1TUFSS
Thread Killer
Progress is brutally slow, checking all components (especially diodes and resistors) against the good channel, but no breakthroughs yet. Calling on our god, @markthefixer, for advice.
SX-3900 FM: Nothing but static
- Thread starter 1TUFSS
- Start date
So current issue is the right amp will not bias up properly?
Comparing the good channel against the bad one is a good method. do they have the same parts installed? If you changed device types, it is possible you threw off the whole bias arrangement and need to re-calc the values for the specific devices used. I'd have to investigate further to give better advise.
Also suggest to read/understand the pioneer tuning fork "Non-switching Amp" technical guide. It is written for techs trying to understand and repair the topology. It is in 3 parts found in this forum
Comparing the good channel against the bad one is a good method. do they have the same parts installed? If you changed device types, it is possible you threw off the whole bias arrangement and need to re-calc the values for the specific devices used. I'd have to investigate further to give better advise.
Also suggest to read/understand the pioneer tuning fork "Non-switching Amp" technical guide. It is written for techs trying to understand and repair the topology. It is in 3 parts found in this forum
Many of these posts are hard to follow, since you only visit them once and a while, then you have to refresh, which can mean re-reading them all over again and again. it would be nice to summarize the faults and the fixes. Like in this case. wrong order?
Fault:
1) No FM.
Cleaned switch = no fix,
mtf: finger test, possible HA1201 if amp, change, no result,
mtf: cap couple RF FE to IF system(PAxxx) detctor input = audio
guess at ceramic filters, change all ceramic filters = fix = yeah, need ctest/check alignmnet
2) Audio Muting issue: R?? open? see I have to look
3) left power amp will not bias.
offset adjusts okay?
what actual step in the manual?
Check all bias voltages against sch?
you are ordering drivers, what part#, determined the original parts(Q?) as what condition, shorted/ open
Sorry I am an old retired EE, we were taught to document our work so others could follow along easy (like if I quit or get fired or run over) and not waist precious resources
Just think we may see even more tariffs when they tear up trade deals back to the old days.
Think about it, most of the parts are made in Asia, US is just a place to put your inventory and mark it up, as service.
Mouser gives you printouts of places of origin as part of a ROHS filing are required to sell into the EU.
Fault:
1) No FM.
Cleaned switch = no fix,
mtf: finger test, possible HA1201 if amp, change, no result,
mtf: cap couple RF FE to IF system(PAxxx) detctor input = audio
guess at ceramic filters, change all ceramic filters = fix = yeah, need ctest/check alignmnet
2) Audio Muting issue: R?? open? see I have to look
3) left power amp will not bias.
offset adjusts okay?
what actual step in the manual?
Check all bias voltages against sch?
you are ordering drivers, what part#, determined the original parts(Q?) as what condition, shorted/ open
Sorry I am an old retired EE, we were taught to document our work so others could follow along easy (like if I quit or get fired or run over) and not waist precious resources
Just think we may see even more tariffs when they tear up trade deals
back to the old days.Think about it, most of the parts are made in Asia, US is just a place to put your inventory and mark it up, as service.
Mouser gives you printouts of places of origin as part of a ROHS filing are required to sell into the EU.
Last edited:
1TUFSS
Thread Killer
All imports are subject to VAT, and those valued above £200 or so get slapped with duty. Bit of a bugger, but at least we had no customs borders within the EU - until now. Thanks, Brexit.
Farnell didn't have them, no.
Many of these posts are hard to follow, since you only visit them once and a while, then you have to refresh, which can mean re-reading them all over again and again. it would be nice to summarize the faults and the fixes. Like in this case. wrong order?
Fault:
1) No FM.
Cleaned switch = no fix,
mtf: finger test, possible HA1201 if amp, change, no result,
mtf: cap couple RF FE to IF system(PAxxx) detctor input = audio
guess at ceramic filters, change all ceramic filters = fix = yeah, need ctest/check alignmnet
2) Audio Muting issue: R?? open? see I have to look
3) left power amp will not bias.
offset adjusts okay?
what actual step in the manual?
Check all bias voltages against sch?
you are ordering drivers, what part#, determined the original parts(Q?) as what condition, shorted/ open
Sorry I am an old retired EE, we were taught to document our work so others could follow along easy (like if I quit or get fired or run over) and not waist precious resources
Just think we may see even more tariffs when they tear up trade deals
Think about it, most of the parts are made in Asia, US is just a place to put your inventory and mark it up, as service.
Mouser gives you printouts of places of origin as part of a ROHS filing are required to sell into the EU.
FM: No front-end signal, just inter-station hiss. HA1201 ICs tested OK, all 3 filters tested open, possibly something to do with the circuit board gunk. All filters changed for a matched set, FM is back and kicking ass. Could benefit from an alignment (as all old tuners can) but it's not greatly in need.
-20dB audio mute: Left channel was not muting. All switches and pots were already flushed, so that couldn't be the culprit. Traced to R149 & R150 on main board, with one testing way out of value and the other a dead short. Washed the board to remove gunk, and problem disappeared. Changed resistors for 1% tolerance anyway.
Power amp: Right channel quieter than left, and audible distortion at high power output. Left was OK. 2SA985 driver found shorted base to emitter, swapped for TIP30c (temporary measure, Ft too slow) and channels are now equal. Distortion caused by lack of bias; with both pots cranked fully clockwise, it won't give more than 3mV. That's where we are now. Going through the whole board, a little each day, checking and re-checking values against the left channel, wondering whether I'm missing something. Waiting for 2SC5171 & 2SA1930 to arrive from US. Left channel is spot-on 70mV and 1mV offset.
All power supply voltages are 100% spot on.
Great, keep going as you are, mark up a sch with your measured voltages, comparing against what they printed and the other working channel.
doing comparison of one ch vs the other powered down, ohms test can reveal differences, areas to concentrate on.
3mV bias across all the o/p emitter R's, have to concentrate on the bias generator. Re's in tolerance?
I am going to sit down and sim this ckt one day to get a better idea of how it works,
I sim'd the Sansui F-2980 (G-7500) yesterday, was bored doing garage wiring With MJL1302/3281 it performs not too bad.
Could take the risky step of using parts from one channel to test the other
doing comparison of one ch vs the other powered down, ohms test can reveal differences, areas to concentrate on.
3mV bias across all the o/p emitter R's, have to concentrate on the bias generator. Re's in tolerance?
I am going to sit down and sim this ckt one day to get a better idea of how it works,
I sim'd the Sansui F-2980 (G-7500) yesterday, was bored doing garage wiring
With MJL1302/3281 it performs not too bad.Could take the risky step of using parts from one channel to test the other
Last edited:
1TUFSS
Thread Killer
Soooooooooooooooo... after many months of cursing, desk-thumping, hair-tearing, and honestly just finding better and/or more urgent things to do, I'm bloody thrilled to say all is well in this big 3900.
The bias faults were traced to a trio of partially shorted Zener diodes: D102, 104 and 106 on the power amp board. Not completely shorted, reverse-bias was okay, but forward voltage was off. Seems that bad things come in threes with this unit. Anyway, they were replaced, and the bias suddenly started behaving properly. Not to be outdone in the calamity stakes, Mouser notified me that Toshiba have axed the 2SC5171/2SA1930 combo which is the go-to for these non-switching Pioneers, meaning I was waiting for a big bag of nothing to arrive. Fortunately, a local supplier had a large inventory, so I stocked up for the oncoming apocalypse by ordering 50 of each for future units; the supplier needed an extra day for the PNP devices, so they were sent out later, whereupon Royal Mail lost the whole lot of NPN! Fear not, they showed up a couple of weeks later, they got bunged into the unit, and it's all singing and dancing.
Now to source a good faceplate, volume knob, and build a new cabinet, as the old one is falling apart. Fun, fun, fun.
The bias faults were traced to a trio of partially shorted Zener diodes: D102, 104 and 106 on the power amp board. Not completely shorted, reverse-bias was okay, but forward voltage was off. Seems that bad things come in threes with this unit. Anyway, they were replaced, and the bias suddenly started behaving properly. Not to be outdone in the calamity stakes, Mouser notified me that Toshiba have axed the 2SC5171/2SA1930 combo which is the go-to for these non-switching Pioneers, meaning I was waiting for a big bag of nothing to arrive. Fortunately, a local supplier had a large inventory, so I stocked up for the oncoming apocalypse by ordering 50 of each for future units; the supplier needed an extra day for the PNP devices, so they were sent out later, whereupon Royal Mail lost the whole lot of NPN! Fear not, they showed up a couple of weeks later, they got bunged into the unit, and it's all singing and dancing.
Now to source a good faceplate, volume knob, and build a new cabinet, as the old one is falling apart. Fun, fun, fun.
Last edited:
Imanoldee
Super Member
Hello 1TUFSS,
If you still have your SX3900 opened up is there any way you can measure a couple of points on the power supply section (AWR210)?
I am worried about R20 being run with to much power for its rating.
I sold my SX-3900 some time back and so can not confirm or dismiss this possible problem.
I was reviewing its schematic recently when the possibility of a problem came to mind.
Take a look at my thread the Pioneer SX-3900.
Any one have a Pioneer SX-3900 opened up?
Any chance you can you measure the 2 sides of R20 for voltage and see what size resister is used there?
Thanks
If you still have your SX3900 opened up is there any way you can measure a couple of points on the power supply section (AWR210)?
I am worried about R20 being run with to much power for its rating.
I sold my SX-3900 some time back and so can not confirm or dismiss this possible problem.
I was reviewing its schematic recently when the possibility of a problem came to mind.
Take a look at my thread the Pioneer SX-3900.
Any one have a Pioneer SX-3900 opened up?
Any chance you can you measure the 2 sides of R20 for voltage and see what size resister is used there?
Thanks
