Pioneer SX 939 Repair and Restore

[Spenser]

Report:

Not out of the woods on the power amp board.

You know I wanted to get the 1451s out of there as a high priority -- distaster prevention. Was nervous about running it any length of time until that was done. Then circle back and start checking things.

Here is the concern now. Idle (pin 10) was high. Adjusting the pot gets it no lower than 133mV. Other channel was close to zero anyway and responds well enough to adjustment of its pot .

Bias both channels fine

Still has the old pots.

Kicking myself did not take any in-circuit component readings while I had the board loose.

Have not yet checked power amp voltages.

As far as other transistors, none were touched during replacement of electrolytic capacitor and replacement of the pre-driver transistors. This receiver has the 2SA763s for the matched pairs not the 2SA726s. My understanding is the 2SA763s are relatively bullet proof (did see carbon on their legs and wonder if this can creep up into the cases). Thought ALL the other transistors were relatively reliable ones, too.

I had a feeling I would be going back in. On the up side, could be homing in on something here.

Questions:

Most probable cause(s) idle problem as described above?

Edit: The Pioneer OCL Amplifier Service Tech Guide talks about disconnecting the output transistors while taking power amplifier voltage measurements. How do you guys feel about this? Under what circumstances how highly recommended is it to actually disconnect the output transistors when taking power amplifier voltage readings?

Thanks.

 

[Spenser]

Next Steps:

I'm planning on doing the power supply and then coming back to the power amplifier board.

I'm also planning on getting the tech guides printed out and trying to discipline myself to do read and learn them and my Malvino and Boylestad.

Thanks.

 

[Spenser]

I thought it might be of interest for readers to be able to see X-Ray Tony B's thinking with respect to diffferent pre-driver transistors found in some models (see Post #33), so I've reprinted it below. Good stuff. Enjoy.

Hi Matt,

Sorry for the delay, as I am really busy these days! Finally getting a chance to answer some of the mountain of e-mails. It’s not uncommon for Pioneer (as well as other manufacturers) to change parts during production. Sometimes, it’s because they are using a different parts vendor and other times it’s to improve or correct an issue with the circuit. The older SX-series (like the SX-939) are very picky about transistors. Just because another transistor substitute has the same specs on paper, it still may not work in the circuit. I’ve found that sometimes the transistors that work best as a substitute may have very different specs on paper. Sometimes you have to try several different types until you get one that works properly. I wish there was an easier answer, but that has been my experience. Also, parts of the circuit may use matched pairs of transistors. Putting unmatched pairs in those parts of the circuit will cause big problems with bias and/or offset. Of all the testers out there, I’ve found that my little homemade matching circuit and an oscilloscope work the best for small signal transistors. Anyway, I hope that answers your question.

Thanks for watching!

Tony

 

[rcs16]

Edit: The Pioneer OCL Amplifier Service Tech Guide talks about disconnecting the output transistors while taking power amplifier voltage measurements. How do you guys feel about this? Under what circumstances how highly recommended is it to actually disconnect the output transistors when taking power amplifier voltage readings?

Be very careful with that advise, if you break the DC feedback loop the amp will swing/clamp to one of the supply rails and the voltage measurements will be useless. So this depends on the amp design topology if the drivers feedback to the diff pair.
rather than trying to figure out the circuit design using Malvino and Boylestad,.try simulating the design in ltspice, your results will better reflect the actual circuit and the substitute bjts that you decide to use. I have supplied simulations in my stk-0050 thread that you can use to get started, PM me once you get the circuit entered or I can email you what you need and i can check it over for you. I have the QX-9900 simulated, it is basically the same design but more power.

 

[Spenser]

Be very careful with that advise, if you break the DC feedback loop the amp will swing/clamp to one of the supply rails and the voltage measurements will be useless. So this depends on the amp design topology if the drivers feedback to the diff pair.

OK. This is good to know. Thank you!

I like the idea of using the free computer aided design modeling software as a learning tool. Good grief -- ltspice -- first I'd heard of it. Took a quick look at a tutorial. Holy cow. Incredible.

Could you please post a link to your stk-0050 thread? Thanks.

I'm going to give it a go. It will take me some time to work through. Excited.

No electronics background. Chemistry, physics and biolgical sciences. No computer science. They were still teaching punch card programming back then (late '70s early '80s).

 

[Spenser]

Report:

Got the power supply substitutions done except for the filter capacitors. All the old components tested very good or reasonably good coming out. Resistors all stayed in; all good or not suspiciously off.

Nevertheless -- still have the DC offset issue -- now seem to have even more millivolts DC sitting on the one channel -- Pin 10. Was 133 before and now is 175 (was gradually subsiding but saw to reason to let it cook). Other channel idle is still good.

Next Steps:

I will check the power supply voltages for good measure; then go back to the power amplifier board and look there

 

[Oldsansui441]

I will check the power supply voltages for good measure; then go back to the power amplifier board and look there.

Sounds like a good plan.

 

[Spenser]

Detour / Sidecar: What do you think of this "Small Signal Transistor Vbe Matching Tool" designed by X Ray Tony B?

I picked up on this a few weeks ago in one of his SX-939 videos. Thought it might be worth checking out.

Pictured here is X Ray Tony B's draft schematic for the device I wrote on an index card.




The potentiometer is to balance the resistors which is done whilst shorting out the transistor test sockets to the volt meter test lead connectors.

When you are ready to go, you insert one transistor as your control, and then start going through possible matches close to the control using the other socket.

He put up a couple of videos in February of a prototype he has been using that he cobbled together on a breadboard. Go to his Youtube, select videos, and look for "inexpensive method for matching transistors" video and followup video. He likes this better than a real deal curve tracer he built some time ago from a kit.

It would be nice to have a PCB. One of his watchers sent a comment there are companies such as jlcPCB.com that will take a file you send them and make a run of minimum 10 for $2-a-board free shipping (can that be right!?!?). The setup for the transistor sockets or hooks would need to be determined. The termporary sockets he used in his prototype would soon become troublesome from wear.

 

[Oldsansui441]

Looks interesting, would be a handy circuit for matching input pairs esp on amplifiers without offset adjustment.

 

[rcs16]

test jig is okay if you are going to do bunches of them. all you need to do is measure bjt Vbe (DMM in diode test mode) and match Vbe to <2mV and you will be just fine. use zif sockets for longer life

As far as learning and using ltspice, I learnt using Bob Cordell's Audio power amp design book and peoples design examples on diyaudio. Bob has tutorials, models, and actual designs to learn from. the book is out of print now, but the 2nd edition is in the works. there is also a thread on diyaudio about using ltspice.
There is the assumption that you learn or understand about some diode/transistor and basic electronics theory to make any sense out of ltspice simulations.

stk-0050 thread
http://audiokarma.org/forums/index....x-780-and-others.721181/page-24#post-11981609

I meant to say the sx-939 is higher power than the qx-9900 but the amp design topology is the same boot-strapped voltage amplifier design

 

[Spenser]

Thanks for everything in Post #50. I've heard of the Bob Cordell book. I'm going to take a look at it and his website. I will look at your stk-0050 thread, too.

I've begun moving around the top of the power amp board taking some in-circuit measurements. So far diodes only, and these all seem good.

Started to try and get some voltage readings around the first stage differential amplfier. I'm not 100% clear on where the test points are supposed to be. Thinking might be more productive to try trouble shooting protocol items #36-42 in Pioneer "Trouble Shooting Method O.C.L. Main Amplifier Circuits."

http://www.kallhovde.com/pioneer/ocl-amp-techguide.pdf

 

[Spenser]

I'm back. How is everyone?

Bad DC Offset Idle Current on One Channel -- Continued

To recapitulate, I've got a reading of 175 mV or more on one channel with the pot cranked all the way. The other channel is good.

I scouted resistors in proximity differential pair. All good. Only thing -- probably not really noteworthy -- saw that R19 and R20 schematic specification 100K but color coding actual resistors is 390K. In-circuit readings of 384K and 387K respectively.

OK anyway so here are voltage readings and then a couple QUESTIONS; please take a look and let me know if you see anything here; not positive got the right test points; going to sign off now (eyes crossing), and check back tomorrow THANKS:

R9 .28
R10 .28

R15 .13
R16 -.02

R13 -.41
R14 -.45

R17 -3.9
R18 -3.9

Q5 B 42.7 E 43.4 C 1.329
Q6 B 42.7 E 43.4 C 1.136

Q7 C -1.006
Q8 C -1.166

Should I go ahead and go in and replace the long tailed pairs? KSA 992 is the proper substitution right? This receiver has the 2SA763s not the 2SA726s. My understanding is the 2SA763s are relatively bullet proof but I do see carbon creep.

Opinions on whether to replace any diodes in the power amplifier while I have the board free? If so which ones?

Sorry to be jumping around here. Back to the power supply for a moment. Opinions on whether there are diodes that should be replaced in the power supply as a matter of routine? I did not replace any there while I had it apart. Sigh.

 

[avionic]

Random entertaining sidenote: yesterday Rico126 was shy one (1) KSC2073 for his SX-737 currently underway; we were both too busy to meet; so I left him two transistors under a rounded rock at the edge of a hardware store parking lot; masking tape arrow at edge of asphalt pointing to rock; Rico reported transistors successfully retrieved.

Sounds like a drug deal..LOL

 

[Spenser]

Glad you saw that. Hi hi.

 

[Spenser]

Voltages at pins on the main amplifier board:

Pin 7 45.5
Pin 4 33.9
Pin 2 -42
Pin 14 48.1
Pin 15 -48.1

Observation: these readings are a bit off from schematic-identified-specifications. It could be that design specs changed and were not updated on manual -or- also could be that something is awry upstream in the power supply.

An aside: house line AC voltages vary. The other day it was 125v AC. At the time of the above readings, I had 118v AC going into the receiver.

Before I show a picture of my "variac" for fun, I thank rcs16 for getting me to add Bob Cordell's book -- Designing Audio Power Amplfiers -- to my reading. What an amazing book. Out-of-print -- but, he is working on another -- according to rcs16.

Something for us all to ponder: a person as obviously brilliant as Bob Cordell begins his acknowledgements, "My Lord and Saviour Jesus Christ has given me the peace and guidance that allowed me to complete this undertaking."

Variac

 

[Spenser]

I'm working on my parts list for "Phase II" (equalizer, control, muting, tuner).

I will post my Phase II list once I've got it put together; see if anyone wants to take a peek.

Don't worry I won't feel neglected if there aren't any takers. We've had a bunch of Pioneer SX-939 threads as you know.

 

[Spenser]

We always verify what's actually in a unit when we go about putting together a parts list. Pioneer and others made engineering changes in succeeding production runs so we never know what we have until we check.

You know the control board on these babies has to come out. A straightforward removal: it easily flips right out of there with plenty of wire slack; but good grief -- this is my first SX-939 -- so I did not know the control board was put together like this. Look at the "brick ____ house" cover arrangement. Also, see picture with the cover removed.

Wow the build quality. Oldsansui441 said it. The 939 is a beast.

Confirmed equalizer and control boards contain transistors that are known, failure prone, so we will replace with better ones.

Specifically, we have some 2SA725; 2SC1312; and 2SC1313.

http://audiokarma.org/forums/index....ilure-prone-whatever-and-replacements.731653/

 

[rcs16]

If voltages are within 5-10% of printed you are usually fine, you are working with 5% resistors and AC line voltages vary as well.

 

[Oldsansui441]

Looks like you are making good progress, yes more KSA992's and KSC1845's needed .

 

[Spenser]

Thank you for replies rcs16 and Oldsansui441.

The following is pretty unremarkable. Will mention only a couple things.

One, seems like multi-turn might be a better option than single-turn for the varialble resistors in the power amplifier. So I chose those. Two, control amplifier C49-52 this particular unit has 47uF instead of 100uF (who knows), so I went with 47uF substitutions.

DRAFT === DRAFT === DRAFT === DRAFT === DRAFT === DRAFT

THIS LIST IS UNDER DEVELOPMENT -- DO NOT USE

Edit: changed signal path caps in control amplifier assembly to KEMET film.
Edit (10-30-18): corrected errors in parts list,


Pioneer SX-939 Phase II Parts List


Power Amplifier (plan to circle back on the DC offset idle current issue still remaining in one channel; see first if we have a quirky pot; also take a look at the differential pairs)

VR1-2 10K Trimmer 652-3296Y-1-103LF (2) multi-turn

VR2-3 100Ohm Trimmer 652-3296Y-1-101LF (2) multi-turn

Q1-4 2SA763 512-KSA992FBU (100)


Equalizer Amplifier


C1-2 cssa 2.2uF 25V 505-MKS22.2/50/5 50V (2)

C5-6 ceanl 470uF 6.3V* 647-UKL0J471MPD 6.3V (2)

C9-10 ceanl 100uF 10V* 647-UKL1E101KPDANATA 25V (2)

C11-12 ceanl 10uF 25V* 647-UKL1E100MDDANA 25V (2)

C19-20 ceanl 100uF 25V 647-UKL1E101KPDANATA 25V (2)

C21 ceanl 220uF 16V 647-UKL1C221KPDANA 16V


Q1-4 2SA725* 512-KSA992FBU (4) has 2SA763 in these spots not 2SA725

Q5-8 2SC1313 512-KSC1845EBU (4)


Control Amplifier


C1-2 cssa .33uF 25V 80-MMK5334M50J01L4 50V (2)

C7-8 cea 100uF 10V 647-UPW1E101MED 25V (2)

C13-14 cssa 2.2uF 25V 80-R82CC4220Z370J 50V (2)

C15-16 cea 4.7uF 35V* 647-UPW1H4R7MDD 50V (2)

C23-24 ceanl 4.7uF 25V 80-R82CC4470Z330K 50V (2)

C25-26 cea 10uF 16V 647-UPW1E100MDD6 25V (2)

C29-30 cea 100uF 10V 647-UPW1E101MED 25V (2)

C31-32 cea 4.7uF 35V* 647-UPW1H4R7MDD 50V (2)

C39-40 ceanl 4.7uF 25V 80-R82CC4470Z330K 50V (2)

C41-42 cea 10uF 16V 647-UPW1E100MDD6 25V (2)

C45-46 cea 100uF 10V 647-UPW1E101MED 25V (2)

C47-48 cssa 2.2uF 25V 80-R82CC4220Z370J 50V (2)

C49-52 cea 47uF* 16V 647-UPW1E470MDD 25V (4) manual specs 100uF but 47uF actually what was in there so went with that


Q1-2 2SA725 512-KSA992FBU

Q3-8 2SC1312 512-KSC1845EBU


Muting Assembly


C1-2 cea 4.7uF 25V 647-UPW1V4R7MDD 35V (2)

C3 cssa 1uF 25V 647-UKL2A10KDD 100V (no 50V in stock)


Tuner


C22 cea 4.7uF 25V 647-UPW1V4R7MDD 35V

C24 cea 10uF 16V 647-UPW1E100MDD6 25V

C25 cea .47uF 50V 647-UKL1HR47KDDANA 50V

C26 cea 10uF 16V 647-UPW1E100MDD6 25V

C31 cea 1uF 50V 647-UKL2A10KDD 100V no 50V in stock

C33 cea 100uF 16V 647-UKL1E101KPDANATA 25V

C34 cea 47uF 16V 647-UPW1E470MDD 25V

C37 cssa .33uF 25V 647-UKL2AR33KDD 100V no 50V in stock

C38 cssa 1uF 10V 647-UKL2A10KDD 100V no 50V in stock

C39 cssa .47uF 25V 647-UKL1HR47KDDANA 50V

C42 cea 100uF 16V 647-UKL1E101KPDANATA 25V

C43-44 cea 2.2uF 50V* 647-UKL1H2R2KDDANA 50V (2)

C45-46 cea .47uF 50V 647-UKL1HR47KDDANA 50V

C52 cea 4.7uF 25V 647-UPW1V4R7MDD 35V

C53 cea 4.7uF 10V* 647-UPW1V4R7MDD 35V
EDIT: TYPO C53 should be 47uF!!! I did not notice my error until I was doing the recap on the tuner board. Sorry!!!

C58 cea 10uF 16V 647-UPW1E100MDD6 25V

C60 cssa .1uF 25V 647-UKL1H0R1KDDANA 50V

C63 cea 220uF 16V 647-UKL1C221KPDANA 16V


Lamp Board


220uF 16V 647-UPW1C221MPD 16V


*different actual in unit than specs in manual


____________________________________________________________________

I don't kinow that any of the following are really needed. Would be interested to hear opinions. Edit: are there any diodes you would recommend replacing; how 'bout resistors in the power amplifier; any that get a lot of wear that you would naturally replace as a matter of due course while you have it apart? I think Q9-12 and the outputs in these are pretty tough and probably ok. Judging from the healthy condition of the power supply board I doubt this unit has seen a lot of hard duty over the years.

Power Amplifier Optional / Backup Parts


Q9-10 2SD381 512-KSC2073TU (2)

Q11-12 2SB536 512-KSA940TU (2)

D1-4, 7-8 1S2473 512-1N4148 (6)


Power Supply Optional / Backup Parts


D1-4 SIB01-04 863-1N4004RLG (4)

D5-8 SR3AM8 863-1N5404RLG (4)

D9-11 1S1885 863-1N4004RLG (4)

D12-13 WZ-130 512-1N5243B (2)

D14 WZ-140 512-1N5244B


Protection Circuit Optional / Backup Parts


D1-8 1S2472 512-1N4148 (8)


Optional / Backup Parts -- Outputs


Q1-2 2SD388 863-MJ21194G

Q3-6 2SB541 863-MJ21193G

Q7-8 2SD388 863-MJ21194G


Insulators (8) 534-4662