• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

sx 3900 question

steamwhistle

steamwhistle

AK Subscriber
Subscriber
I opened up an old dead sx 3900 and found the usual problems such as a blown resistor and cracked connections running to the outputs. But I also ran across a lot of add-ons to the board, pictured. So far I haven't run into similar modifications. Why would someone add capacitors and resistors and jumpers like this? Does this have to be undone? Not quite sure how to proceed. Any ideas?
 

Attachments

  • 3900a.jpg
    3900a.jpg
    100 KB · Views: 93
Register to hide this ad
Looks like factory tweaks to me. I see it all the time where I work. I would leave it and address the obvious problems.
 
steamwhistle said:
I opened up an old dead sx 3900 and found the usual problems such as a blown resistor and cracked connections running to the outputs. But I also ran across a lot of add-ons to the board, pictured. So far I haven't run into similar modifications. Why would someone add capacitors and resistors and jumpers like this? Does this have to be undone? Not quite sure how to proceed. Any ideas?
Click to expand...

In all probability, Factory mods to the tuner section - leave them alone. Sometime later we'll talk about what they are and where they are connected, and I'll see what sense can be made of them.

As for fixing a bad 3900 amp - it's DEEP DEEP muck loaded with Gators.

100 percent component testing... no exceptions - no slip ups and probably a complete power transistor socket replacement (all 8) with something a lot more reliable. Mini euro terminal strips from Radio Shack - as documented on a recent thread.

IF any of the outputs are blown - this type of amp is very sensitive to mis-matched output transistor characteristics - a complete set of 4 for each channel affected is called for... I have TRIED it the other way... and the amp is VERY unforgiving. I MATCHED the original salvaged replacements, and things went from bad to worse - and a set of NEW transistors cleaned up many strange things happening - NOTHING else was changed.

The replacements are 2x 2sa1216 & 2x 2sc2922 at $30 (per channel) from ampslab.com
OR make bars to clamp down to-3p packages and use
863-MJL1302AG pnp TO-3P/TO-264 bce 260V 15/30A 200W 45-150 30MHZ $3.50 ea
863-MJL3281AG npn TO-3P/TO-264 bce 260V 15/30A 200W 45-150 30MHZ $3.50 ea
(these are the little brothers of the transistors EW uses on the sx-1980)

I have done it both ways. Including drilling and tapping the heat sinks to properly mount the to-3p transistors.
 
Last edited:
I'm glad to hear the add-ons are probably factory. The output transistors are all over the place in both location and price and range from 2.99 on Ebay to $5 at B&D Enterprises to $15+ at certain well-known sources. I'd think the Ebay ones were counterfeit, but there are no customer complaints I can find. I'll test these first off with a DMM. When I first got this thing years ago, it worked fine for about a half hour, then gradually went on the fritz with display and audio problems. Next morning I turned it on, if I remember correctly, and a little rattail of smoke came out of the grille about where the burned-in-half resistor is today. Now it lights up, but no second click. Will check the outputs now. Thanks for all the info.
 
Okay, I've got two dead output transistors and several cracked hard plastic harnesses that connect to the outputs. Both transistors are on the right channel, if that is the one toward the rear of the case. That's the channel with the exploded resistor. This thing sure isn't built like a 50-series. Like comparing a sixties Renault to a Checker cab. Any thoughts on a next step?
 
You might want to check your speaker relay. I have an SX-3900 on the bench right now - I completely rebuilt the blown channel and couldn't figure out why I still wasn't getting sound even after I confirmed proper operation of the driver and output stages. Turns out the speaker relay contacts had apparently melted down before the protection circuit could disengage them. Hopefully your relay is still good because replacements are almost impossible to find for this receiver...
 
I gave you two perfectly good TRUSTWORTHY output transistor replacements, and adequate if not detailed "marching orders" as well as the benefit of some HARD WON experience.

This is the HARDEST type of Pioneer Amplifier to fix.

The adage about " leading a horse to water but can't make him drink" seems to apply.

You have dead diodes, blown driver transistors, and a bunch of hidden - blown out resistors - which can ONLY be found by 100 percent component testing. This is mostly the way I have to do it.

There ARE lists of the components necessary to do this board posted on the Forum, in the last year and the original protection relay IS still available from Pioneer for about 50 bucks, 60 shipped.

This looks right now like an impending train wreck.

The thing is that the diodes, transistors and resistors - excluding the outputs - are pretty cheap to replace - even if a bit tedious - but you did that pretty well on the 1050, so I hope you pay attention now.


http://www.audiokarma.org/forums/showthread.php?t=183134

http://www.audiokarma.org/forums/showthread.php?t=161866

http://www.audiokarma.org/forums/showthread.php?t=190127

markthefixer said:
This is one of those innocent sounding posts that is going to eat up my entire evening......... sigh......

Also I have the sneaking suspicion that I have already posted much of this somewhere else in the Pioneer forum.

I have ALREADY done the shotgun thing with the 3900's amps. They can be finicky.

FIRST: Q101, Q102, Q103, Q104, Q105 & Q106 are NOT replaced willy nilly. They have to be PROVEN BAD.
Q101 and 102 are offset compensation transistors for the output offset compensation setup.
Q103 and Q104 are the differential stage transistors, are FIVE leads each, and are TWO transistors in a single case, and they share EMITTER leads.
IF you do INDIVIDUAL eplacements, you need to find the highest gain matched pair.
Q105 and Q106 make up the current mirror portion of the differential stage, they have FIVE leads each, are TWO transistors in a single case, and they share BASE leads.

I HAVE done what you propose - shotgunning the ENTIRE amp, with those exceptions above. Don't forget to do ALL the diodes too, including the zeners. When I used ON-Semi MJE15032G and MJE15033G transistors in the Q134 - Q136 positions ( the four heatsunk to-220 drivers across one edge of the board) I had severe bias control problems, I could NOT dial the bias down to zero, and I had NEW power output transistors in place as well, which when I had FIRST used KSA940 & KSC2073 transistors the bias adjusted properly. I used the mje's as an "upgrade" so that there could NEVER EVER be a situation where the drivers were loaded down. Needless to say the MJE's came out and the KSA/C's went back in.
The outputs were On-semi:
863-MJL1302AG pnp TO-3P/TO-264 bce 260V 15/30A 200W 45-150 30MHZ $3.50 ea
863-MJL3281AG npn TO-3P/TO-264 bce 260V 15/30A 200W 45-150 30MHZ $3.50 ea
These are the smaller brothers to the ones EW used in the sx-1980 restoration



These are the six transistor types I used across the amp:

512-KSA992FBU (ln)to-92 ecb 120v .05a .5w 100mhz 150-800hfe $0.05 ea Q107, Q108, Q111, Q112, Q117, Q118, , Q103, Q104
512-KSC1845EBU (ln)to-92 ecb 120v .05a .5w 100mhz 150-800hfe $0.05 ea Q113, Q114, Q119, Q120 , Q101, Q102, Q105, Q106

512-KSA910YSHTA to-92 ecb 150v .05a .8w 100mhz 40-240hfe $0.07 ea Q115, Q116, Q121, Q122
512-KSC2310YBU to-92 ecb 150v .05a .8w 100mhz 40-240hfe $0.11 ea Q109, Q110, Q123, Q124

512-KSA940TU_Q pnp to-220 bce 150v 1.5a 25w 4mhz 40-140hfe $0.54 Q135, Q136
512-KSC2073TU npn to-220 bce 150v 1.5a 25w 4mhz 40-140hfe $0.54 Q133, Q134

misc diodes:
512-1N5234B 6.2v zener D101 - D106
512-1N4148 D113 - D116
512-UF4004 ultra fast rectifier 1amp 400v $0.13 ea D111, D112, D117, D118
512-1N4454 D129, D130, D132, D133
512-1N5245B 15v zener D131

and two transistor types I used in the protection circuit:

512-KSA1013YBU to-92 ecb 160v 1a .9w 50mhz 160-320hfe $0.17 ??Q139 (if 2sa1100)??
512-KSC2383YBU to-92 ecb 160v 1a .9w 50mhz 160-320hfe $0.17 Q125, Q126, Q137, Q138, ??Q139(if 2sc2575)??, Q140

edit, Here's the caps:
two 4.7uf 25v amplifier input caps 647-UPW1H4R7MDD 0.14 ea
one 100uf 25v protection cap 647-UPW1H101MPD 0.36 ea
one 470uf 16v protection cap 647-UPW1E471MPD 0.68 ea
two 0.22uf 50v protection caps- changed to poly metallized film 598-DSF050J224 0.50 ea


I have decided to NOT include the original transistor specifications in this post, nor detail (argue) the reasons for my choices of either which transistor replaced what, but even why there MAY be two different replacement numbers used for the same original number - it has to do with what is doing what, and where in the amp. (hint - a few get WARM, and I'm glad I did those substitutions)

The results speak for themselves, I have about the same distortion numbers that original versions of working amps produce.

PAY VERY CLOSE ATTENTION TO THE LEAD ARRANGEMENTS. The emitters are marked on the board.

You have a choice for Q139, depending upon it's type and the variant of the protection circuit on this unit. If it's the 2SA1100, it is the newer circuit.

I also did NOT have positive results using a zetex ztx1056a in the Q140 position, it immediately pulled the relay in on powerup. It could have been a leaky D131, but I just returned the original 2sc1384, after which it operated normally.

this post is provisionally finished. see post 41 for when....
Click to expand...
 
Skibjr,

Appreciate the kind reminder about the speaker relay. Will check it if there's still no sound after repairs.
 
markthefixer said:
As for fixing a bad 3900 amp - it's DEEP DEEP muck loaded with Gators.

100 percent component testing... no exceptions - no slip ups and probably a complete power transistor socket replacement (all 8) with something a lot more reliable. Mini euro terminal strips from Radio Shack - as documented on a recent thread.

IF any of the outputs are blown - this type of amp is very sensitive to mis-matched output transistor characteristics - a complete set of 4 for each channel affected is called for... I have TRIED it the other way... and the amp is VERY unforgiving. I MATCHED the original salvaged replacements, and things went from bad to worse - and a set of NEW transistors cleaned up many strange things happening - NOTHING else was changed.
Click to expand...

Is it correct, that the SX-D7000 has the same amp section as the 3900?
Then I'd agree! It was a pitty to repair this amp and after the first try it blew also a bit of fire out of the top with QUEEN ! :D
 
agger said:
Is it correct, that the SX-D7000 has the same amp section as the 3900?
Click to expand...

Identical. Possibly some of the older 3900's have an older variant of the protection circuit, but the amp circuits are = .

agger said:
Then I'd agree! It was a pitty to repair this amp and after the first try it blew also a bit of fire out of the top with QUEEN ! :D
Click to expand...
 
...and after you do everything Mark tells you to on the amp section, don't forget to pull the P/S board out and dress all of the cracked solder joints and replace the dried out caps.
 
markthefixer said:
I gave you two perfectly good TRUSTWORTHY output transistor replacements, and adequate if not detailed "marching orders" as well as the benefit of some HARD WON experience.

This is the HARDEST type of Pioneer Amplifier to fix.

The adage about " leading a horse to water but can't make him drink" seems to apply.

You have dead diodes, blown driver transistors, and a bunch of hidden - blown out resistors - which can ONLY be found by 100 percent component testing. This is mostly the way I have to do it.

There ARE lists of the components necessary to do this board posted on the Forum, in the last year and the original protection relay IS still available from Pioneer for about 50 bucks, 60 shipped.

This looks right now like an impending train wreck.

The thing is that the diodes, transistors and resistors - excluding the outputs - are pretty cheap to replace - even if a bit tedious - but you did that pretty well on the 1050, so I hope you pay attention now.


http://www.audiokarma.org/forums/showthread.php?t=183134

http://www.audiokarma.org/forums/showthread.php?t=161866

http://www.audiokarma.org/forums/showthread.php?t=190127
Click to expand...
Question: Thanks for posting this, it gives me hope. Which link is for the pioneer sx-3900 (white silicon plug thingies on underside att to heat sink?) silicon output transistor placement. My unit is working great but after finding all of the output transistor sockets cracked I'm a little frightened to run it. looking forward to a response - thanks a ton.
 
You must log in or register to reply here.
Back
Top Bottom