a NOS 2SB509 off eBay for TR613 and installed it, also replacing R623 and D61 which were toasted pretty good. After powering the unit up I found there were no shorts or smoke (yay!) but also found 5V across the left channel speaker outputs (whaaaa?!!!).

Further probing has found absurdly high voltages at all of the left channel transistors all the way back to TR605 but not before, and 5V on the main output line. I attached an image with red numbers to show what I am measuring. The right channel voltages are all very close the factory specs (in the second attached image), and no high voltages are coming into the amp board on either channel.

I saw that TR605 had correct emitter and base voltages but high collector voltage so I figured it was leaking and replaced it with a KSA1013. But when I powered the unit back up with the new TR605 in place I found the exact same high voltages on the left channel!

I am admittedly stumped where to look from here. I thought about another transistor leaking, but since high voltages are even seen at TR605 (which is a “pre amp drive” according to the service manual) I am questioning that. C605 and C607 are not shorting according to my DMM. The 2SB509 I put in is a rev 6J not a D7G as original unit, but I can't image how that would cause an issue like this since I'd think a 509 should be a 509 regardless of revision...right?.

I will continue searching for bad caps or resistors, but am interested in getting your feedback on what should be my next area of investigation. If this unit was on your bench, what part of the circuit would you try looking at next?

Thanks in advance!

upload_2019-1-5_17-16-6.png

upload_2019-1-5_17-16-26.png
 
You mentioned d61?? Do you. Ean d601? If so make sure you Installed a 13v zener diode. Item #2. I always suggest replacing output as a pair not just a single transistor buying transistors on Ebay can be iffy although purchasing used outputs from reputable sellers. You also need to check tr601 to see if it open b to e. If it's bad you need to replace tr601 and tr603 together. Mouser has ksc1845 that work good as differential PAIRS. Then check the trimpot tr601 to see if it burned open also check the drivers tr607 and 609 for opens shorts or leaky
 
.edit I'm very suspicious of TR607. The 7.7 volts on it's base really has no source - TR605 should be totally off. And with a base voltage of 7.7 volts, the emitter should be near 7.0 volts, not 16.

TR611 also looks bad. +16.5 volts on the base but only +5.1 on the emitter? Not on any good NPN I'm aware of.
 
Last edited:
Thanks for the quick replies!

@Ylli: This is why I wanted to post, I was looking so deep into the problem I missed the obvious! You're right, those voltages do look suspicious, I will check out those transistors shortly and post the results.

@Tom B: I will check out TR601 and TR603 as suggested, perhaps 603 is leaking too. However, can you please restate "You mentioned d61?? Do you. Ean d601?" I didn't quite catch what you said here...I didn't replace D601 but I did check it and it is not shorted or open.
 
Might also want to recheck R623. If you truly do have (16 - 5.1 =) 10.9 volts across it, you are dissipating over 1W in a 1/2W resistor.
 
Well folks, looks like I might be in the market for replacement 2SD315 and 2SD438 transistors.

Per suggestion, I checked out TR601 and TR611 again, and their voltages seem to indicate trouble. The below values are what I got with both my Fluke 45 and 115; both pieces are disconnected from the circuit:
upload_2019-1-6_22-7-44.png

TR611 having conduction from Collector to Base looks to be a smoking gun. I will probably get a TR607 replacement even though it seems OK since I agree that that 7V just doesn't come from no where.

I then checked out TR601 and TR603, and neither produced terribly odd readings except for TR601's C>E, which was 1.41 and TR603's C>E which showed 0.67. However, I didn't remove either from the circuit so it might be their parallel configuration causing this:
upload_2019-1-6_22-11-35.png

Checked D601 again, has 0.7 to ground and OL the other way, as expected. VR601 measured 407 ohms across the wiper while VR602 (right channel) measured 207, but I think that is from factory as both have different physical settings on the wiper. Measured R623 in circuit as 99 ohms, but I will need to pull it out tomorrow to double check that it's not being misrepresented.

I will heed your warning about fake transistors, so I presume that something like this (from China) may not be trustworthy? https://www.ebay.com/itm/1PCS-Trans...h=item25e6a7d9e1:g:AlIAAOSwKQ9aIWWr:rk:1:pf:0

Also, since TR607 is a driver will I need to replace it and TR609 with a matching pair? Or can I just replace TR607 on its own? So far my research suggests that the KSA2690 or KSC2383 might be good replacements.
 
I will probably get a TR607 replacement even though it seems OK since I agree that that 7V just doesn't come from no where.
Might want to check TR605 for C-E leakage.
Also, since TR607 is a driver will I need to replace it and TR609 with a matching pair? Or can I just replace TR607 on its own?
The circuitry certainly isn't mirror image, so i don't think a match between TR607 and TR609 is particularly important.
 
Alrighty, sounds good about the matching! I see some posts about good replacements so will get those along with the TR611 replacement.

I checked out TR605 again and while I got .8 V from C to E and 1.8 V from B > E, I'm not sure if these are leaky readings or not. TR605 was already replaced by me with a KSA1013 before I started this post so it should be still good unless the high voltage fried it too.

I also checked out all of the resistors near these transistors, and all measured within a few percent of their spec'd value in-circuit, so looks like I'm safe there. I'll post an update once I get the replacement transistors installed and measured.
 
A BJT with a floating base should read open circuit from collector to emitter in both directions. If you are getting conduction, the part is gone.
 
Did you measure it in circuit or out of the circuit? I assumed out of the circuit. Perhaps a bad assumption. A floating base just means the base is not connected to anything.
 
Ah OK, fair enough. I did measure TR605 out-of-circuit to be sure, by unsoldering it and performing the 6-way test with my multimeter.
 
Just an update, I ordered a 2SD315 I found on eBay from a decent seller and am waiting for it to arrive. For some reason he hasn't shipped it yet even though I ordered it over a week ago...
 
Buying transistors from ebay is not recommended. Specially if you already bought transistors from ebay and you don't trust them. Try to buy transistors and semiconductors from a known components seller like Mouser or digikey just to mention 2 Well known sources.
 
Yeah I'm finding that out haha. I had good luck with the eBay 2SD509 that I put in this unit, though unfortunately both transistors are no longer made or have equivalents that I can find so I'd have to get new TO-220 pairs if I have to get new ones.
 
OK, so as an update I gave up on the eBay fellow and have ordered two 2SD315D's from littlediode.com, which I found recommended here in an older thread. Hopefully this time I actually get them!
 
Alrighty, so I FINALLY have made some progress this evening. For whatever reason, the eBay seller took literally a month to get the 2SD315's I ordered shipped to me, but they finally arrived today so I installed them this evening. After firing it up we have good news and some not-so-great news:
Good news: only 30 mV at the speaker terminals on the left!!! :banana:Still a bit high compared to the right channel, but the spec is "less than 25 mV" so I am happy to even be close there. Now I won't fry a speaker connected to it. I also checked the voltages at TR605 (replaced with KSA1013), TR607 (replaced with KSC2383), TR611 (replaced with eBay 2SD315D), and TR613 (replaced with eBay 2SB509D), and all were within 1 volt if their spec!

Bad news: one of the left channel transistors, and I think it is the one I received tonight (TR611), becomes very hot very fast. Within a minute of idling with no audio on the AUX input I measured 115 F and climbing on the left hand part of the heat sink. Both TR611 and TR613 were too hot to touch while I could easily keep my fingers on the right hand transistors. I then measured the voltage across R33 (emitter resistor for TR11) and found it to be 380 mV while the manual specs it to be 5 mV....

I think I may have been bitten by the Fake eBay Transistor snake, as the new one was a couple millimeters shorter than the old one, and didn't say "Japan" on it like the eBay picture showed. I do have more 2SD315s coming from Little Diode (ordered when I thought the eBay order was lost), but does anyone have suggestion on anything else I can check? Besides being faulty or fake, what else could cause output transistors to become very hot without any load or input? All the voltages on them and other left channel transistors check out OK with no load.
 
There is an idle current adjustment available ion that unit. Have you tried adjusting VR601? CCW should reduce that idle current.
 
I was tempted to check that out, though was concerned that if I connect a load that it may stress them more. I do have a dummy 8 ohm load from Parts Express so am more than willing to give that adjustment a shot if you think that it won't (immediately) blow the transistors. I have the procedure from the service manual, seems pretty straight forward.
 
Check idle current with no load. When bias is too high, I usually do this in order to minimize the risk of burning something: Connect a DBT if you have one. Check the bias with the DBT. Power off the unit. Move the bias trimpot a bit in one sense. Power On the unit again with the DBT. Check if bias moved up or down. Try again until setting the bias near spec. Without the DBT bias will be higher, so you'll need to set it properly without the DBT. I use the DBT when I don't know how the bias trimpot work or if it's damaged.
 
Back
Top Bottom