Technics SA-5760 Stereo troubleshooting techniques

[tdlunsfo]

Ok, I need some help from the guru's on this one. I have done a handful of stereo's over the years and since there aren't a lot of people doing this work in my area, I took this in against my better judgement. This amp came to me with blown outputs in one channel, I replaced those and of course had issues with the driver board. Transistors are all obsolete so I had to get creative with replacements, using a combo of old and new parts got this thing sounding pretty good. Hours later poof, blown outputs. I replaced them with some older pulls and got it up and running again and noticed that I had it biased wrong (creeping current on my variac) and it was going into thermal runaway... my mistake.
I found the bias instructions in the service manual (in a picture in the bottom corner.. I obviously missed this before), biased it up and everything was sounding really good. Poof again! Except this time it really let out the magic smoke on the driver board - outputs, drivers, a handful of resistors all toasty. I replaced the parts, pulled the output board and now I still have 68v on the diff amp feedback and the NPN driver getting really hot with no signal. All the parts test good individually but obviously something isn't right or is still wounded.

With the goal like before to get the driver board working first, my questions are around how I can troubleshoot this circuit without throwing parts in it? My knowledge holes with this design have me hampered. For instance, can I disconnect the drivers and still get a balanced circuit or do these have to provide the negative feedback for the diff amp? Can I disconnect the VAS and constant current sources and still troubleshoot the diff amp? What about the current mirror.. how do I troubleshoot that? I think I can disconnect the current limiters for now but want to verify. I want to put some brand new parts in this thinking my older parts from before where a culprit, but TO3 output transistors are expensive and I don't want to go through this exercise again if I can get away with it.

TR605 and TR607 are mentioned as pre-divers in the manual but I view 605 as a constant current source and 607 as the voltage amp.
Obviously 601 is the diff, 615/617 are drivers.
609/613 are current limiters and 611 is the overload alarm switch (which worked last time after the driver board almost burned down..lol)

As always any help or transfer of knowledge is appreciated.

[J M Fahey]

To each his own, Iīll show my way, others may differ:

1) I donīt use Variacs, not for Technical reasons but because of Human error, they are not automatic but require *your* action and vigilance.
Nice to rise voltage slowly and watching current meters, good for troubleshooting, but you leave it on and turn your attention elsewhere and you are alerted by smoke or blown fuses.
Bulb limiters worst case shine bright but much better: limit current/damage even if you are not nearby.

2) SS amps are DC coupled and high NFB throughout, repair it as a unit (which it is), separate driver board is only a manufacturing/assembly convenience.

3) the basic circuit is quite straightforward/standard, *good* , being Japanese I expected some crazy nerdy ideas thrown in.

4) that said, I find it weird they use a 60īs technology thermistor for bias control: unobtainable, not easy to replace, they vary a lot in charasteristics even if you get a NOS one inside a bamboo and rice paper lacqured box, the works.
Hope it *at least* is touching the heatsinks!!!!

Found annoying that they unnecessarily list each and every jumper as a "component" on the schematic. Oh well, Japanese OCD.

5) replace all suspect transistors and visibly burns/toasted parts, plug it into a smallish tungsten bulb, say 40W or so, 75W max, and turn it ON.
No load of any kind, of course.

Then the actual troubleshooting starts.

6) measure voltages , starting with speaker out, and : Vce and Vbe for all power handling transistors, so

* TR 621 to 628 POWER
Vce and Vbe for each of them

* TR 615 to 618 DRIVERS
same thing.

* TR 605 to 608 VAS and constant current sources

* since we are at it, some "auxiliaries" too:

TR 609 610 613 614 short protection.

TR 611 612 overload detectors???????

TR 603 604 VAS darlingtons?

and to fill the basket: all 5 pins of input differential pairs: both Vbe and Vce so 4 +4 measurents in total.

* a pet peeve: do NOT pull parts at random just to measure them "outside" which destroys PCB, I much prefer "functional" readings which "tell the truth" and if anything remove real suspect parts to test outside.

Ok, roll up your sleeves and ......

[tdlunsfo]

Juan thanks for the reply. I wasn't clear before but both times the stereo failed it had operated for at least an hour. The second time for 8 hours so I was at full power. I'll pull out my dim bulb this time though. I typically don't pull parts except when they test bad but I had some other issues separate from this with some "popping" that suggested some of the old parts needed to go. As of right now all the resistors, transistors and electrolytics test ok but I have 68v to ground at test point Z and it tries to burn the driver emitter resistors if I leave it on for more than 20 seconds. I still have the finals disconnected as I am down to my last pair and they are pulls from an old Peavey that I don't really trust. Also there is a transistor near the output that helps with the bias. It's tough to see but in the middle of the schematic with all the heavy lines TR619/620.

[tdlunsfo]

So I got the driver board working by swapping out the Diff, VAS and current mirror. Although they tested good with my DMM diode test they certainly were not operating well in circuit. I had 68v on the base of my drivers sending them into saturation and burning up the driver emitter resistors. Now.. I feel like I can plug in my finals. I plan to buy brand new finals, bias it up, and try to run it on my dim bulb...but I'd be lying if I said I feel good about it as this thing has me spooked. Should I just run it with no finals on the dim bulb for awhile to see if the driver board survives?

[tdlunsfo]

I got this up and running again. I tried testing on the dim bulb but its not allowing enough voltage to fire the speaker relay so back on the variac, luckily I have a 3A and 1A mode. Since this stereo idles around 600mA, I feel confident I can catch a fault with the 1A fuse before it melts down the driver board again. On this unit you can set Icq on each side and I have been monitoring it nonstop as I suspect I got into a thermal runaway scenario last time. How long would you all give it till you call it good? I'm thinking a couple of days of playtime at low volume, then a day on 3A variac at higher volumes.

[Jazz P Bass]

Good luck.

SS repairs can be a bear!

[Mars Amp Repair]

To each his own, Iīll show my way, others may differ:

1) I donīt use Variacs, not for Technical reasons but because of Human error, they are not automatic but require *your* action and vigilance.
Nice to rise voltage slowly and watching current meters, good for troubleshooting, but you leave it on and turn your attention elsewhere and you are alerted by smoke or blown fuses.
Bulb limiters worst case shine bright but much better: limit current/damage even if you are not nearby.

2) SS amps are DC coupled and high NFB throughout, repair it as a unit (which it is), separate driver board is only a manufacturing/assembly convenience.

3) the basic circuit is quite straightforward/standard, *good* , being Japanese I expected some crazy nerdy ideas thrown in.

4) that said, I find it weird they use a 60īs technology thermistor for bias control: unobtainable, not easy to replace, they vary a lot in charasteristics even if you get a NOS one inside a bamboo and rice paper lacqured box, the works.
Hope it *at least* is touching the heatsinks!!!!

Found annoying that they unnecessarily list each and every jumper as a "component" on the schematic. Oh well, Japanese OCD.

5) replace all suspect transistors and visibly burns/toasted parts, plug it into a smallish tungsten bulb, say 40W or so, 75W max, and turn it ON.
No load of any kind, of course.

Then the actual troubleshooting starts.

6) measure voltages , starting with speaker out, and : Vce and Vbe for all power handling transistors, so

* TR 621 to 628 POWER
Vce and Vbe for each of them

* TR 615 to 618 DRIVERS
same thing.

* TR 605 to 608 VAS and constant current sources

* since we are at it, some "auxiliaries" too:

TR 609 610 613 614 short protection.

TR 611 612 overload detectors???????

TR 603 604 VAS darlingtons?

and to fill the basket: all 5 pins of input differential pairs: both Vbe and Vce so 4 +4 measurents in total.

* a pet peeve: do NOT pull parts at random just to measure them "outside" which destroys PCB, I much prefer "functional" readings which "tell the truth" and if anything remove real suspect parts to test outside.

Ok, roll up your sleeves and ......

One SS troubleshooting important issue is when firing up, DO IT WITH NO LOAD. Reason being that if there is an imbalace between positive half and negative half of the output stage, you will end up with DC offset at the speaker terminals sometimes at full rail potential & the amp will do its level best to drive that offset into the load (speaker or dummy). Yes if you have a bulb limiter in the ckt that might keep the amp from destroying itself, but why take the chance.
I typically, as Jazz P listed, monitor the output dc offset on meter and scope (to be certain it's not going into oscillation) before the speaker relay as well as the bias with a meter and the bulb limiter inline when firing up the amp, Quite often the bias point is between the emitters of the output xistors and the speaker line so you can use one of the bias meter leads to monitor the offset between that point and ground.
Beware tho that the output stage typically needs around 40Vac line voltage in order for the bias ckt to start functioning so you may see the bias increase initially until you get past whatever the threshold is for that partiular amp. Typically tho the DC offset does not go crazy before you reach that threshold.
As P mentioned, HIFI troubleshooting on that old equipment can be tricky as there are about 100 electrolytic caps in some of those old receivers in the sigal path, but for your type of issue only the ones on the driver boards would affect the output stage.
Remember, the input capacitor to the driver board can also be leaky causing DC offset & general mahem in the output stage. g