Pioneer SX-9 Where to start, if at all
- Thread starter weeber
- Start date
Ok, don't fool around with q115, q116, R127, R128.
D115 and D135 are in the good channel? :scratch2:
That's all for tonight - I HAVE to do some other things.
D115 and D135 are in the good channel? :scratch2:
That's all for tonight - I HAVE to do some other things.
Last edited:
weeber
Active Member
D115 & D135 are on the good (right) channel.
weeber
Active Member
Help me to understand the flow of electrons
To help me understand better what's going on and to learn how to read schematics better I've been studying the diagrams and identifying all the parts Mark wanted me to check and also found all the parts I've replaced.
When I originally got this unit the left channel idle current wouldn't adjust down, it was too high. The right side was fine.
In the process of checking the function of the VR pots, I accidentally left VR5 on the fight channel fully clockwise and turned on the amp. Fireball ensued.
Post mortem revealed I blew up on the right side the outputs q3, q4, q7 and q8. Q133 and Q135 where also gone. The emitter resistors r183 and r185 were toast. Two diodes, D115 and D135 needed replacement as did resistors r163, r177 and r181.
I noticed that at some previous time before me it looks like the left side had destructed. The original outputs 2sa1076 and 2sc2526 had been replaced with ecg33 and ecg34. The transistors q134 and q136 which originally were 2sa985a and 2sc2275a had been replaced with 2sb595 and 2sd525. this was the side that wouldn't originally adjust.
Well, so everything was equal on both sides I replaced all the outputs with Sanken 2sa1216 and 2sc2922. I replaced q133, q134, q135 and q136 with 2sa1930 and 2sc5171.
After all this, the left side still doesn't adjust correctly. It's better though. Originally the best it would adjust to 139mV. Now it starts at a better 58mV but drifts upto 78mV.
So even after fixing my mistake on the right side, the left still has issues, although less than before, and probably less just because the trimpots were all replaced with better new ones.
Correct me when I'm wrong, the voltage at VR4 is too high even at it's zero setting. Therefore, there is something (one or more components) BEFORE VR4 causing the problem. BUT the voltages at the thermal diodes were the same side to side. Are these before or after VR4?
To help me understand better what's going on and to learn how to read schematics better I've been studying the diagrams and identifying all the parts Mark wanted me to check and also found all the parts I've replaced.
When I originally got this unit the left channel idle current wouldn't adjust down, it was too high. The right side was fine.
In the process of checking the function of the VR pots, I accidentally left VR5 on the fight channel fully clockwise and turned on the amp. Fireball ensued.
Post mortem revealed I blew up on the right side the outputs q3, q4, q7 and q8. Q133 and Q135 where also gone. The emitter resistors r183 and r185 were toast. Two diodes, D115 and D135 needed replacement as did resistors r163, r177 and r181.
I noticed that at some previous time before me it looks like the left side had destructed. The original outputs 2sa1076 and 2sc2526 had been replaced with ecg33 and ecg34. The transistors q134 and q136 which originally were 2sa985a and 2sc2275a had been replaced with 2sb595 and 2sd525. this was the side that wouldn't originally adjust.
Well, so everything was equal on both sides I replaced all the outputs with Sanken 2sa1216 and 2sc2922. I replaced q133, q134, q135 and q136 with 2sa1930 and 2sc5171.
After all this, the left side still doesn't adjust correctly. It's better though. Originally the best it would adjust to 139mV. Now it starts at a better 58mV but drifts upto 78mV.
So even after fixing my mistake on the right side, the left still has issues, although less than before, and probably less just because the trimpots were all replaced with better new ones.
Correct me when I'm wrong, the voltage at VR4 is too high even at it's zero setting. Therefore, there is something (one or more components) BEFORE VR4 causing the problem. BUT the voltages at the thermal diodes were the same side to side. Are these before or after VR4?
I repair MORE damage after that is said. It's the CARDINAL sin in the MTF universe. It's called "jumping ahead" RIGHT into the middle of the damn minefield.
:nono: :bash:
:tears: ... moving on ... :sigh:
Once again, the student has failed to heed the master.
Just kidding guys. One of the most entertaining threads on Ak right now, and such a small audience. :scratch2:
weeber
Active Member
Once again, the student has failed to heed the master.
Just kidding guys. One of the most entertaining threads on Ak right now, and such a small audience. :scratch2:
There's a peanut gallery of one!
I would have given up long ago, but I know you need closure. :yes:
So, this is for you onwardjames
Thanks man, and hang in there.
There's a peanut gallery of one!
I would have given up long ago, but I know you need closure. :yes:
So, this is for you onwardjames
Heh, well maybe two in the peanut gallery. I've also been following this thread, mostly because I also like to watch MTF work a problem, and this repair has had the excitement of blowing up due to operator error. LOL, as MTF would say, BTDT, got the tshirt.
Only a few months ago, I had a repair that was COMPLETE - everything done, tested and adjusted. Then I started reassembly without draining the main caps. Blew it all up again! So, I feel for ya buddy.
weeber
Active Member
I've been researching other's threads about sx-3900 and sx-d7000 with idle current issues.
I found one http://www.audiokarma.org/forums/showthread.php?t=540812&page=2 that sounded familiar to the issue I'm having on the left channel. The idle current was close but still too high.
This person checked r139 and r141 and found their values low at around 216ohms instead of the required 220. These were replaced with ones with better values of around 221ohms. This bought a few mVs, but still not enough. They were next replaced with higher value resistors of 270ohms, which actually measured about 261ohms.
This allowed the idle voltage to be adjusted properly.
I downloaded a service manual for the SX-3900 and found the amp boards almost identical. R139 And r141 are for the right side and would say that these would correspond with r140 and r142 on the left side for an SX-9. Does this sound reasonable?
My question: would it be prudent to check the values of these 2 resistors (lifting the leg of course)? Barking up the wrong tree?
I won't make a move without the advice of an expert, I'm just trying to sort stuff out a little on my own, as a learning exercise.
I found one http://www.audiokarma.org/forums/showthread.php?t=540812&page=2 that sounded familiar to the issue I'm having on the left channel. The idle current was close but still too high.
This person checked r139 and r141 and found their values low at around 216ohms instead of the required 220. These were replaced with ones with better values of around 221ohms. This bought a few mVs, but still not enough. They were next replaced with higher value resistors of 270ohms, which actually measured about 261ohms.
This allowed the idle voltage to be adjusted properly.
I downloaded a service manual for the SX-3900 and found the amp boards almost identical. R139 And r141 are for the right side and would say that these would correspond with r140 and r142 on the left side for an SX-9. Does this sound reasonable?
My question: would it be prudent to check the values of these 2 resistors (lifting the leg of course)? Barking up the wrong tree?
I won't make a move without the advice of an expert, I'm just trying to sort stuff out a little on my own, as a learning exercise.
Yes to your 1-4 post, check the resistances.
What are the parts numbers of q112(- side) , q114(+ side), q118(+ side), q120(- side) transistors left channel
What are the parts numbers of q111(- side) , q113(+ side), q117(+ side), q119(- side) transistors right channel
If the eight transistors are of the two parts numbers only (thus identical) does swapping these transistors from channel to channel make any difference?
that is: (q111<>q112) (q113<>q114)(q117<>q118)(q119<>q120)...
DOUBLE CHECK MY STUFF ABOVE SO THAT YOU DON'T CHANGE ONE TRANSISTOR (ksa,2sa,ksb,2sb are PNP and ksc,2sc,ksd,2sd are NPN)
WITH AN OPPOSITE POLARITY TRANSISTOR.
If the trouble doesn't move we have exonerated the transistors. So then take a very close look at all the values of all the resistors connected to these transistors.
Do a daily bump when I don't respond, so it doesn't fall in the cracks (again).
There ARE several day gaps in my posting....
What are the parts numbers of q112(- side) , q114(+ side), q118(+ side), q120(- side) transistors left channel
What are the parts numbers of q111(- side) , q113(+ side), q117(+ side), q119(- side) transistors right channel
If the eight transistors are of the two parts numbers only (thus identical) does swapping these transistors from channel to channel make any difference?
that is: (q111<>q112) (q113<>q114)(q117<>q118)(q119<>q120)...
DOUBLE CHECK MY STUFF ABOVE SO THAT YOU DON'T CHANGE ONE TRANSISTOR (ksa,2sa,ksb,2sb are PNP and ksc,2sc,ksd,2sd are NPN)
WITH AN OPPOSITE POLARITY TRANSISTOR.
If the trouble doesn't move we have exonerated the transistors. So then take a very close look at all the values of all the resistors connected to these transistors.
Do a daily bump when I don't respond, so it doesn't fall in the cracks (again).
There ARE several day gaps in my posting....
Last edited:
weeber
Active Member
All 8 transistors are only 2 part numbers.
Q111 q112 q117 and q118 are 2sa1115
Q113 q114 q119 and q120 are 2sc2603
They all checked good when tested out of the circuit and when swapped from side to side there was no change in symptoms.
Right side starts its adjustments at 0mV and adjust to 56 and 70mV just fine and steady.
Left side starts at 55mV and slowly drifts up to 66mV in about 2 minutes without any adjustments. It seems to stay at about 66mV once settled.
Any clues as to which resistors to test in particular? R140 and r142 both checked out OK at 220 Ohms.
Q111 q112 q117 and q118 are 2sa1115
Q113 q114 q119 and q120 are 2sc2603
They all checked good when tested out of the circuit and when swapped from side to side there was no change in symptoms.
Right side starts its adjustments at 0mV and adjust to 56 and 70mV just fine and steady.
Left side starts at 55mV and slowly drifts up to 66mV in about 2 minutes without any adjustments. It seems to stay at about 66mV once settled.
Any clues as to which resistors to test in particular? R140 and r142 both checked out OK at 220 Ohms.
Wow, you are really putting in the hours here.
OK, identify all the interrelated components around those 8 transistors, and get a can of keyboard duster.
COOL each component individually, do NOT BLAST them hard, they will get cold enough to form condensation, ice, or white stuff - which we do not want.
Try to shield the non-cooled components with some cardboard etc...
Sometimes a small non-conducting tube like a large straw can isolate a component from all others.
WATCH the idle current like a hawk. when it starts to dip back to the 55mv you are getting close.
To confirm, when THAT component is swapped to the other channel, the problem will move.
You MAY see other cooling effects on innocent components. We don't want to drive them to extremes.
DO NOT cool down the transistors - we swapped them without changes, but the temperature changes WILL induce significant masking changes in them. Changes we are not interested in chasing.
The idea is to cancel out that 2 minutes of warm up, not to turn the area into the Antarctic.
The technique works, we once found a temperature sensitive quartz crystal for an oscillator in a receiver doing this. I would have NEVER thought of THAT otherwise.
You CAN use a hairdryer to reverse the cooling, but again, take it easy, if it's too hot to touch, it's way too damn hot for this situation.
This is probably faster and easier than taking a gazillion voltage measurements.
I say keyboard duster, because it's far less expensive and much less aggressive than freeze mist. But it WILL freeze if someone gets enthusiastic with it.
OK, identify all the interrelated components around those 8 transistors, and get a can of keyboard duster.
COOL each component individually, do NOT BLAST them hard, they will get cold enough to form condensation, ice, or white stuff - which we do not want.
Try to shield the non-cooled components with some cardboard etc...
Sometimes a small non-conducting tube like a large straw can isolate a component from all others.
WATCH the idle current like a hawk. when it starts to dip back to the 55mv you are getting close.
To confirm, when THAT component is swapped to the other channel, the problem will move.
You MAY see other cooling effects on innocent components. We don't want to drive them to extremes.
DO NOT cool down the transistors - we swapped them without changes, but the temperature changes WILL induce significant masking changes in them. Changes we are not interested in chasing.
The idea is to cancel out that 2 minutes of warm up, not to turn the area into the Antarctic.
The technique works, we once found a temperature sensitive quartz crystal for an oscillator in a receiver doing this. I would have NEVER thought of THAT otherwise.
You CAN use a hairdryer to reverse the cooling, but again, take it easy, if it's too hot to touch, it's way too damn hot for this situation.
This is probably faster and easier than taking a gazillion voltage measurements.
I say keyboard duster, because it's far less expensive and much less aggressive than freeze mist. But it WILL freeze if someone gets enthusiastic with it.
weeber
Active Member
I used canned air, for cleaning keyboards and such with a bendy straw to direct the air and try to shield each component. I couldn't get all the components well as the back cover needs to be moved a little and I'm scared of shorting something with the back cover moved and the amp plugged in. To move the cover that components for the multi voltage switch need to be unscrewed from the back cover and there are a lot of un-insulated exposed wire connections about.
Anyway, the voltage drifted up to 72mV and I started spring things. i gave each one a couple second blast of air. But nothing changed. Once, the voltage dropped a little to 68mV when I sprayed d102. I let it drift back up to 72mV, but couldn't reproduce the results.
I'll try again later with better access (I couldn't get many components on the lower part of the board as the canned air was tilted so that liquid would come out the tube at the angle to reach them)
Anyway, the voltage drifted up to 72mV and I started spring things. i gave each one a couple second blast of air. But nothing changed. Once, the voltage dropped a little to 68mV when I sprayed d102. I let it drift back up to 72mV, but couldn't reproduce the results.
I'll try again later with better access (I couldn't get many components on the lower part of the board as the canned air was tilted so that liquid would come out the tube at the angle to reach them)
weeber
Active Member
Obviously, I'm doing something wrong.
This is what I'm using to cool the components
I've put it on its side and moved the back panel
Here's a close up over the board
I've determined these are the target components :
Th102
R130, 132, 136, 138, 140, 142, 144, 146, 152, 200
D112, 114, 116, 136
C136
The amp on its side settled at 66mV. But each thing I blasted with air would cause an increase in voltage by a mV or two.
So I started testing all the components on the board. The only one that seemed to do anything was r174 which seemed to drop the voltage down a bit where it settled back at 64mV. I couldn't reproduce these results though when the board was back up to 69mV.
This is what I'm using to cool the components
I've put it on its side and moved the back panel
Here's a close up over the board
I've determined these are the target components :
Th102
R130, 132, 136, 138, 140, 142, 144, 146, 152, 200
D112, 114, 116, 136
C136
The amp on its side settled at 66mV. But each thing I blasted with air would cause an increase in voltage by a mV or two.
So I started testing all the components on the board. The only one that seemed to do anything was r174 which seemed to drop the voltage down a bit where it settled back at 64mV. I couldn't reproduce these results though when the board was back up to 69mV.
Last edited:
straw independent of can, notch in bottom of straw to fit around component leads, spray is shot in at middle of straw, in small volume, with back spray coming back UP and out the same end of the straw.
A full VOLT!! that's too much, 10 millivolts is our error.
I will have to review your component list and the circuit in question. Been a long day...
A full VOLT!! that's too much, 10 millivolts is our error.
I will have to review your component list and the circuit in question. Been a long day...
weeber
Active Member
Okay. I tried the straw method how you described, Mark, but the result was still the same.
Which ever component I blasted I either got no response or the mV would only increase (then drop back down again after 30 seconds or so) .
I tried every component on the left side that wasn't a transistor.
Which ever component I blasted I either got no response or the mV would only increase (then drop back down again after 30 seconds or so) .
I tried every component on the left side that wasn't a transistor.
Ok you swapped the NSA section channel to channel: 8 transistors per channel. Without the problem moving.
Now it's time to swap the drivers side to side. 2 per channel, Q133 <> q134 and q135<> q136.
Step after that is to start measuring a LOT of voltages on both channels.
I am trusting / assuming that the components you named above for the freeze treatment, were all tested for values with one leg lifted each.
Now it's time to swap the drivers side to side. 2 per channel, Q133 <> q134 and q135<> q136.
Step after that is to start measuring a LOT of voltages on both channels.
I am trusting / assuming that the components you named above for the freeze treatment, were all tested for values with one leg lifted each.
Last edited:
weeber
Active Member
The drivers, all 4, are brand new. Is it possible new components could be faulty?
The resistors and other components were all lifted and tested when I replaced all the blown pieces on the right side. It is possible I could have missed or read something wrong.
Before I tear into it again, can I check a few voltages to see if the above is necessary? Voltages are easier to take then dis assembly again.
I'm just the padawan learner, though, so I will do whatever you advise and follow your wisdom.
The resistors and other components were all lifted and tested when I replaced all the blown pieces on the right side. It is possible I could have missed or read something wrong.
Before I tear into it again, can I check a few voltages to see if the above is necessary? Voltages are easier to take then dis assembly again.
I'm just the padawan learner, though, so I will do whatever you advise and follow your wisdom.