Hello Kind AK People,

I'm reopening my SX-850 repair project to the public (AK) after making exactly zero progress on my own. Retro Stereo was exceedingly kind and lent me his Service Manual, which I have been hording for much too long now. So sorry Retro!!!

Let me bring you up to speed. I bought this SX-850 for parts for someone on this board who due to miscommunication decided not to purchase it from me. Consequently, I decided (perhaps over-ambitiously) to make it a repair project. The receiver is in nice cosmetic condition, and really deserves to be set right.

The current consenus is that due to excessive heat a regulator cap on the POWER SUPPLY ASSEMBLY (AWR-101C) went bad and in so doing took a transistor and zener diode with it. I can confirm that one of the 80v 470uf caps is shot. It's bulging and leaked fluid on the board. The one that is shot is symbol C11, part number ACH-038 (Electrolytic). The other cap looks like it's ok, but I imagine it would make sense to replace both at once so that they match and to save trouble down the road.

After that, I'm a little lost. There are 10 (5 per side?) transistors on this board, and I do not know how to determine which are bad. Further there are two zener diodes with different part numbers (WZ-140 and WZ-130), so it seems that they are not specific to one channel or the other. I believe I have located them on the board (they are very small - correct?), but my multimeter is old and does not have a diode test function. This leads me to my first two official questions:

Q1. How do I determine which transistors are good and which are bad?
Q2. How do I determine which zener diodes are good and which are bad?

Here is my roadmap to fix this unit:


1. Locate the following replacement components:

a. 2 x 80v 470 caps (or better? Panasonic FC-series)
b. 1 x WZ-140 (zener diode)
c. 1 x WZ-130 (zener diode)
d. An unknown number / variety of transistors.
e. 2 x 1A 125v fuses

2. Remove the damaged components.
3. Install the replacement (upgraded?) components.
4. Improve heat dissipation (add heat sink to???)
5. Enjoy!


Obviously this is an incomplete list. I hope that certain knowledgable members of this board will help me fill it out. I've posted several pictures to help orient everyone. I apologize for the quality of the schematic, but Retro's SM has a very weak binding. Be sure to click to see the fullsize versions.

SX-850 Pictures (http://www.ambitious.ca/~stewey/sx850/)

I know you sent a PM, but I'll answer here..


The zeners are standard stuff...a WZ-140 is a 500mW 14V zener, and a WZ-130 is a 500mW 13V zener.

With a diode test, a zener looks like any other diode.... 0.7V or so in one direction, open in the other. A zener works in avalanche mode by breaking down at the specified voltage, which your meter can't do. However, if it measures like a normal diode, it is probably OK (but in cases like this I replace them anyway) Do not be tempted to step up to a 1W zener...not enough current for proper regulation, and there is no such thing as a 1W 14V zener anyway.

If you are going to troubleshoot this, you absolutely have to have a meter that has a diode test mode. I'm not a big fan of RatShack stuff, but this ( http://www.radioshack.com/product.asp?catalog%5Fname=CTLG&category%5Fname=CTLG%5F011%5F008%5F002%5F000&product%5Fid=22%2D811) meter is excellent for the price. They show it as no longer available online, but I was at RatShack to get a few resistors tonight and it was there.

Get that cap out of there...the electrolytic is corrosive and will ruin other parts.

Once you are set with a proper meter, you need to learn how to test transistors. Come back when you are ready...

Thank you.

Assuming that they do not have that model in stock (it's not even listed on the .ca website), would this model do?

22-813 (http://www.radioshack.com/product.asp?catalog%5Fname=CTLG&category%5Fname=CTLG%5F011%5F008%5F002%5F000&product%5Fid=22%2D813)

Aside from being a 29 range model, it doesn't seem to measure capacitance, frequency or 'detect electric fields' as the 42 range model does. Otherwise it appears to have the same features.

Thanks!

Looks fine...

Get a pair of these (http://www.radioshack.com/product.asp?catalog%5Fname=CTLG&category%5Fname=CTLG%5F011%5F008%5F001%5F000&product%5Fid=270%2D334) too...

Picked up the 29 range model, (no 42 range model in sight). I also bought the test hooks, a 9v battery and LED. I'll have to head to a real electronics store to get the resistor, but I think I'm almost ready to test the transistors.

BTW: those mini-hooks are great! I used them in my CENG 355 lab, but I wouldn't have thought to buy them if you hadn't mentioned it. Thanks!

Stew.

Is there a way to access the foil side of this board and desolder/remove components from it without disconnecting it completely? There are over 30 different wires connecting to this board and I'm not eager to label & disconnect all of them if it is at all possible to avoid.

Also, I tested both of the zener diodes in circuit and they appear shorted in each direction. Do I have to take them out of circuit to get an accurate reading?

Please advise.

I'm not specifically familiar with the layout of the 850, but you should be able to access the foil side by removing the bottom. If not, then somehow, someway, there has to be a method of getting to it by disassembling surrounding components. Most of these old Pioneers had wires connected with wirewrap...taking them off is not an option, and I've never worked on a unit where it was necessary (but on a lot of 'em it was tempting, as they were a PITA to fool with).


Many components will not test properly in circuit. Removing them for test is just part of what comes with the territory.

Originally posted by EchoWars
I'm not specifically familiar with the layout of the 850, but you should be able to access the foil side by removing the bottom.

That's what puzzled me about this. The power supply board is mounted upside down on the bottom the unit, so I couldn't access the foil side from either top or bottom. Still, I persisted and was able to gently flip the board up on its side by pulling some wires out of the harnesses and releasing the clips that hold the board in place.

I pulled C11 (the 80v 470uf cap) out. Wow, that was exciting :) ...seriously, I was almost certain I'd melt something in the process, but I didn't!!! I also pulled the zener diodes out of circuit (free'd one side) and tested them again. They both checked out this time. Each one measured about .68v in one direction and open in the other. Should I still replace them?

Can you recommend which transistor(s) to test first? How do I know which posts correspond to base, emmitter and collector? How do I know if they are PNP or NPN transistors?

Sorry about all the questions; thanks for all your help! I think I made some major progress today!

Stew.

Can you recommend which transistor(s) to test first? What voltage(s) was missing? What fuse was blowing, if any? Look at the right side of the power supply print...there are well over half a dozen voltages generated by this supply...which ones are missing or problematic?How do I know which posts correspond to base, emmitter and collector? In the service manual there should be a page, somewhere near the prints, showing the physical layout of the devices. Below is a page from a Sansui manual...the Pioneer should have something similar.How do I know if they are PNP or NPN transistors?2SAxxx and 2SBxxx are PNP devices, 2SCxxx and 2SDxxx are NPN devices.

What voltage(s) was missing? What fuse was blowing, if any?

The 3rd picture in my first post shows the two 1A 125v fuses that are blown (referred to as FU2 and FU3). I'm trying to trace these fuses back to transistors as it seems to me that would indicate ideal places to test. FU2 is wired in parallel with a diode and cap to the 'collector' of Q2 (a PNP transistor). C11 (the bad 80v cap) is also tied to the collector of Q2. FU3 takes a path that is precisely the same as FU2 (at least back as far as Q2).

I should mention that FU2 & 3 take another path also through a diode and cap in parallel, but if I'm reading the diagram correctly it is biased in a direction that is opposite to the flow of current.

I THINK (emphasis on think), that what I'm describing is a bridge rectifier, but I don't know for certain. It seems to me that this (Q2) would be a good transistor to test first. Does this make sense to you? ...or is all that just crazy talk?

Look at the right side of the power supply print...there are well over half a dozen voltages generated by this supply...which ones are missing or problematic?

I don't see how I can test these now that I've removed the bad cap... or is that the idea? Find out which voltages are missing now that the cap is gone?

In the service manual there should be a page, somewhere near the prints, showing the physical layout of the devices. Below is a page from a Sansui manual...the Pioneer should have something similar. 2SAxxx and 2SBxxx are PNP devices, 2SCxxx and 2SDxxx are NPN devices.

The SM I have doesn't have any page like this (not that I can find and I've looked several times), but I think I can read the transistor symbol properly now. Question: does the direction of the arrow indicate whether a transistor is PNP or NPN, or is that a coincidence?

Q2 is a big transistor screwed into the largest heatsink on the board (with lotsa heat transfer paste). It is labelled:

SANYO
B507D
6E

The SM I have doesn't have any page like this (not that I can find and I've looked several times) Well that kinda sucks...

Anyway, yes, an arrow pointing towards the base of the transistor indicates a PNP, and one pointing away indicates NPN. Also, most Jap transistors won't have the whole part number on them... a 2SB507 might be labled 'B507'. It's still a PNP device, the 'B' gives it away.

Until you have the cap(s) replaced, don't power it up. Additionally, replace them all in that regulator. It's pretty inexpensive, so do it. Put in new zeners too. Use good high-temp caps, like the Panasonic FC caps. Do not screw up the direction that they go in...electrolytic caps can make really cool mini-hand grenades when reverse-biased.

Yes, those fuses go to a diode bridge. Since the bridge supplies both + and - voltages, there are several possibilities for dead transistors. The most likely candidates are Q1, Q2, Q3, and Q4, with Q1 and Q2 being the serious suspects.

Testing a transistor:

For rookies, I recommend removing the transistor. Other circuits will interfere with your readings, and unless you've done this a lot, it may confuse you. So just take 'em out...good soldering practice.

1. Identify the device (NPN or PNP).

2. Identify the leads (base, emitter, collector). If not sure, check where you removed it. If the collector is supposed to go to 'X' component, well then, see which lead of the transistor went in that place. The vast majority of Japanese transistors are either BCE layout, or ECB layout...collector almost always is the center lead. If you are not sure, follow the leads on the board to see what runs where. For TO-3 power transistors, the layout is always the same as the pic below.

3. Set meter to diode test. In the pic below, you can see the equivelent of a NPN and PNP (for testing only...) With an NPN device, put posiive lead on the base, negative on the emitter. The meter will show the voltage required to overcome the diode depletion layer, generally .4 to .6V. Now move the negative lead to the collector. Again, you should read something like .4 to .6V. Now move the positive lead to the emitter. It should be an open circuit. Now reverse the leads...positive on the collector, negative on the emitter. Same deal, it should read as an open circuit. Anything else, and the transistor is toast (As an aside, dead transistors usually short from the collector to the emitter).

4. For PNP transistors, reverse the leads as described above (negative on the base, positive on the emitter etc, etc..) Again, read .4 to .6V from B to E, and B to C, open circuit from C to E.

Obviously the two hardest parts are identifying which lead is which, and removing the components. The former takes experiance and the latter requires a good solder sucker.

BTW, these tests are NOT foolproof. Some transistors will test fine out of the circuit, but fail once any appreciable voltage is applied (but are rare in comparison to the normal 'just plain dead'). Such is life...

Thank you. I will pick up a 'solder sucker' tomorrow, do a few tests and report back.

Do not screw up the direction that they go in...electrolytic caps can make really cool mini-hand grenades when reverse-biased.

As if this wasn't stressful enough already! Now I have to worry about killing myself as well as the amp.

:para:

This (http://www.hvwtech.com/pages/products_view.asp?ProductID=444) is the solder sucker you want. Anything else is a waste of time and money.

Also, when desoldering (and soldering), don't heat up the board traces any more than necessary, or they will come loose from the epoxy. Lifted traces are a PITA to deal with, so get in there, get it soldered (or desoldered), and get out.

Hi EchoWars,

Picked up that Soldapullt today. Worked great once I got the hang of it. So, I pulled out Q1 & Q2. They both checked out. Here's what I did for you to double check if you are so inclined.

Q1 (NPN):
a. Pos to base, Neg to emitter = ~.5v
b. Pos to base, Neg to collector = ~.5v
c. Pos to emitter, Neg to collector = OPEN
d. Pos to collector, Neg to emitter = OPEN

Q2 (PNP):
a. Neg to base, Pos to emitter = ~.5v
b. Neg to base, Pos to collector = ~.5v
c. Neg to emitter, Pos to collector = OPEN
d. Neg to collector, Pos to emitter = OPEN

Hmmm... so that means the zener diodes checked out, and Q1/2 checked out. I suppose I should move on to Q3 & Q4 then?

What do you think?
Stew.

Yep....keep chuggin'.

Back at it tonight. I pulled Q3 and Q4 which both checked out. Then I pulled Q8 & Q9 simply because they're big and heatsinked. I figured they might be candidates. Anyway, they checked out too. I would have kept going, but this work is quite tedious due to my lack of experience desoldering components.

There are only 4 more trannies to go on this board. Unless I hear an alternate suggestion from you I will pull them tomorrow and test.

If they all check out, where do I go from there?

Thanks for your continued support on this project!
Stew.

Put new caps in, install new fuses, fire it up and cross yer fingers...

That sounds much more exciting!!! Ok, time to build a shopping list. Should I get new transistors for the ones I removed? Should I replace EVERY cap, even the small ones? There are 23 in total. Gotta do the zener diodes too - right?

Damn that's a lot of soldering / desoldering.

What about improved heat dissipation?

If the transistors test OK, put 'em back in.

The caps are cheap. replace 'em. The zeners too.

Well, after a bit of a break from this project, I thought I should attempt to make some progress again, before I forget everything you have taught me!

I looked locally for Panasonic FC caps, and no one in Victoria carries them. I was referred to "digikey". I can supposedly order these caps from them off their website. Here's the problem: I do not know enough about what I'm looking for to pick the right capacitors.

The capacitors part list in the Service Manual is as follows:

http://www.ambitious.ca/~stewey/sx850parts.gif

As I quickly found out, there are many different kinds of "ceramic" capacitors, which style do I choose? Same with "electrolytic" capacitors. What if Panasonic FC caps aren't available in a given value, should I pick the next largest available FC cap, or switch to a different series?

Please help!
Thanks, Stew.

BTW, here's a link to the digikey capacitors parts list:

http://dkc1.digikey.com/CA/PDF/C042/P1.html

Leave the ceramic caps alone. They rarely fail.

All the rest of the values are available in FC caps except the 470µf 80V. Panasonic does not make a 80V cap...they go from 63V to 100V. The problem with the 470µf 100V cap is that it's long, like 40mm.

Check page 785 here: http://dkc3.digikey.com/PDF/T042/0769-0797.pdf

These 80V caps are often underrated in Pioneer gear...often there is over 80V on them (which is why it's nice to put in a 100V cap). So check out the TSHA caps and see if you can use one. The beginning of the section shows the measurements of the pins (they are all 10mm from center of '+' pin to center of '-' pin), but do require a slightly larger PC board hole than normal. If you can, use part #P6695-ND, which is a 560µf 100V cap that measures 22mm dia by 25mm tall. It should fit...

The power supply board is mounted upside down on the bottom the unit, so I couldn't access the foil side from either top or bottom. Still, I persisted and was able to gently flip the board up on its side by pulling some wires out of the harnesses and releasing the clips that hold the board in place.

I had the same problem and figured out the same solution for my board. One note, though, about those plastic clips.

In the design of the amps, these offsets were not intended to be manipulated much after assembly. Furthermore, due to heat and time, the plastics can become amazingly brittle. Remove and replace this board from those standoffs as little as possible - and better yet, have candidates for replacement on hand. I'm not sure what I would use to replace them, but just something important to be aware of.

Hope the restoration is going well!

Thanks for the advice!

I would say the restoration is going well, albeit much slower than I had originally hoped. I'm finding it hard to work on this project lately. My girlfriend and I are trying to buy a condo and that seems to be consuming all of our spare time. Ah, but that's just an excuse.

I should really get around to ordering all those caps and I know Retro Stereo wants his service manual back. EchoWars has been so generous in extending his expertise that I'm determined not to let this become one of THOSE projects that never gets completed.

Check back in a few months and this unit might just be up and running :)

SX-850 Shopping List:

Now, I certainly don't expect anyone (EW) to go over this list, but if they wanted to, well that would be nice. I'll wait a little while before I order this stuff, so if there are any glaring errors or ommissions please let me know. I don't know how significant the leakage, resistance and ripple are, but I put them in just in case they're important and so that you don't have to go cross-referencing them. Whenever there were two or more variations for a specific v/uf cap, I selected the smaller or mid-size one, unless I felt the size wasn't going to be an issue in which case I selected the one with a longer lifetime. If you feel that this is a poor method of selection, please let me know and I'll re-assess my choices.

TSHA:

2x P6695-ND
100v, 560uf, .7mA leakage, 1.10A ripple @ 120hz/85deg.


FC Series:

1x P10310-ND
35v, 2200uf, 770uA leakage @ 20deg, 0.020ohms @ 100khz/20deg, 2740mA ripple @ 105 deg.

1x P10278-ND
25v, 1000uf, 250uA, 0.038ohms, 1655mA.

1x P10220-ND
10v, 220uf, 22uA, .35ohms, 290mA.

1x P10351-ND
63v, 470uf, 296.1uA, 0.055ohms, 2090mA.

2x P11196-ND
16v, 47uf, 7.52uA, .80ohms, 175mA.

1x P10321-ND
50v, 47uf, 23.5uA, 0.60ohms, 260mA.

1x P11232-ND
35v, 47uf, 16.45uA, 0.35ohms, 290mA.

1x P10343-ND
63v, 100uf, 63.0uA, 0.256ohms, 535mA.

1x P10294-ND
35v, 100uf, 35.0uA, 0.117ohms, 555mA.

If that is the value of the original caps (except for the 100V TSHA), then I don't see a problem. All should be much smaller than the originals, and should fit fine. The only 'if' is getting the TSHA cap in. You need to look at the mounting for the original 470µf 80V cap and see if the holes are 10mm apart. If so, you can carefully drill out the mounting eyelet (if necessary, and it may not be) with a 1/16" drill bit (no larger).

G/L