Pioneer SA8100 refurbishing issues

gslikker

Super Member
Bought myself a dirty and defective SA8100. While troubleshooting a DC-unstable output first, there was a bang and some smoke.
A 50 volts capacitor blew, it was sitting at a location supposed to be 50 volts....
Now, I decided stop troubleshooting before hearing the next bang and just start replacing some components and cleaning up this very nice amp.

I always start the easy way to familiarize myself with the amp. ("Always" is my seventh amp refurbishment, in my case). The first thing to do will be the protection unit. On the internet manuals/schematics I find the AWM027 A board. Mine is AWM027-D, having no R19 resistor compared to -A, and two 0.22 uF electrolytics for C1 and 2. Q1 and Q2 are corroded leads 2SC869

issue 1: The power supply board employs 50 volts caps, where some 50 volts to be measured according the schematics. This is rather odd (no margin).

issue 2: There is no protection diode over the AWM027 relay coil, it may be no good for the transistor.

issue 3: two 0.22 uF 25V caps, C1 and C2, are small electrolytics. The "+" legend on the circuit card is the wrong way around for both. As a result, they were mounted the wrong way around..


This is the first circuit card, guess I'll find more "issues" as I go along.
So, I guess I post the interesting "issues" as I find them, if anybody interested?
 
continuing:

Issue 1 again: probably this was one 2SB527 found designated "diode funtion" (shorted) by my Peak Atlas transistor checker, and one 2SB530 having low current HFE of only 17 (other one around 70). which means 2 transistors dead.

Issue 4: one 2SA628 pulled somewhere HFE only 59 (another one 274 which is a rather big difference.

Issue 5: many transistors have leg corrosion

Issue 6: control amp has different caps compared to the schematic.

For the circuit cards containing semiconductors, only the RIAA equalizer amp and the control amp have yet to be recapped and new transistors. more difficult will be soldering everything back and testing so I am some 25% done, now.
No pictures avail yet....
To be continued....
 
I recently repaired an SA8100 - twice. After the first repair, I damaged it AGAIN by not discharging all the caps before re-assembly. doh!

Good luck with your work there. Keep us updated.
 
1. yes use higher voltage caps, 63s 80 or 100's as the voltage can go higher because that DC is NOT regulated. AFTER the regulator (35v circuit) I would use 50v caps. The caps have shrunk over the years so going up in voltage will still make the size constraints.

P.S. a lot of the 9100 stuff is applicable to the 8100.

2. so add a flyback diode (1n4004) , cathode (banded end) goes to the more + polarity relay coil connection, pin 15 IIRC, and pin 14 for the anode...

3. use film caps as replacements, polarity won't matter as they are non-polarized.

4. yecch, but 5 covers this. the 628's are the current source stage, but gain is still desired...

5. leg discoloration is NOT the source of problems. Still, most modern transistors of the appropriate class will be superior to what was used that long ago. Quite frankly, I have had excellent results with almost complete audio transistor replacements. IIRC there are parts lists of sa-9100 transistor substitutions that should answer most questions... My preferred replacements are the ksa992 and ksc1845 but these can also be 2sa992 and 2sc1845 or equivalents. Low noise higher voltage tolerant small signal transistors. You are NOT on your own in finding substitutes, I can help...

6. re: different caps, 0.1 to 1uf are best replaced with stacked film caps. Obviously go with the values found for the replacements, as for the TYPE of the replacements use the parts lists in the manual to discern the cea general purpose caps from all the other more specialized caps (cssa, ceanl, csza) and replace the specialized caps with Nichicon UKL caps. There are places where noise will be an issue if general purpose caps are used instead. The stacked film caps are low noise already so no worries there. The Nichicon UKL caps are the ONLY ones left with this lower leakage / lower noise property, and there ARE critical locations where they WILL make a difference.
 
Thank you for your input Mark.
I am already busy repopulating the boards Equalizer, Control, protection and main amp board I took out totally, and the protection diode I did put on the solder side. On the control and equalizer board I also replace some resistors using metal film where I "believe" they can add noise

btw, do you have an opinion about METAL CAN tantalums?
I have a lot (2 pounds of them at least) and they look good on the LCR meter.
Better than the average drop shape tantalum "mini bombs". However, I never tested them on leakage...
Also when looking on a huntron tracker component tester I have a nice elliptical shape. If the shape is not elliptical, the capacitor would be having rectifying properties or other irregular behavior (have seen that testing a drop shape tantalum which I removed from a pre-amp because of obvious leakage infuencing a DC voltage).

I did put these metal can caps in several of my amps already.
Of course, I did read all the opinions on tantalums already, it is not enough to keep me from using them yet :thmbsp:
 
Now, issue 6 is less valid as there exists another schematic. Although there is still a cap size mistake in there.

and now:

Issue 7: The heatsink is not flat! See the middle of the picture: around one leg hole there was some "aluminium-distortion" and apparently it was flattened but not enough....Although I noticed one transistor did come loose very easy and also having trouble getting this glue off (the mica washers were glued to heatsink and transistor), I just did see this now, wanting to mount new transistors. I wonder why I did not notice it earlier, but is is most evident when using a lot of light from a certain angle.

pioneer-sa8100-heatsink1.jpg
 
Last edited:
And some more troubles putting things back together.

Wheras the bias was not properly adjustable on one main amp channel measuring 0.3 volts over a b-to-e juction of a driver transistor, took the scope and found out both channels starting oscillating when bias of output transistor just coming in.

Of course, this is not uncommon, just replacing transistors selecting different types "thinking" they are the right ones. Especially if data available is limited and direct 1:1 replacements are probably not there.

This is a kind of moment where I am stuck, lacking the knowledge of calculating an amp for AC behavior.

Just guessing, I put a 100 pF cap between B and C of Q9 (Q10 other channel) it worked fine.
Then tried 15 pF it still works fine, so I left it that way (15 pF it is).

After this a new "issue number" 6 came up: the protection was working intermitting. Apparently, one of the 2SC945 in the protection circuitry was "sensitive" . Q4 was defecitve, Q3 and Q5 were replaced also as precaution.
Running at sine wave 40 watts for a while, heating up the amp, the bias stays quite the same. So this diode bias thing works very well.

Now putting the last bits and pieces together.
Anyway some more testing to do but already the music sounds fine.

See the funny six brown "things" looking like "film caps" on the closest circuit card on the picture, actually they are ancient precision resistors for the phono amp.

pioneer-sa8100-inside1.jpg
 
Fun, I just received one for repair. Indeed not the easiest amp to work on.

This one had the infamous bad RIAA equalizer amplifier board issue. Playing for some while, then one channel losing bass and generating a loud hiss instead (that at least was what the owner told me).

I replaced the electrolytics, tantalums and transistors on this board. I guess the issue was with the egg shaped tantalums, one was shorted, the other one had a value of 25uF, while according to the service manual it should have been 2,2uF. All ESR's were ok though. I replaced both with regular electrolytics (Panasonic FC) and the thing plays fine. Apart from the tantalums it also had one below spec electrolytic (470uf was 280) and the gain of the right side transistors was noticeable lower than those at the left side.

Just wonder why you use BC's? I replaced 2SA774 / 2SC1345 with KSA992 / KSC1845 and all is ok. The ECB KSC1945's even fitted better than the BCE 2SC1345's, while the board seems to be designed for ECB.

Nice, I did not realize these brown things were resistors!
 
Hi Jeroen, it seems we get the same type of amps many times, or is it just because we apparently like this kind of stuff.
I just guess the BC's are not worse than 2sa992 in the input stage, they are designed for low noise at low current / not too low impedance. However I never did a noise test. My BC559c have higher HFE compared to the Mouser KSA992 I have.
I still have the BC so I use them.

The ONLY dropshape tantalums I use, are a lot of 10 uF 35 V and 50V ones I pulled from some professional equipment, I am 100% sure they are very good (never had a failed one although hundreds of them were in a system) and they measure very good (lower ESR than metal ones or new Panasonics on all frequencies of the lcr meter).

I have found several defective dropshapes already (not in this amp), although it does not discourage me of using ones I consider better. But my small box of new (NOS) drop shapes, guess you see them on "marktplaats" one day ;-)

Now, i have to fix my Philips 212 turntable I got from kringloopwinkel (local goodwill store alike for USA readers) it doesnt do 33 rpm. Decided I should have one, I try it for a while and if a turntable is too much in the way for me it will probably go back to kringloopwinkel (unfortunately I sold my MC cartridge and homebrew MC amp >25 years ago after deciding I outgrew wasting my time on audio stuff...)
 
Well, this Pioneer is a coincidence, but I think we indeed share the same love for at least Sansui and Luxman, the real vintage ones that is. I tried to think of a common name to indicate exactly which ones and I think I have found a nice one; "the charming amps".

I too think the BC's and KSA/KSC can be used to replace older 2SA's and 2SC's but since you mentioned you had loads of KSA992/KSC1845's I just wondered if you had a specific reason to use them here. Well, stock is a plausible reason I guess :).

But now that you mention Philips, I am working on one too. Well, 2 in fact, 2 MFB speakers 22AH532. What a disaster to work on those, there are no marks at all at the pcb and it looks like being produced by a bunch of looneys with a lot of left hands. Still, I want to have them ok again, so it's a nice challenge.

As for the drop shape tantalums, I wouldn't see any reason not to use them, by the time someone comes to file a claim because they exploded and ruined their complete amp I will have safely been put away in an elderly home, so just let them come and get it.

:D
 
Back
Top Bottom