Yes it does.
Pioneer SX-737 Distortion left channel
- Thread starter daveyh
- Start date
Yes it does.
BTW, the Q1, Q3, Q5 cluster of transistors on the foil diagram in the manual has a mis-identification!!
What IT calls Q8 is really Q3 (2sa726 > ksa992). Did you did notice that?
Q5 looks backwards to me.
This is after a LOT of fiddle pharting with images, service manual and your photograph.
I am going by the un-symmetric seam on the case of the c3503. I have one in front of me.
The correct orientation will have the Q5 lettering facing AWAY from VR3. Is it?
On Q6, the lettering will FACE VR4. They will be pointing the same way, towards VR1 and the pin 1 end of the board.
That puts both Q5 and Q6 emitters closer to the off-board power transistor connections.
That puts the bases closer to the pins 1-19 row, with the collectors off to the side between the emitter and base.
IF Q5 isn't backwards, take voltage readings between:
R19 (inboard side lead, red probe) and Q5 collector on the nicely sticking out to the left center lead.
r20 (inboard side lead, red probe) and q6 collector on the nicely sticking out to the left center lead.
Vary VR1 (r19) and VR2(R20) and watch the readings. DON'T crank it too hard, it's the idle current, the working VR2 side can be turned down and back up again, while vr1 - having no idle current effect, can be turned a bit further.
What IT calls Q8 is really Q3 (2sa726 > ksa992). Did you did notice that?
Q5 looks backwards to me.
This is after a LOT of fiddle pharting with images, service manual and your photograph.
I am going by the un-symmetric seam on the case of the c3503. I have one in front of me.
The correct orientation will have the Q5 lettering facing AWAY from VR3. Is it?
On Q6, the lettering will FACE VR4. They will be pointing the same way, towards VR1 and the pin 1 end of the board.
That puts both Q5 and Q6 emitters closer to the off-board power transistor connections.
That puts the bases closer to the pins 1-19 row, with the collectors off to the side between the emitter and base.
IF Q5 isn't backwards, take voltage readings between:
R19 (inboard side lead, red probe) and Q5 collector on the nicely sticking out to the left center lead.
r20 (inboard side lead, red probe) and q6 collector on the nicely sticking out to the left center lead.
Vary VR1 (r19) and VR2(R20) and watch the readings. DON'T crank it too hard, it's the idle current, the working VR2 side can be turned down and back up again, while vr1 - having no idle current effect, can be turned a bit further.
Last edited:
Q: BTW, the Q1, Q3, Q5 cluster of transistors on the foil diagram in the manual has a mis-identification!!
What IT calls Q8 is really Q3 (2sa726 > ksa992). Did you did notice that?
A: Yes I did.
Q: Q5 looks backwards to me.
This is after a LOT of fiddle pharting with images, service manual and your photograph.
I am going by the un-symmetric seam on the case of the c3503. I have one in front of me.
The correct orientation will have the Q5 lettering facing AWAY from VR3. Is it?
On Q6, the lettering will FACE VR4. They will be pointing the same way, towards VR1 and the pin 1 end of the board.
A: Yes they are installed correctly. I have a Peak Pro Transistor Checker and checked the transistors before I installed them. I also went by the screen printing on the board.
I hope I did this right. I placed the red mini clip on the inside of the r19 and r20 resistor and the black mini clip on the collector.
VR1 (r19)
Turned the variable resistor all the way to the right and it reads 2.781v.
All the way to the left it reads 2.171v
VR2(R20)
Turned the variable resistor all the way to the right and it reads 2.749v.
All the way to the left it reads 2.152v
ALL the way to the right was
2.71 v means huge idle currents flowed, fortunately no magic smoke got out.
Ok, those readings exonerate the idle current circuits R19(20), Vr1(2), Q5(6) and so forth.
I TRICKED you into reading the bases of Q7(8) & Q9(10).
Read left channel output voltage on pin 3 voltage (red) to ground(black), CAN vr3 adjust pin 3 to 0.000v
Read pin 17 voltage (red) to ground(black), CAN vr4 adjust pin 17 to 0.000v?
read and post these voltages to ground (black)
pins 20, 21, 22
pins 23, 24, 25
pins 30, 31, 32(33)
pins 26, 27, 28
While adjusting the idle current pots (again)
read BETWEEN pin 27 (black) and pin 31 (red) for the left channel (vr1)
and pins 24 (black) and pin 21 (red) for the right channel (vr2).
IT should be in the range of (2.7 - 1.2 = 1.5) and (2.17 - 1.2 = 0.97) BUT DON'T GO OVER 1.3 VOLTS!!!
OR if you have a second voltmeter, WATCH the idle current and don't go over 100mA (0.1v) idle current.
2.71 v means huge idle currents flowed, fortunately no magic smoke got out.
Ok, those readings exonerate the idle current circuits R19(20), Vr1(2), Q5(6) and so forth.
I TRICKED you into reading the bases of Q7(8) & Q9(10).
Read left channel output voltage on pin 3 voltage (red) to ground(black), CAN vr3 adjust pin 3 to 0.000v
Read pin 17 voltage (red) to ground(black), CAN vr4 adjust pin 17 to 0.000v?
read and post these voltages to ground (black)
pins 20, 21, 22
pins 23, 24, 25
pins 30, 31, 32(33)
pins 26, 27, 28
While adjusting the idle current pots (again)
read BETWEEN pin 27 (black) and pin 31 (red) for the left channel (vr1)
and pins 24 (black) and pin 21 (red) for the right channel (vr2).
IT should be in the range of (2.7 - 1.2 = 1.5) and (2.17 - 1.2 = 0.97) BUT DON'T GO OVER 1.3 VOLTS!!!
OR if you have a second voltmeter, WATCH the idle current and don't go over 100mA (0.1v) idle current.
ALL the way to the right was
2.71 v means huge idle currents flowed, fortunately no magic smoke got out.
Ok, those readings exonerate the idle current circuits R19(20), Vr1(2), Q5(6) and so forth.
I TRICKED you into reading the bases of Q7(8) & Q9(10).
---Read left channel output voltage on pin 3 voltage (red) to ground(black), CAN vr3 adjust pin 3 to 0.000v
** Already at 0.000v. Yes it can be adjusted.
---Read pin 17 voltage (red) to ground(black), CAN vr4 adjust pin 17 to 0.000v? Read and post these voltages to ground (black)
** Already at 0.000v. Yes it can be adjusted.
pins 20, 21, 22; (pin 20 = 0.055v); (pin 21 = 0.683v); (pin 22 = 35.00v)
pins 23, 24, 25; (pin 23 = -0.055v); (pin 24 = -0.662v); (pin 25 = -34.91)
pins 30, 31, 32 (33); (pin 30 = 0.003v); (pin 31 = 0.832v); (pin 32 = 35.07v); (pin 33 = 35.07v)
pins 26, 27, 28; (pin 26 = -0.004v); (pin 27 = -0.513v); (pin 28 = -34.93v)
** Did not know what you were referring to. But the voltages are below. They were adjustable but I did not want to go to far.
While adjusting the idle current pots (again)
read BETWEEN pin 27 (black) and pin 31 (red) for the left channel (vr1)
and pins 24 (black) and pin 21 (red) for the right channel (vr2)
** Pin 27 (black) and 31 (red) = 1.330v
** Pin 24 (black) and 21 (red) = 1.22v
Note: IT should be in the range of (2.7 - 1.2 = 1.5) and (2.17 - 1.2 = 0.97) BUT DON'T GO OVER 1.3 VOLTS!!!
I suspect (MORE analysis needed, later tonight - time crunch) the output transistors, one is probably bad. +.832, -0.513 yet 0.00x offset.
Notice on the right channel, +/- 0.055 ? you should be measuring idle by putting the red lead on pin 18 and the black lead on pin 19 (at the same time), that would give you a reading of 0.110v - which is twice as much idle current as should be flowing. I suspect you used ground referenced readings to set idle current.
Notice on the right channel, +/- 0.055 ? you should be measuring idle by putting the red lead on pin 18 and the black lead on pin 19 (at the same time), that would give you a reading of 0.110v - which is twice as much idle current as should be flowing. I suspect you used ground referenced readings to set idle current.
These are brand new Onsemi outputs. I followed the service manual to set the voltages. I have done this many times before on many receivers. This has been the hardest one I think I have tried to tackle in a long time. I will pull the outputs and check them.
http://www.audiokarma.org/forums/showthread.php?t=464083
AWH033 q11 2sd370 863-MJ21194G
AWH033 q12 2sd370 863-MJ21194G
AWH033 q13 2sb530 863-MJ21193G
AWH033 q14 2sb530 863-MJ21193G
Last edited:
Ok, if the output transistor are STILL out, leave them out, INSULATE the sockets and turn ON the amp. Then repeat the measurements. This type of amp (having r23, r25 and feedback tied together) will balance AND the offset adjust will work. THEN the "idle current" adjustment can be twisted any which way safely, as all it is now doing is varying a base voltage, and somewhat affecting the driver (q7,q9 & q8, q10) pairs idle current - and IIAC it can't be turned up high enough to hurt THEM in THIS circuit.
later we might swap r31 with R32, then swap r33 with r34 when the outputs go back in.
since they are 0.5 ohm resistors, you will have difficulty reading 1/2 of an ohm accurately.
Because SOMETHING caused :
When the other channel
.832 + .513 = 1.345 a theoretical offset voltage of +0.319 v would be generated on the working channel.
.683 + .662 = 1.345 a theoretical offset voltage of +0.021 v would be generated on the working channel.
Another "given" is that the idle current measurement is ACROSS BOTH 0.5 emitter resistors of the output. Pin 20 being +0.055v and pin 23 being -0.055v just does NOT square with a reported successful 20mA idle current measurement.
I cannot imagine the various errors and faults necessary to result in these readings.
I understand the circuits, from DECADES of experience, technician based and ENGINEERING based. I have seen countless fault conditions in these amps.
It literally "doesn't add up".
later we might swap r31 with R32, then swap r33 with r34 when the outputs go back in.
since they are 0.5 ohm resistors, you will have difficulty reading 1/2 of an ohm accurately.
Because SOMETHING caused :
When the other channel
is symmetrical.
.832 + .513 = 1.345 a theoretical offset voltage of +0.319 v would be generated on the working channel.
.683 + .662 = 1.345 a theoretical offset voltage of +0.021 v would be generated on the working channel.
Another "given" is that the idle current measurement is ACROSS BOTH 0.5 emitter resistors of the output. Pin 20 being +0.055v and pin 23 being -0.055v just does NOT square with a reported successful 20mA idle current measurement.
I cannot imagine the various errors and faults necessary to result in these readings.
I understand the circuits, from DECADES of experience, technician based and ENGINEERING based. I have seen countless fault conditions in these amps.
It literally "doesn't add up".
Took a little while but I made it back!
pins 20, 21, 22; (pin 20 = (-0.012v); (pin 21 = (0.427v); (pin 22 = (35.95v)
pins 23, 24, 25; (pin 23 = (-0.012v); (pin 24 = (-0.762v); (pin 25 = (-35.85)
pins 30, 31, 32 (33); (pin 30 = (-0.012v); (pin 31 = (0.422v); (pin 32 = (35.97v); (pin 33 = (35.95v)
pins 26, 27, 28; (pin 26 = (-0.012v); (pin 27 = (-0.778v); (pin 28 = (-35.83v)
While adjusting the idle current pots (again)
read BETWEEN pin 27 (black) and pin 31 (red) for the left channel (vr1)
and pins 24 (black) and pin 21 (red) for the right channel (vr2).
** Pin 27 (black) and 31 (red) = (1.193v)
** Pin 24 (black) and 21 (red) = (1.195v)
I could adjust from Left to Right
VR1 - from 1.657v to 1.057v
I could adjust from Left to Right
VR2 -1.630v to 1.056v
Note: The outputs are still out of the unit and it's on the bench. If I did anything wrong above let me know and I will do it again.
pins 20, 21, 22; (pin 20 = (-0.012v); (pin 21 = (0.427v); (pin 22 = (35.95v)
pins 23, 24, 25; (pin 23 = (-0.012v); (pin 24 = (-0.762v); (pin 25 = (-35.85)
pins 30, 31, 32 (33); (pin 30 = (-0.012v); (pin 31 = (0.422v); (pin 32 = (35.97v); (pin 33 = (35.95v)
pins 26, 27, 28; (pin 26 = (-0.012v); (pin 27 = (-0.778v); (pin 28 = (-35.83v)
While adjusting the idle current pots (again)
read BETWEEN pin 27 (black) and pin 31 (red) for the left channel (vr1)
and pins 24 (black) and pin 21 (red) for the right channel (vr2).
** Pin 27 (black) and 31 (red) = (1.193v)
** Pin 24 (black) and 21 (red) = (1.195v)
I could adjust from Left to Right
VR1 - from 1.657v to 1.057v
I could adjust from Left to Right
VR2 -1.630v to 1.056v
Note: The outputs are still out of the unit and it's on the bench. If I did anything wrong above let me know and I will do it again.
Last edited:
Davey; Hope you are aware of the Bias and offset pots being reversed in the manual. See this link for info. correct procedure is in 1st post . http://www.audiokarma.org/forums/showthread.php?t=464083
\
OK, first and foremost, leave the outputs out of the unit for now. Once I judge the unit and it's readings ready for outputs I will tell you explicitly that we are ready to try it with outputs again.
pins 20, 21, 22; (pin 20 = (-0.012v); (pin 21 = (0.427v); (pin 22 = (35.95v)
pins 23, 24, 25; (pin 23 = (-0.012v); (pin 24 = (-0.762v); (pin 25 = (-35.85)
0.427 + 0.762 = 1.189
pins 30, 31, 32 (33); 30 = (-0.012v); (pin 31 = (0.422v); (pin 32 = (35.97v); (pin 33 = (35.95v)
pins 26, 27, 28; (pin 26 = (-0.012v); (pin 27 = (-0.778v); (pin 28 = (-35.83v)
0.422 + 0.778 = 1.200
** Pin 27 (black) and 31 (red) = (1.193v)
VR1 - from 1.657v to 1.057v
** Pin 24 (black) and 21 (red) = (1.195v)
VR2 -1.630v to 1.056v
Now, overall these are good driver voltage TOTALS, HOWEVER (there's always a "BUT") the outputs don't care about the totals, they care about the emitter to base voltages.
so this: 0.427 + 0.762 = 1.189 is "nice" BUT 0.427 says one transistor is not turned on and the other transistor is turned on too hard (I would expect that transistor to be damaged if it took 0.778 volt to turn it on)
BECAUSE
pins 30 and 26 are at -0.012 volts
0.422 + 0.012 = 0.434 is what the base to emitter voltage for Q11 is.
0.778 - 0.012 = 0.766 is what the base to emitter voltage for Q13 is.
That's an imbalance, and that causes distortion.
BUT (another but)
BOTH channels are behaving the same.
so instead of doing the:
pins 20, 21, 22; (pin 20 = (-0.012v); (pin 21 = (0.427v); (pin 22 = (35.95v)
pins 23, 24, 25; (pin 23 = (-0.012v); (pin 24 = (-0.762v); (pin 25 = (-35.85)
and
pins 30, 31, 32 (33); 30 = (-0.012v); (pin 31 = (0.422v); (pin 32 = (35.97v); (pin 33 = (35.95v)
pins 26, 27, 28; (pin 26 = (-0.012v); (pin 27 = (-0.778v); (pin 28 = (-35.83v)
measurements referenced to ground, do them
to pin 3 (20,21,22,23,24,25)
and pin 17(27,28,29,30,31,32)
BUT FIRST, with power off check R23, r24, r25, r26 resistances, all should be within 5% of 150 ohms.
Basically, this should be happening (with ground based readings)
pins 20, 21, 22; (pin 20 = (-0.012v); (pin 21 = ( 1.189 / 2 - 0.012 = 0.582v); (pin 22 = (35.95v)
pins 23, 24, 25; (pin 23 = (-0.012v); (pin 24 = -( 1.189 / 2 + 0.012 = 0.607v); (pin 25 = (-35.85)
pins 30, 31, 32; (pin 30 = (-0.012v); (pin 31 = (1.200 / 2 - 0.012 = 0.588v); (pin 32 = (35.97v))
pins 26, 27, 28; (pin 26 = (-0.012v); (pin 27 = -(1.200 /2 +0.012 = 0.612v); (pin 28 = (-35.83v)
The only other possibility is that wherever you got a ground point for the black dmm lead, it was at -0.762 - . 427 = +0.167v. with respect to true ground.
calculating THAT:
-0.762 + .167 = 0.594
0.427 + .167 = 0.595
gives us reasonable base voltages for the outputs.
Last edited:
Got that. Posted a reference to the link in #47.
* BUT FIRST, with power off check R23, r24, r25, r26 resistances, all should be within 5% of 150 ohms.
* All the resistors for R23-R26 were replaced with new ones from mouser: 660-MOS1/2CT52R151J since I have posted this problem to the forum. They are KOA Speer - Metal Oxide Resistors RSS1/2 150 5%TR.
I measured them again and all measure above 149.5 ohms.
Please clarify. Do you want me to put the black lead to pin 3 and pin 17 and measure voltages 20,21,22,23,24,25 and 27,28,29,30,31,32 or place the black probe to ground and measure:
pin 3 (20,21,22,23,24,25)
and pin 17 (27,28,29,30,31,32) ?
Voltages from all pins above to ground are:
Pins
3 (-0.012v)
20 (-0.013v)
21 ( 0.400v)
22 ( 36.22v)
23 (-0.013v)
24 (-0.789v)
25 (-36.77v)
Pins
17 (-0.012v)
27 (-0.792v)
28 (-36.05v)
29 (-36.05v)
30 (-0.012v)
31 ( 0.415v)
32 ( 36.17v)
Looks like I am going to have to order a new output transistor socket. I broke off one of the internal pins on one side while trying to insert the transistor. The service manual says it is part number AKH-001-0 but I can not find a reference to one online. Anyone know the modern part number so I can order and replace it?
This has nothing to do with my original problem. I just now did this.
This has nothing to do with my original problem. I just now did this.
Last edited:
What wasn't clear about that? Strike one - let others explain what that means to you.
The socket, it's going to have to come from a donor unit (ebay - classicstereoparts no affiliation), not many of those sockets of that style are still being made new.What they DO make, I have found, is constructed quite differently and very non-compatible with the heat sinks. It has to do with the locating bumps. And they will generally ding you from 4 to 30 bucks APIECE!!! Look at mouser and digikey..
And the readings:
they were to connect the black probe to pin 3 for one channel and measure voltages on pins 26,27,28,30,31,32 with the red probe
then connect the black probe to pin 17 for the other channel and measure voltages on pins 20,21,22,23,24,25 with the red probe
Getting the desired circuit readings with you is like hammering bent nails blindfolded. I have to trick you to get the probes where I need them and I have for the life of me not figured out your blind spot that I have to carefully tip-toe around, and over the years I have had plenty of practice with that.
The "feedback" system at the front end of the amp (q1,q3) tries to achieve a balanced condition in the amplifier. The need for the balance is mandated by the split nature of the push-pull amplifier. The necessary internal operational voltage(s) are generated internally, it starts as 2.4 volts developed across D1 and VR1, then is split into +1.2v and -1.2v when referenced to the output (which ISN'T ground - it only PLAYS one on TV and should be near 0.000v and adjustable) by the characteristics of of the q7 - q9 driver stage. They should be roughly symmetrical - with a loss of 0.6v across the stage, from 1.2v (positive or negative depending upon which it is) at the base (INPUT) of the driver transistors to 0.6v at their emitters (OUTPUT).
Your voltages are reported as 0.4 volt and 0.8 volt. IF they had been reported as 0.580 and 0.620 that would have been an acceptable variation in transistors, 20 millivolts is reasonable. However it is 200 millivolts which is NOT acceptable, in fact for both transistors to be conducting equally it is impossible. ASSUMING the transistors are conducting as physics dictates, the output of the amplifier would be at -0.200 volts.
Thus I question your frame of reference - the relationship of the voltages to ground which DOES NOT MATTER, and the relationship of the voltages to the OUTPUT connection of the amplifier which is all important. There are certain relationships between the voltages in an operating push-pull amp, and troubleshooting and repair is the process of bringing the non-conforming voltages back into conformity by replacing malfunctioning parts.
I have to TRUST your measurements, and all the "QC" that I have built into the process I have used over the years repeatedly reinforces the belief that the readings are wrong, ESPECIALLY when there is insistence that one channel is operating correctly, in the face of frankly impossible readings.
For an operating transistor to run on 0.4 volts base bias referenced to the emitter voltage - just plain impossible.
plus 0.6v base to emitter (not ground - ground is a convenience) for a npn transistor,
minus 0.6v base to emitter (not ground - ground is a convenience) for a pnp transistor.
Variations for this generally run in the tens of millivolts, while multiple hundreds of millivolts is flat out impossible.
A transistor is a bit of a funny device.
First until it is turned on, it is a voltage type device, requiring 0.6v or so to "light it up"
THEN
it becomes a current ratio device, where the increase or decrease of base to emitter current controls or affects a MUCH LARGER Collector to Emitter current.
Thus there is a sort of minimum current to keep things alive - and in the circuit of the power amplifier - it is called the idle current.
Dropping BELOW the idle current causes a distortion of the audio signal, as the changes in the audio signal DO NOT get through the amp, they are chopped off.
Thus the 1.2volts I spoke of, when it enters the q7 driver stage, LOSES 0.6v due to the physics of the driver transistor, coming out at 0.6v.
WHY do we want 0.6v here?
To turn on the QA11 output transistor! IT gets 0.6v at IT'S base, and loses another 0.6v, resulting at 0v at the emitter.
Then we turn it up a bit with vr1 until enough current is flowing through q11 AND R31 (0.5 ohms) that there is 10 millivolts at Q11's emitter. coupled with an additional 10 millivolts across r33 (0.5 ohms) at q13's emitter gives you your 20 millivolt idle current specification.
Thus there is 20 milliamps of current flowing from +36v, through q11, R31, r33 and q13 to -36 volt.
Then stick a speaker between R31 and R33 (pin 3) TO GROUND, and when the audio voltage at the amplifier input rises - saying the speaker cone should move out a bit, Q11 turns on a bit more and the extra current flows from pin 3 to ground, THROUGH the speaker, pushing the cone forward.
when the audio input voltage DROPS negative then Q13 turns on (q11 has also turned off) pulling the cone backward.
Last edited:
It's 5:55, and I did not see your 5:15 post, nor the 5:26 edit. I just finished the post, while my browser ate a last paragraph.
I suggest a re-read, perhaps study to assist your analysis. Better data to operate on.
Maybe we can get on the same page. Get this amp FIXED. Then I can devote more concentration into getting my toilet and car fixed.
I suggest a re-read, perhaps study to assist your analysis. Better data to operate on.
Maybe we can get on the same page. Get this amp FIXED. Then I can devote more concentration into getting my toilet and car fixed.
Mark:
With your help I did get the amp fixed. It is now running fine. Both voltages on both sides are performing without distortion.
Earlier in one of your posts you said it sounds like a bad transistor.
QUOTE: so this: 0.427 + 0.762 = 1.189 is "nice" BUT 0.427 says one transistor is not turned on and the other transistor is turned on too hard (I would expect that transistor to be damaged if it took 0.778 volt to turn it on)/QUOTE
So then I pulled the outputs. They were new and checked good. I also pulled and checked the Q7 and Q9 and Q8 and Q10 transistors and they checked good. They were also new.
I installed a new TO-3 socket I had in stock from another project I did not know I had. Since it was on the left channel I replaced the Q7 and the Q9 transistors with new ones. I fired it up on my DBT and light was completely off unlike before where I could see a slight dim. I adjusted the VR3 and VR4 to 0v. I was able now to adjust the voltages on VR1 and VR2 to 20mA.
Relay kicks in within 3-4 seconds.
I plugged it directly in to the AC and set the voltages again. Connected speakers and sounds excellent.
I could not have done this without your dedication and time you spent helping me overcome this problem. I truly appreciate it. :thmbsp:
