Returning to Complete Repair of an SA 9500II: an Old Friend

I am returning now to repair and recap my old favorite, the SA 9500II that I started to fix a while back and accidentally blew the outputs and some other unknown things yet to be found, when I was trying to set the bias. A couple of years have gone by and I have had more time inside some Onkyos, Pioneers and other units and I feel more confident that I will be able to accomplish this task, My thanks go out to Avionic, Markthefixer, Echowars and the others who helped and contributed to this attached thread: http://www.audiokarma.org/forums/showthread.php?t=108465 , Which was phase I of this project. I now settle in to phase II. I wish for it to go better.

My first question gets directed specifically to Markthefixer regarding the output transistors that I will use for this project. I already have the OnSemi MJE15032/33 purchased and the little brackets and jumpers to the heat sinks ready to go. I did all that two years ago. I do, however remember a more recent thread that mentioned a replacement transistor that was more robust than the 32/33s. Mark, if you remember that discussion, what was that transistor and would it be worth it for me to purchase them, even though I already have the 32/33s ready to go? I only would do so if there was a solid technical reason to do so. If the ones I have now will do the job, then I will stick with them.

Tonight I will do a total cap count, from all the boards, and inventory and prepare my order. Since I have this unit already totally disassembled, I plan to do all the things that need to be done. Can anyone tell me if there are some other things that should be replaced, like the tantalum caps in an SX 1250 or the "Flying Saucers" in the Sansui Model 8, that I would not normally know about. I am open to all suggestions and hints. I especially would be glad to hear from those of you that have done the 9500II before. I will try to post as I go and add photos of what I have done.

If those of you that read or remember the thread I have attached here, and have any suggestions of where I should be looking on the driver boards for problems, hints are welcomed. If I should be looking on the PS board, please let me know because I am assuming it is alright and not the reason the relay is not engaging. The power on light was engaging when I last turned this on. Also, would replacing the relay, or its accompanying transistors or resistors, as a matter of course, be a good idea?

Thanks again for any assistance rendered. It is important to me that I accomplish this, but I readily admit that I will be needing help.

First the (easy) caps.....

(read this:http://www.audiokarma.org/forums/showthread.php?p=3312664&highlight=permission#post3312664 to see what I did for the 727, post 35 is what he sent me and post 42 are the results, I will do the same for the 9500ii , and the added advantage is the BOM entry block of text at the end that saves a LOT of alphabet soup typing...)


list out ALL the caps labeled electrolytics in the service manual
that'll get all the tantalums and solid aluminums etc...
(and eyeball check all of them!),
in this format (for example):

awg-045 C15 4.7uf 50v ceanl

an explanation:
awg-045 is the control board
c15 is the position
4.7uf
50v
ceanl <<<<----- this is THE most IMPORTANT PART of the description

this is from the parts list on page 43 of 47 in the hi-res pdf manual....

and I'll get the replacement cap parts numbers to you....
it can either be a text document
or a zipped up spreadsheet...

AND
watch out for things like :
awf-024 C21 4.7uf 25v ach-318 (which doesn't tell me that it's a NON-POLAR cap)
awf-024 C22 4.7uf 25v ach-318 (which doesn't tell me that it's a NON-POLAR cap)

which either the schematic lists as "n.p." or when you LOOK at it, there's NO negative stripe on it.

as for the transistors:
the sa-9500ii originals are

2sb618a pnp batwing bce 150v 7a 80w 80hfe 14mhz
2sd588a npn batwing bce 150v 7a 80w 80hfe 14mhz


863-MJE15033G pnp to-220 bce 250v 8a 50w 30mhz 70hfe $0.87
863-MJE15032G npn to-220 bce 250v 8a 50w 30mhz 70hfe $0.87

(more powerful to-220 replacements for the batwing transistors...
512-FJP1943OTU pnp to-220 bce 230v 15a 80w 30mhz 80-160hfe $1.21
512-FJP5200OTU npn to-220 bce 230v 15a 80w 30mhz 80-160hfe $0.85

and since you are ordering capacitors anyway, it's not hard to guess what my recommendation would be... especially at 2 bucks per pair....

Mark, I did not expect you to go to that degree of trouble for me, but I will gladly accept the ordering assistance. I am presently printing out the appropriate portions of the SM but I am, of course, running into printer issues. I'll have to continue tomorrow. I still have to suss out the issue on the amp boards but first things first. I am not going to rush this whole procedure. As you subtle hint suggests, I will order the other transistors. Since they are a to-220 style, I will assume the brackets setup I have already put together will work for them too. Thanks for the reply, I'll update as soon as I can.

WE (everybody) end up with a good recap list for the SA-9500ii.

I benefit from having the list mostly entered (time consuming), and with the way I have things set up here, the selection goes quickly, with the side benefit of the easy extraction of the BOM list.

I'll email the results back to you and you can post them.

AK Email sent to Mark with the Caps list. Thanks. The nut holding on the Alp volume pot was installed by Godzilla, arrrrgggghhhh!!!

An update. At this point I have taken pics of all board from multiple sides. I have done a search through the driver boards left and right for problems. There are some due to my previous error. Some issues are discrepancies in resistor ratings which I attribute to the resistor being tested in circuit. Here are the results:

Driver Transistors 2sb536 are shot on both sides
Driver Transistors 2sd381 are good on both sides
Resistor 1 - 330k is open on both boards
Resistor 3 - 62k is open on right board
Resistor 4 - 43k is open on left board
Resistor 26 - 10ohm 2watt shows no resistance at all on both boards yet resistor 27, which is also a 10 ohm 1w resistor shows 10 ohms. Resistor 26 is not a meter sensitivity issue.( After pulling 1 leg and retesting it reads 10 ohms, right where it should be)
Resistors 22 >25 show 50ohm, not 100 as listed. This is true on all resistor on both boards.
All diodes on both board show good.
Transistors Q 1>4 and 7 have not been tested yet.

My questions for the AK masses, at this point, are; are the resistors 22>26 that showed different readings, but were consistent for both boards, bad? I suspect that they may be okay except for R26. I have never heard of a resistor turning into a dead short, but then again, I don't get out much.

Next, since I have to replace the 2sb536 transistor should I replace the 2sd381 transistor? I have not done any replacement research yet. Are there any suggestions for replacement from any AKers.

I completed the cap list from the service manual and then verified it visually on all boards. Board AWF-024 contained two non-polar caps on C21 an C22. Here are the results after they were addressed by Mark and updated to include Mouser order #s. Many thanks Mark. I don't know how you do all this stuff and still have time to fix things. Here's the list


AWH-052r C4 22 50 CEA220P50 647-UPW1H220MDD 22 50
AWH-052l C4 22 50 CEA220P50 647-UPW1H220MDD 22 50
AWH-052r C5 470 6 CEA471P6 647-UPW1E471MPD 470 16
AWH-052l C5 470 6 CEA471P6 647-UPW1E471MPD 470 16
AWH-052r C15 1 50 CEANL010P50 647-UPW1H010MDD 1 50
AWH-052l C15 1 50 CEANL010P50 647-UPW1H010MDD 1 50

AWS-109 C7 0.68 35 CSZA-R68M 35 667-ECQ-V1H684JL 0.68 50
AWS-109 C8 0.68 35 CSZA-R68M 35 667-ECQ-V1H684JL 0.68 50
AWS-109 C9 0.22 10 CSSA-R22M 10 598-DSF050J224 0.22 50
AWS-109 C10 0.22 10 CSSA-R22M 10 598-DSF050J224 0.22 50
AWS-109 C11 330 6 CEA-331P 6 647-UPW1A331MPD 330 10
AWS-109 C12 330 6 CEA-331P 6 647-UPW1A331MPD 330 10
AWS-109 C13 100 16 CEA-101P 16 647-UPW1H101MPD 100 50
AWS-109 C14 2.2 50 -7.5vac cap 647-UPW1H2R2MDD 2.2 50
AWS-109 C19 330 80 CEA-331P 80 647-UPW2A331MHD 330 100
AWS-109 C20 330 80 CEA-331P 80 647-UPW2A331MHD 330 100
AWS-109 C21 220 63 CEA-221P 63 647-UPW1J221MPD6 220 63
AWS-109 C22 220 63 CEA-221P 63 647-UPW1J221MPD6 220 63
AWS-109 C25 47 16 CEA-470P 16 647-UPW1V470MED1TD 47 35
AWS-109 C26 47 16 CEA-470P 16 647-UPW1V470MED1TD 47 35
AWS-109 C29 47 35 CEA-470P 35 647-UPW1V470MED1TD 47 35
AWS-109 C30 47 35 CEA-470P 35 647-UPW1V470MED1TD 47 35
AWS-109 C34 3.3 50 CEA-3R3P 50 647-UPW1H3R3MDD 3.3 50
AWS-109 C35 3.3 50 CEA-3R3P 50 647-UPW1H3R3MDD 3.3 50

AWF-024 C1 1 50 CEANL-010P 50 647-UPW1H010MDD 1 50
AWF-024 C2 1 50 CEANL-010P 50 647-UPW1H010MDD 1 50
AWF-024 C9 470 6 CEA-471P 6.3 647-UPW1E471MPD 470 16
AWF-024 C10 470 6 CEA-471P 6.3 647-UPW1E471MPD 470 16
AWF-024 C17 3.3 50 CEA-3R3P 50 647-UPW1H3R3MDD 3.3 50
AWF-024 C18 3.3 50 CEA-3R3P 50 647-UPW1H3R3MDD 3.3 50
AWF-024 C21 4.7 25 ACH-318NP 647-UVP2A4R7MED1TD 4.7 100 (Mark caught this and C22 as Non Polar.
AWF-024 C22 4.7 25 ACH-318NP 647-UVP2A4R7MED1TD 4.7 100 I missed this. Thanks Mark.)
AWF-024 C31 47 35 CEA-470P 35 647-UPW1V470MED1TD 47 35
AWF-024 C32 47 35 CEA-470P 35 647-UPW1V470MED1TD 47 35
AWF-024 C35 47 35 CEA-470P 35 647-UPW1V470MED1TD 47 35

AWG-045 C15 4.7 50 CEANL-4R7P 50 647-UKL1H4R7KDDANA 4.7 50
AWG-045 C16 4.7 50 CEANL-4R7P 50 647-UKL1H4R7KDDANA 4.7 50
AWG-045 C17 4.7 50 CEANL-4R7P 50 647-UKL1H4R7KDDANA 4.7 50
AWG-045 C18 4.7 50 CEANL-4R7P 50 647-UKL1H4R7KDDANA 4.7 50
AWG-045 C25 4.7 50 CEANL-4R7P 50 647-UKL1H4R7KDDANA 4.7 50
AWG-045 C26 4.7 50 CEANL-4R7P 50 647-UKL1H4R7KDDANA 4.7 50
AWG-045 C27 4.7 50 CEANL-4R7P 50 647-UKL1H4R7KDDANA 4.7 50
AWG-045 C28 4.7 50 CEANL-4R7P 50 647-UKL1H4R7KDDANA 4.7 50
AWG-045 C31 100 16 CEANL-101P 16 647-UPW1H101MPD 100 50
AWG-045 C32 100 16 CEANL-101P 16 647-UPW1H101MPD 100 50
AWG-045 C33 4.7 50 CEANL-4R7P 50 647-UKL1H4R7KDDANA 4.7 50
AWG-045 C34 4.7 50 CEANL-4R7P 50 647-UKL1H4R7KDDANA 4.7 50
AWG-045 C41 4.7 50 CEANL-4R7P 50 647-UKL1H4R7KDDANA 4.7 50
AWG-045 C42 4.7 50 CEANL-4R7P 50 647-UKL1H4R7KDDANA 4.7 50
AWG-045 C43 4.7 50 CEANL-4R7P 50 647-UKL1H4R7KDDANA 4.7 50
AWG-045 C44 4.7 50 CEANL-4R7P 50 647-UKL1H4R7KDDANA 4.7 50
AWG-045 C47 100 16 CEANL-101P 16 647-UPW1H101MPD 100 50
AWG-045 C48 100 16 CEANL-101P 16 647-UPW1H101MPD 100 50
AWG-045 C49 4.7 50 CEANL-4R7P 50 647-UKL1H4R7KDDANA 4.7 50
AWG-045 C50 4.7 50 CEANL-4R7P 50 647-UKL1H4R7KDDANA 4.7 50
AWG-045 C51 100 35 CEA-101P 35 647-UPW1H101MPD 100 50
AWG-045 C52 100 35 CEA-101P 35 647-UPW1H101MPD 100 50
AWG-045 C53 100 35 CEA-101P 35 647-UPW1H101MPD 100 50
AWG-045 C54 100 35 CEA-101P 35 647-UPW1H101MPD 100 50


598-DSF050J224
598-DSF050J224
667-ECQ-V1H684JL
667-ECQ-V1H684JL
647-UPW1H010MDD
647-UPW1H010MDD
647-UPW1H010MDD
647-UPW1H010MDD
647-UPW1H2R2MDD
647-UPW1H3R3MDD
647-UPW1H3R3MDD
647-UPW1H3R3MDD
647-UPW1H3R3MDD
647-UVP2A4R7MED1TD
647-UVP2A4R7MED1TD
647-UKL1H4R7KDDANA
647-UKL1H4R7KDDANA
647-UKL1H4R7KDDANA
647-UKL1H4R7KDDANA
647-UKL1H4R7KDDANA
647-UKL1H4R7KDDANA
647-UKL1H4R7KDDANA
647-UKL1H4R7KDDANA
647-UKL1H4R7KDDANA
647-UKL1H4R7KDDANA
647-UKL1H4R7KDDANA
647-UKL1H4R7KDDANA
647-UKL1H4R7KDDANA
647-UKL1H4R7KDDANA
647-UKL1H4R7KDDANA
647-UKL1H4R7KDDANA
647-UPW1H220MDD
647-UPW1H220MDD
647-UPW1V470MED1TD
647-UPW1V470MED1TD
647-UPW1V470MED1TD
647-UPW1V470MED1TD
647-UPW1V470MED1TD
647-UPW1V470MED1TD
647-UPW1V470MED1TD
647-UPW1H101MPD
647-UPW1H101MPD
647-UPW1H101MPD
647-UPW1H101MPD
647-UPW1H101MPD
647-UPW1H101MPD
647-UPW1H101MPD
647-UPW1H101MPD
647-UPW1H101MPD
647-UPW1J221MPD6
647-UPW1J221MPD6
647-UPW1A331MPD
647-UPW1A331MPD
647-UPW2A331MHD
647-UPW2A331MHD
647-UPW1E471MPD
647-UPW1E471MPD
647-UPW1E471MPD
647-UPW1E471MPD

Just for a reference, the filter Caps were replaced in May of 2007. It was after that that all this madness started. The caps I used were Panasonic ECE-T1JA153FA, CAP 15000UF 63V ELECT THA SERIES, Digikey part #P10638-ND. These are the snap in type caps to which I added ring lugs to make them easier to connect to.

I'll keep this thread up to date as I progress. I hope tonight to be able to post pics of all the boards as a reference for other AKers.

Here are the pics of the SA9500II. These are really nude photos as the unit is completely broken down to its component parts. Unfortunately these amps were wired with wire wraps and did not use plug in modules as some of the other Pioneers have so dis-assembly requires a very time consuming tear down. This will also require some considerable unit juggling to get each board in a position where it can be worked on. I've been in my SX1280 and it takes some manipulation but the cards can be separated and that is a big benefit for repair work. With removable cards, it is possible to repair one section, reinstall it and check for proper operation. With the SA 9500II, that will not be possible. When I am done I will have to try some type of partial reassembly to test it first. I think it will be a bit tough keeping the cards safely isolated.

First are photos of the unit before I replaced the filter caps:

These next photos were taken today and show numerous views of all the cards:

Continued:











These tiny caps between the switches are the two electrolytic caps that are the aluminum ones that Mark was referring to----I think:scratch2::



More card shots, I hope these will be useful later for AK reference:







Here are some full body nudes showing the new filter caps installed 2 years ago: I have not yet removed the faceplate because a monster tightened the nut to the volume pot and I have to get my deep well socket set to remove it.







Here is the sheet I use when I tested the amp driver board today. Because there are two boards I use a color code to ID left from right. The blue dots on the paper read as black. I mark one side of the component with one color to show it tested good. Any notations of issues or problems are written and marked with a pink or yellow highlighter. This works well for me.



Tomorrow will start the testing of the rest of the transistors on the driver boards. Thanks for visiting my thread. I hope it is helpful to others on AK.

It looks like the OnSemi MJE15028/29 or the MJE15030/31 might be the right choice for the driver transistor replacements. The 28/29 complimentary pair seems to be closer to the original but the 30/31 pair are a little beefier. The TIP31C/32C combo is also a possibility, but I am still inclined toward the OnSemis. If anyone has any opinions on the choice of either of these, please speak your mind.

I was unable to come up with time to test the other transistors on the driver boards today. Maybe tomorrow. I finally got the nut off the volume pot. I had to use a deep well socket and ratchet to do it. The screws holding the third faceplate on were also installed by Godzilla and I have stripped three trying to get it off. Now I have to order a screw extractor online so I can get them out. This will stall me for time a little. I ordered a set of Pro-Grabits to see how well they work. Normally I won't order anything Billy Mays was connected with, but these extractors look like they really do the job. I'll give them an AK review when I see how well they do on this job.

I thought we had been using the mj15032/33's.....

I'm talking about Driver transistors being 28/29s or 30/31s; not as the outputs transistors. If the 32/33s will work as the drivers, I would be glad to use them as I am already in possession of the 32/33s.

It's nice to be able to read these things without worrying about doing the work myself...

All this info should be helpful down the road to lots of folks.
Appreciate the effort you are putting in.

Good luck and keep us posted.

I want to be able to give back to AK at least a little something. I am not a guru and quite honestly this project has me scared sh**less, but, if I can do this and accomplish my goal to restore this great old amp, then others should be able to do likewise. :thmbsp:

I tested the transistors Q2 > 4 on the right AWH-052 board and they show to be alright. I'll try the other side later in the day. BTW, thanks for the comments. For a while, I thought it was just me and Mark all by ourselves. This is lonely, scary work. Speak out AK.

markthefixer said:

I thought we had been using the mj15032/33's.....

Click to expand...

KingBubba said:

I'm talking about Driver transistors being 28/29s or 30/31s; not as the outputs transistors. If the 32/33s will work as the drivers, I would be glad to use them as I am already in possession of the 32/33s.

Click to expand...

There's an old adage about there being no limit as to how SOFTLY you can hit something with a big hammer......

A slight detraction might be a difference in the gain of a 3 amp transistor versus an 8 amp transistor at lower currents in both of them.
Another would be that it might not blow quick enough on a fault, and pass the load back into the circuit...

Quite frankly I would be more tossed up about using a slower 4 mhz 1.5amp fairchild ksa transistor versus a 30mhz 8amp onsemi. The original being the 2sd381 at 40 mhz and 1.5 amps.

In other words, you can use it...

Thanks for the reply Mark. It's nice to know the 32/33s will not be a wasted purchase after all, not that they were all that much. Actually I think I got them as a free sample from OnSemi because they were, at the time, the "G" version no lead introduction.

I'm setting up now to test the other AWH-052 board's transistors. I'm a little puzzled how to test the 2SA798, which is a double transistor. It has five pins and I am not sure of the configuration. All the datasheets are Japanese. I'll have to figure it out from the diagram on the foil side layout in the SM. I get the feeling these remaining transistors will be okay.

I'm going to pick up the needed resistors at a local parts house tomorrow. I wanted to ask if there was any benefit, or problem, to replacing the 1/4 watt carbon resistors, that are bad, with 1/2 watt resistors. My only thought was that the wattage rating may be part of a protection scheme by the designers.

I have to say that I am surprised that more people have not visited/posted in this thread. It may be that not to many AKers have the 9500II. All comments are welcomed here. I'll keep this thread updated as I progress.

KingBubba said:

Thanks for the reply Mark. It's nice to know the 32/33s will not be a wasted purchase after all, not that they were all that much. Actually I think I got them as a free sample from OnSemi because they were, at the time, the "G" version no lead introduction.

I'm setting up now to test the other AWH-052 board's transistors. I'm a little puzzled how to test the 2SA798, which is a double transistor. It has five pins and I am not sure of the configuration. All the datasheets are Japanese. I'll have to figure it out from the diagram on the foil side layout in the SM. I get the feeling these remaining transistors will be okay.

I'm going to pick up the needed resistors at a local parts house tomorrow. I wanted to ask if there was any benefit, or problem, to replacing the 1/4 watt carbon resistors, that are bad, with 1/2 watt resistors. My only thought was that the wattage rating may be part of a protection scheme by the designers.

I have to say that I am surprised that more people have not visited/posted in this thread. It may be that not to many AKers have the 9500II. All comments are welcomed here. I'll keep this thread updated as I progress.

Click to expand...

2sa798 bcecb
test as two transistors, one at a time, sharing the emitter lead, bases are the furthest pins out, shared emitters pin in the center.

Yes the wattage is part of the protection scheme, ESPECIALLY if it has a triangle symbol with an exclamation point inside.

I have completed the removal and checking of transistor Q 1 > 4 on both AWH-052 boards. They all tested good, including the 2SA798, which, being a double transistor, threw me for a loop at first. The five pins of the 798 are simple. The center pin is the common emitter with the outside pins being the separate bases; so the pin configuration is B-C-E-C-B. It is a double PNP transistor. I have used Echowars transistor thread many times before and used it again tonight. Not doing this for a living, I forget which pin tests which. I also forget what the transistor diagrams mean so this thread is a great refresher.

http://audiokarma.org/forums/showthread.php?t=43186

Thanks to Echowars, Markthefixer and Shimniok for their diagrams and worksheets. It has helped me a lot.:thmbsp:

The driver transistors are attached, as is usual, to a small heat sink. I was surprised to find that there was no mica being used. Is this normal? Is this recommended? Should I add some mica to this setup? I understand that the heat sink is electrically separated from everything, but I am uncertain as to how I should proceed here.

Tomorrow I will get the replacement resistors and Bourne pots, replace the original drivers with the OnSemi MJE15032/33s and resolder the AWH-052 cards, feeling confident that the issues with those cards are all dealt with. If I have time, I will put in my mouser cap and transistor orders. I'm tempted to separate all the cards and fire this thing up to see if it will come out of protection. Mark, you brought up the need to make a jumper in order for the amp to come out of protection. If you remember what you suggested could you please elaborate. I reread my old thread, posts 55 >57, but I am not exactly clear on what you meant for me to do. It sounded like I need to replace R28 with a piece of wire as a jumper. Before I do that I would like a little bit of reassurance that that was the real plan. Thanks for your time in this regard. As always, I will keep this thread up to date as I progress, or fail.

I wouldn't power it up until everything was back together.

The temporary short across R28 was to power up with no power output transistors in place. BUT NO SIGNAL, it's just to troubleshoot - and at this point with all you have done it's more of a risk than a step forward to do it.