Pioneer SA-8100: Protection (yes, again)

dbxdx5

Super Member
A friend of mine asked me to take a look at his SA-8100. He said he had brought it to a tech because the protection light would randomly come on. The tech said the amp couldn't be fixed. Well, we'll see about that....

When I tested the unit with some sacrificial speakers, I initially got no sound out of the left channel. Thinking perhaps the balance pot was oxidized, I exercised it a bit while continuing to play music. After a minute or so, there was faint sound from the left channel, followed by the protection light going on and off, and then music from both channels. Then the I started getting static in the left channel, at which point I stopped the music. But before I could turn the amp off, the left speaker got a good hit of DC.

Left channel DC offset can be adjusted close to 0.00mV, but it's less stable than the right channel (no surprise there), spiking when the protection kicks in. I was able to set bias to 20mV for both channels.

My best guess at this point is that the 2sc1451's and 2sa726's may be causing issues in both the amp board and the protection board, so I plan to replace those, along with the light blue Sanyo caps. Suggestions for other areas to consider before I place my order?

Pin readings and a few transistor readings below.

Control Board AWG017 B

Pin 3: -16.73

Pin 4: 17.99

Q5
E: 16.82
C: 0.83
B: 17.44



Power Supply AWR030 A



Pin 1: 0.00

Pin 2: 0.00

Pin 3: 0.00

Pin 4: -39.2

Pin 5: 39.2

Pin 6: 0.00

Pin 7: 0.00

Pin 8: 0.00

Pin 9: 51.4

Pin 10: 0.00

Pin 11: 0.00

Pin 12: -40.7

Pin 13: -23.4

Pin 14: -16.6

Pin 15: -12.7

Pin 16: 12.8

Pin 17: 17.9

Pin 18: 24.3

Pin 19: 34.4



Protection Circuit AWM027 D



Pin 1: 0.00

Pin 2: 0.00

Pin 3: 0.00

Pin 4: 0.00

Pin 5:

Pin 6: -12.7

Pin 7: 0.00

Pin 8:

Pin 9: ~0.00 with protection off . Spikes and drops when protection kicks in.

Pin 10: 0.00

Pin 11: 0.00

Pin 12: 0.00

Pin 13:

Pin 14: 08.3

Pin 15: 34.5

Pin 16: 34.5

Pin 17: 0.00

Q1
E: 0.00
C: 19.1
B: 0.00

Q2
E: 0.00
C: 19.1
B: 0.00



Power Amp Board AWH022 D



Pin 1: 0.00

Pin 2: 0.00

Pin 3: up to 0.01 fluctuates during protection

Pin 4: 0.00

Pin 5: 51.5

Pin 6: up to 0.15

Pin 7: 00.0

Pin 8: 00.0

Pin 9: 00.0

Pin 10: 00.0

Pin 11: -40.8

Pin 12: -12.7

Pin 13: 12.9

Pin 14: 0.00

Pin 15: -40.8

Pin 16: 0.00

Pin 17: 0.00

Pin 18: 0.00

Pin 19: 0.00

Pin 20: 0.00

Pin 21: 51.5

Pin 22: 0.00

Pin 23: 0.00

Pin 24: 0.00

Pin 25: 0.00



Pin 26: 00.6

Pin 27: 39.3

Pin 28: 00.0

Pin 29: -00.5

Pin 30: -39.2

Pin 31: 00.0

Pin 32: 00.5 – 00.6

Pin 33: -39.2

Pin 34: 00.0

Pin 35: 00.7

Pin 36: 39.3

Pin 37: 00.0



Q3

B: 0.00 fluctuates when protection kicks in
C: -40.7
E: 0.6
 
You're on the right track suspecting 2SA726 and 2SC1451. Replace all of them that you see with KSA992, and KSA2383, respectively.
Also, don't even dick around with the bias trimmers (VR3,4) throw them in the trash and replace them with some new Bourns 3386P or K. Deoxit all of the switches and pots while you're at it.
It wouldn't hurt to clean (or better yet, replace) the speaker relay.

BTW, any one who calls himself a tech and can't fix one of these amps should sell his tools and go to work at McDonald's.
 
MarZutra, stories like yours piss me off to no end. The 9900 is a bit more sophisticated than the 8100, but both are pretty straightforward, conservative, very stable designs.

Pioneer amps from the 70s are not tough to fix.
 
I just rehabbed my 8100. Replaced all suspect transistors: 2SA726, 2SC1312, 2SC1451. Get the service manual and locate all of them and replace. I also replaced all e-caps on all boards except a couple of small ones on the switch board that I just could not get to. You probably won't have to go to all that trouble to get this unit running, but I'd for sure start with the power supply board (source of the protection issue I was having), and the power amp board. After that I'd do the protection board, control amp board, and equalizer.

It really depends how far you want to go, but on my unit, I did not clear up all of the static issues until I had replaced every last one of the suspect transistors.

I love my 8100 and I think the effort to rehab is really worth it!
 
You're on the right track suspecting 2SA726 and 2SC1451. Replace all of them that you see with KSA992, and KSA2383, respectively.
Also, don't even dick around with the bias trimmers (VR3,4) throw them in the trash and replace them with some new Bourns 3386P or K. Deoxit all of the switches and pots while you're at it.
It wouldn't hurt to clean (or better yet, replace) the speaker relay.

BTW, any one who calls himself a tech and can't fix one of these amps should sell his tools and go to work at McDonald's.

Thanks. I was thinking about replacing those bias pots. I'll also swap out the DC offset pots, VR1 and VR2, while I'm at it. Let me know if you recall which Bourns part # is a good sub for those.
 
Please, not KSC2383 for the 2sc1451, use KSC3503!!! They run cooler.
Tarior and I appear to disagree about the maximum dissipation level, and we're
not about to break out the engineering books and slide rules and duke it out!!
BTW I started working on this before Tarior made his first post, so don't think this
is a shooting from the hip reply, it's researched.

The major difference between the ksc2383 and the ksc3503 is physical space, and the old 2sc1451
both allowed more space and also had a built in heat sink over an ostensible TO-5 type case
for heat dissipation.So it becomes a question of gain and how hot it will run.
512-KSC2383YTA to-92L ecb 160v 1a .9w 100mhz 160-320hfe
512-KSC3503dstu to-126 ecb 300v .1a 1.2/7w 150mhz 60-120hfe


Addreesing the OP's question of priorities:
Power supply rebuild
protection rebuild
problem transistors and power amp trimpots
CSSA capacitors (nowadays we spec WIMA's for 0.1uf to 1uf, and if they don't FIT, Nichicon UKL's)
Regular recap

trimpots:
652-3386H-1-101LF 100ohms 3386H Single Turn 3/8" ? 10% parallel to pcb idle current
652-3296P-1-103LF 10Kohms 3296P Multi Turn : 12 3/8" Sealed parallel to pcb dc offset

parts list
http://www.audiokarma.org/forums/index.php?threads/sa-8100-cap-list.463178/page-2

I just added these trimpots to that thread:

trimpots:
652-3386F-1-101LF 100ohms 3386F Single Turn 3/8" ? 10% perpendicular to pcb ccw min#1 triangle
652-3296Y-1-103LF 10Kohms 3296Y Multi Turn : 12 3/8" Sealed facing UP ccw min#1 triangle

and the trimpots
AWH-022 VR1 10k 652-3296Y-1-103LF
AWH-022 VR2 10k 652-3296Y-1-103LF
AWH-022 VR3 100 652-3386F-1-101LF
AWH-022 VR4 100 652-3386F-1-101LF
 
Here is what I was just preparing to post. LOL.


Just an addendum, some techs whom I quite respect, prefer to use KSC3503 in place of 2SC1451 rather than KSC2383. Being a TO126 device, it can dissipate more heat than a TO-92L.
The reason I go with 2383Y (specifically the "Y" gain designator) over KSC3503D, is that it has better gain. I haven't been able to source any 3503 with E or F gain designations.

However, said respected tech is a much better Thevinizer than I am, so if he recommends KSC3503 here, then that is the one to go with.

I'll break out the loop equations book and pocket calculator another time:p:biggrin:
 
I too am concerned with the lower gain of the ksc3503, and it's sitting in the VAS slot in the amp.:dunno:

The lower gain bit me in the SA-9900 preamp section a few weeks back,
as the lower gain caused a biasing resistor correction from 470k to 220k to correctly bias the stage.
That circuit was the ONLY time I had EVER seen Pioneer use that biasing scheme that is so dependent upon gain.

It caused me to reverse myself on my long standing policy of NOT changing parts values from the factory values
 
Many thanks to Mark and tarior. I'll circle back after I've checked the aforementioned work off the list and re-measured everything.
 
Please, not KSC2383 for the 2sc1451, use KSC3503!!! They run cooler.
Tarior and I appear to disagree about the maximum dissipation level, and we're
not about to break out the engineering books and slide rules and duke it out!!
BTW I started working on this before Tarior made his first post, so don't think this
is a shooting from the hip reply, it's researched.

The major difference between the ksc2383 and the ksc3503 is physical space, and the old 2sc1451
both allowed more space and also had a built in heat sink over an ostensible TO-5 type case
for heat dissipation.So it becomes a question of gain and how hot it will run.
512-KSC2383YTA to-92L ecb 160v 1a .9w 100mhz 160-320hfe
512-KSC3503dstu to-126 ecb 300v .1a 1.2/7w 150mhz 60-120hfe


Addreesing the OP's question of priorities:
Power supply rebuild
protection rebuild
problem transistors and power amp trimpots
CSSA capacitors (nowadays we spec WIMA's for 0.1uf to 1uf, and if they don't FIT, Nichicon UKL's)
Regular recap

trimpots:
652-3386H-1-101LF 100ohms 3386H Single Turn 3/8" ? 10% parallel to pcb idle current
652-3296P-1-103LF 10Kohms 3296P Multi Turn : 12 3/8" Sealed parallel to pcb dc offset

parts list
http://www.audiokarma.org/forums/index.php?threads/sa-8100-cap-list.463178/page-2

I just added these trimpots to that thread:

trimpots:
652-3386F-1-101LF 100ohms 3386F Single Turn 3/8" ? 10% perpendicular to pcb ccw min#1 triangle
652-3296Y-1-103LF 10Kohms 3296Y Multi Turn : 12 3/8" Sealed facing UP ccw min#1 triangle

and the trimpots
AWH-022 VR1 10k 652-3296Y-1-103LF
AWH-022 VR2 10k 652-3296Y-1-103LF
AWH-022 VR3 100 652-3386F-1-101LF
AWH-022 VR4 100 652-3386F-1-101LF

Few follow-up questions:

I'm looking at 653-MY4-DC24 for a replacement speaker relay. Correct?

Mouser doesn't have any 512-KSC2073TU in stock. What would be an alternative?
 
MJE15030 would be a big upgrade for the regulator.
Mark compiles these wonderful replacement parts lists, however, I don't think he is implying that every single one of the transistors needs to be replaced. Personally, I advise against it.
Known troublemakers, sure, 'can 'em. Otherwise, leave 'em be.

*Edit* Mouser shows 800+ KSC2073 in stock. The H2 suffix just indicates a higher gain device. Not a problem here.
 
Last edited:
Understood. Speaking of known troublemakers, the list Mark linked to recommends that the 2SC945s at Q3 and Q4 on the Protection Board be replaced with KSC1845s rather than the usual KSC3503s (or your preferred KSC2383s). Is this correct?

AWM-027 Q3 2SC945 r or q : 512-KSC1845FTA
AWM-027 Q4 2SC945 r or q : 512-KSC1845FTA

I only replace them if I have a reason to do so. C945 has proven to be a pretty reliable device.
Now, when I see 2SC869 in a Pioneer protection circuit, I replace those with KSC1845. If they are in a power supply, I replace them with KSC2383.

Addendum; if Mark recommends replacing those proactively in the SA-8800 in particular, then you probably should.
 
Very interesting because the first tech told me different. They 'repaired' it 3 times and told me that they weren't going to work on it anymore because they couldn't fix it. Then I took it to another tech who had it for 8 months. After I was lied to about 5 times, I got so pissed off and drove to his shop, walked past him as he was with another client and I walked behind his desk and back into his shop. Picked up all my gear and put it into the car. Took it to an audio shop in Halifax for repair. They said it was fixed. They replaced all kinds of parts. Their cleaning crew threw out a bunch of parts - my parts - as well the work order..??!?! Then the sales guy tells me the tech didn't do anything. Got 4 or 5 different stories from the two sales guys... When they brought it out to me the top front cover was missing?!?!
Of course they billed me. I wanted to knock the cheeky prick out...
Got it home and put it back into its wooden box - at least that way you cant see the cover missing...
Within a day or two the thing totally pooched....
Was 9.5/10 condition and then for parts or repair... At least a Pioneer fan who was in search for one for ages got it. I hope he managed to fix it up right and source a new cover...

A guy brought me a Marantz 4300 about 17 months ago with a story similar to yours. Several shops over several years, the final one he had to take to small claims in order to get his receiver back.
I had the thing for about a month and a half. It was about $300 worth of work and parts to fix all the cobbled up prior shit repairs, and then proceed with a quality refurbish.
It's still running flawlessly. (Jinx. lol)

What really bothers me is that it's guys who have been at this for 20, 30, or more years doing the kind of work that you describe. It really ticks me off.
As far as I'm concerned, every tech who has been doing this for that long should be as good at it as Mark(thefixer, of course). Both in knowledge and execution.
There is no excuse for an old pro to do crap work.
We have a shop in town that does shitty work like that. The only thing keeping them in business is factory warranty service (it's more lucrative than you think) for major television manufacturers. Warranty service these days consists of diagnosing down to the board level, and requesting a new board from the company. Easy-peasy.
Part of me sometimes wishes that I could birddog their accounts. Same goes for a couple of JBL-authorized service centers in Portland that don't do great work.
 
Tubes are my weak point, I'm just now slowly breaking into that realm (see my sig line). I was in my 30s when I went to school for electronics, and the vintage moniker was just then starting to stick to our beloved 70s gear. I've only worked on a small handful of tube amps/receivers, usually just replacing a few coupling caps or somesuch.
 
I only replace them if I have a reason to do so. C945 has proven to be a pretty reliable device.
Now, when I see 2SC869 in a Pioneer protection circuit, I replace those with KSC1845. If they are in a power supply, I replace them with KSC2383.

Addendum; if Mark recommends replacing those proactively in the SA-8800 in particular, then you probably should.

Thanks tarior. Before I complete my mouser order, I'm looking to confirm that the Omron 653-MY4-DC24 is the drop-in replacement for the original MY4S-US-24DC that it appears to be.
 
Is that one chassis mounted or PCB mounted? PCB is MY4-02-DC24. If it's chassis mount, sometimes the one you listed will actually pop right into the old chassis mount cover. Depending upon what was in there originally. I really wish I took better notes.
 
Scratch that. An LY2 will snap into the old DEC chassis mount cover. It's tight, but it fits.
 
Chassis mounted.

without top mount flange 653-MY4-DC24 $6.63 650 in stock

with top mount flange 653-MY4F-DC24 $14.82 16 in stock
(plus standoffs to mount to the top flanges)
stainless preferred standoff from mouser 15mm (need two per side): 728-FM2110-3005-SS
728-FM2110-3005-A is 4.5mm od 3mm x 0.5 thread 15mm male female aluminum 0.63 /100: 0.43
728-FM2110-3005-SS is 4.5mm od 3mm x 0.5 thread 15mm male female stainless steel 1.07 /100: 0.74
728-FM1262-3005-A is 4.5mm od 3mm x 0.5 thread 15mm female female aluminum 0.59 /100: 0.40
728-FM1262-3005-SS is 4.5mm od 3mm x 0.5 thread 15mm female female stainless steel 0.97 /100: 0.67



If you are going to try the case transplant why spend double.
also there's mattsd's implementation of EchoWar's find, post 38 in the relay thread:
http://audiokarma.org/forums/index....-relay-replacement.150439/page-2#post-9520539
relay flange adapter (528-16-782C1) with a 653-MY4-DC24 (with the plug in terminals).
 
As I mentioned previously, you can snap a LY2 right into the cover of the old bottom flange DEC relay. Since the MY-4 in that application has the contacts connected in parallel, it's not a bad option.
I would have taken a pic of how it works, but all I have on hand is a 48V LY2, and I'm oresently working on a SX-1010 which of course takes a 24V relay.
 
Back
Top Bottom