• 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 Powers on, but no audio output...

Diodes:
measured from a ground reference.
D111 +17.50v (not stabilized, steadily climbs) anode +51v cathode
D112 +17.50v (not stabilized, steadily climbs) anode +51v cathode
D117 -51v anode +16.20v (not stabilized, steadily climbs) cathode
D118 -51v anode +16.10v (not stabilized, steadily climbs) cathode

ABOVE READINGS HAVE BEEN EDITED...
 
SolutionRoom said:
1. turn on the unit, with the DBT connected.


Diodes:
measured from a ground reference.
D111 +17.50v (not stabilized, steadily climbs) anode OL, No Continuity cathode
D112 +17.50v (not stabilized, steadily climbs) anode OL, No Continuity cathode
D117 OL, No Continuity anode +16.20v (not stabilized, steadily climbs) cathode
D118 OL, No Continuity anode +16.10v (not stabilized, steadily climbs) cathode
Click to expand...

OK, there are TWO voltage measurements for the circuit, CONVIENENTLY LOCATED AT the pins of the diode.
+17.50 is a perfectly acceptable reading, +16.20 is a surprising but acceptable reading.

But OL? >>>> I looked at the picture of your meter - it has a 20 volt range and a 600 volt range. When it says OL, it means that you are asking it to read MORE than 20 volts, so click up to the next range, 600 volts, in your case.
This miscommunication of 0v for OL is why we spent so much time checking the capacitor and heatsink voltages. and here I though it had disappeared again

Plus, I think the D117 and D118 surprising readings are negative 16 volts.... look carefully for any indication of a negative sign in front of the numbers. If it really IS positive 16 volts - that's earthshaking.

I'll try this another way:

1. turn on the 3900 with the DBT connected.

What is the voltage when the red lead of the voltmeter is touched to the UNBANDED side of D111 when the black lead of the voltmeter is grounded?

What is the voltage when the red lead of the voltmeter is touched to the BANDED side of D111 when the black lead of the voltmeter is grounded?

What is the voltage when the red lead of the voltmeter is touched to the UNBANDED side of D112 when the black lead of the voltmeter is grounded?

What is the voltage when the red lead of the voltmeter is touched to the BANDED side of D112 when the black lead of the voltmeter is grounded?


What is the voltage when the red lead of the voltmeter is touched to the UNBANDED side of D117 when the black lead of the voltmeter is grounded?

What is the voltage when the red lead of the voltmeter is touched to the BANDED side of D117 when the black lead of the voltmeter is grounded?

What is the voltage when the red lead of the voltmeter is touched to the UNBANDED side of D118 when the black lead of the voltmeter is grounded?

What is the voltage when the red lead of the voltmeter is touched to the BANDED side of D118 when the black lead of the voltmeter is grounded?



And additionally, this is checking that the power supply voltages are making their way to the amplifier board:

What is the voltage when the red lead of the voltmeter is touched to Pin 1 when the black lead of the voltmeter is grounded?

What is the voltage when the red lead of the voltmeter is touched to Pin 3 when the black lead of the voltmeter is grounded?

What is the voltage when the red lead of the voltmeter is touched to pin 5 when the black lead of the voltmeter is grounded?

gazillion. turn the 3900 off. :D

meanwhile I'm analyzing the other readings.
 
markthefixer said:
OK, there are TWO voltage measurements for the circuit, CONVIENENTLY LOCATED AT the pins of the diode.
+17.50 is a perfectly acceptable reading, +16.20 is a surprising but acceptable reading.

But OL? >>>> I looked at the picture of your meter - it has a 20 volt range and a 600 volt range. When it says OL, it means that you are asking it to read MORE than 20 volts, so click up to the next range, 600 volts, in your case.
This miscommunication of 0v for OL is why we spent so much time checking the capacitor and heatsink voltages. and here I though it had disappeared again (Another amateur mistake. Previous post has been edited)

Plus, I think the D117 and D118 surprising readings are negative 16 volts.... look carefully for any indication of a negative sign in front of the numbers. If it really IS positive 16 volts - that's earthshaking. (NO Negative sign)

I'll try this another way:

1. turn on the 3900 with the DBT connected.

What is the voltage when the red lead of the voltmeter is touched to the UNBANDED side of D111 when the black lead of the voltmeter is grounded? (+17.50v and climbs)

What is the voltage when the red lead of the voltmeter is touched to the BANDED side of D111 when the black lead of the voltmeter is grounded? (+51v)

What is the voltage when the red lead of the voltmeter is touched to the UNBANDED side of D112 when the black lead of the voltmeter is grounded? (+17.5v and climbs)

What is the voltage when the red lead of the voltmeter is touched to the BANDED side of D112 when the black lead of the voltmeter is grounded? (+51v)


What is the voltage when the red lead of the voltmeter is touched to the UNBANDED side of D117 when the black lead of the voltmeter is grounded? (-51v)

What is the voltage when the red lead of the voltmeter is touched to the BANDED side of D117 when the black lead of the voltmeter is grounded? (+16.5v and climbs)

What is the voltage when the red lead of the voltmeter is touched to the UNBANDED side of D118 when the black lead of the voltmeter is grounded? (-51v)

What is the voltage when the red lead of the voltmeter is touched to the BANDED side of D118 when the black lead of the voltmeter is grounded? (+16.5v and climbs)



And additionally, this is checking that the power supply voltages are making their way to the amplifier board:

What is the voltage when the red lead of the voltmeter is touched to Pin 1 when the black lead of the voltmeter is grounded? (-18.40v and climbs)

What is the voltage when the red lead of the voltmeter is touched to Pin 3 when the black lead of the voltmeter is grounded? (+2.8mv at the 200m setting on the DMM)

What is the voltage when the red lead of the voltmeter is touched to pin 5 when the black lead of the voltmeter is grounded? (+17.3v and climbs)
Click to expand...

Hope this clears up some things further...

In the future, I won't make the "OL" mistake when measuring voltages. I will bump up the dial to read a higher setting. That one registered firmly in my brain housing group.
 
Remember what we did with the 470 ohm resistor? connecting it from pin 22 to ground?

1. disconnect pin 22 wire
2. turn on unit with dbt and measure pin 22 voltage with red meter lead, with the black meter lead grounded
3 turn off unit
4. get an orange 3300 ohm resistor you removed from the old board.
5. connect it between pin 22 and ground just like you did with the 470 ohm resistor
6. turn on unit with dbt and measure pin 22 voltage with red meter lead, with the black meter lead grounded
7. turn off unit.


I cannot find WHERE enough current is flowing (from those voltage readings across the resistors) to explain why the pin 22 voltage is being pulled down to between 16 and 18 volts.

Did you keep a record of what transistor numbers were removed from Q137, Q138, Q139 & Q140? I can see which way they are facing in the "before" picture. In the after picture it looks strange. But even THAT shouldn't be able to pull down the 54 volts to under 20v.

Check that R194 and R197 are both 100,000 ohms. Use an ohmmeter.

8. Then CUT the jumper wire next to R102, the 470k ohm resistor.
9. reconnect the pin 22 wire to pin 22
10. turn on unit with dbt and measure pin 22 voltage with red meter lead, with the black meter lead grounded
11. turn off unit. report readings.
 
markthefixer said:
Remember what we did with the 470 ohm resistor? connecting it from pin 22 to ground?

1. disconnect pin 22 wire
2. turn on unit with dbt and measure pin 22 voltage with red meter lead, with the black meter lead grounded
3 turn off unit
4. get an orange 3300 ohm resistor you removed from the old board.
5. connect it between pin 22 and ground just like you did with the 470 ohm resistor
6. turn on unit with dbt and measure pin 22 voltage with red meter lead, with the black meter lead grounded
7. turn off unit. report readings.

I cannot find WHERE enough current is flowing (from those voltage readings across the resistors) to explain why the pin 22 voltage is being pulled down to between 16 and 18 volts.

Did you keep a record of what transistor numbers were removed from Q137, Q138, Q139 & Q140? I can see which way they are facing in the "before" picture. In the after picture it looks strange. But even THAT shouldn't be able to pull down the 54 volts to under 20v.
Click to expand...

Doing the readings now. Yes, I have a list of what was removed. I will have to put it together cleanly if you need it. Some of the new transistors do face in a different way than the old ones, but I used all the Data Sheets for the new ones and placed the "E" in the board marking "E" spots. Unless the board is marked wrong, they are in right. I did not verify them with the schematic emitter spots however, nor with the old ones. I simply removed the old ones, and installed the new ones with the emitter lead into the hole marked "E" on the PCB. Like I said, I used the data sheets for ALL of them to determine the "E". (ALL TRANSISTORS HAVE NOW BEEN VERIFIED (AS OF 12-11-08), SEE POST NUMBER 162 FOR COMPLETE LISTING)
 
markthefixer said:
Remember what we did with the 470 ohm resistor? connecting it from pin 22 to ground?

1. disconnect pin 22 wire (Done)
2. turn on unit with dbt and measure pin 22 voltage with red meter lead, with the black meter lead grounded (+54vdc)
3 turn off unit
4. get an orange 3300 ohm resistor you removed from the old board.
5. connect it between pin 22 and ground just like you did with the 470 ohm resistor (Done)
6. turn on unit with dbt and measure pin 22 voltage with red meter lead, with the black meter lead grounded (+54vdc)
7. turn off unit.

The receiver powers on with soft start relay both ways. Without 3.3k and with 3.3k resistor

I cannot find WHERE enough current is flowing (from those voltage readings across the resistors) to explain why the pin 22 voltage is being pulled down to between 16 and 18 volts.

Did you keep a record of what transistor numbers were removed from Q137, Q138, Q139 & Q140? I can see which way they are facing in the "before" picture. In the after picture it looks strange. But even THAT shouldn't be able to pull down the 54 volts to under 20v.

Check that R194 and R197 are both 100,000 ohms. Use an ohmmeter. (They register about 90,000 while in circuit and do not stay steady) They are definitely 100k resistors, but to test them out of circuit would require desoldering. Do you need me to do that?)

8. Then CUT the jumper wire next to R102, the 470k ohm resistor. (Done)
9. reconnect the pin 22 wire to pin 22 (Done)
10. turn on unit with dbt and measure pin 22 voltage with red meter lead, with the black meter lead grounded (+16.5v and climbs as before)
11. turn off unit. report readings.
Click to expand...

This SX-3900 is brutal!
 
Alright Mark. Not sure if you are still chuggin away at this at this late hour, I am going to turn in for the night. Been a looooong day for me. I will pick this back up first thing in the morning because, I am relentless and tireless for challenges such as this AND... as I said before...

THIS SX-3900 WILL NOT WIN ME!!!
 
No, reading 90,000 in circuit is fine, and not the source of the current drain that's pulling down the +54 volts.

When you installed all those resistors, did you measure each of them before installation for the correct value?

And just to be sure, when the +54v is pulled down to 20v the soft start relay doesn't actuate.

1. Now is the time to connect the wire that should go to pin 22, to the 470 ohm resistor instead, and the other side of the 470 ohm resistor should be connected to pin 22.
2. power up unit with dbt. does the soft start relay actuate?
3. measure pin 22 voltage (to ground).
4. measure the voltage (to ground) on the other side of the 470 ohm resistor where the wire that SHOULD be on pin 22 is connected.
5. measure the voltage ACROSS the 470 ohm resistor black probe to pin 22, red probe to the wire on the other side.
6. turn off unit.

"see" you after noon .
 
Si Senor... If you come up with any more things in the process, please post them. I will give them a shot early in the morn. Then I will be back at it in the late afternoon. Got a few tracking sessions to do tomorrow and will be tied up some in the early afternoon. Thanks a ton!
 
yes, a thought occurs - a remote possibility, did q103 / q105 (and q104/q106) possibly get swapped?
inserted backwards wouldn't happen because of symmetry.
 
markthefixer said:
No, reading 90,000 in circuit is fine, and not the source of the current drain that's pulling down the +54 volts.

When you installed all those resistors, did you measure each of them before installation for the correct value? (No. Unfortunately)

And just to be sure, when the +54v is pulled down to 20v the soft start relay doesn't actuate. (Correct, it DOES NOT actuate)

1. Now is the time to connect the wire that should go to pin 22, to the 470 ohm resistor instead, and the other side of the 470 ohm resistor should be connected to pin 22. (Done)
2. power up unit with dbt. does the soft start relay actuate? (Yes, soft start relay DOES actuate)
3. measure pin 22 voltage (to ground). (+53vdc)
4. measure the voltage (to ground) on the other side of the 470 ohm resistor where the wire that SHOULD be on pin 22 is connected. (+12.24vdc)
5. measure the voltage ACROSS the 470 ohm resistor black probe to pin 22, red probe to the wire on the other side. (-41vdc)
6. turn off unit.

"see" you after noon .
Click to expand...

*** NOTE: The jumper next to R102 is still cut.

_________________________________________________
_________________________________________________

markthefixer said:
yes, a thought occurs - a remote possibility, did q103 / q105 (and q104/q106) possibly get swapped?
inserted backwards wouldn't happen because of symmetry.
Click to expand...

No. The Q103-Q106 are all in as they were originally. A CLOSE look at the photos (old vs new) confirms this. I have a VERY high res monitor and can zoom in on a pimple on an ant's ass. I also verified them by the schematic, part itself and the parts listing in the SM. They are the original parts put back in after I tested them. I removed only one at a time and then re-installed them on by one.

ALTHOUGH... I am not 100% sure I tested them properly. I did use your diode testing procedures, but was not as proficient at it as I am now.

I just metered them again for continuity (shorts) WITH POWER OFF. They have no shorts from lead to lead and register a range that indicates they are likely good. BUT, to be 100% sure, I would need to pull them back out and test them out of circuit, AND with my newly acquired testing skills (from testing all of the output transistors) I could get solid confirmation. IF you think they are suspect and it is necessary at this point. Please let me know.

As P/N verification, they are (OEM) as follows (P/N correction made):
Q103 & Q104 2SA979
Q105 & Q106 2SC2291

markthefixer said:
Did you keep a record of what transistor numbers were removed from Q137, Q138, Q139 & Q140? I can see which way they are facing in the "before" picture. In the after picture it looks strange. But even THAT shouldn't be able to pull down the 54 volts to under 20v.
Click to expand...

Yes, I kept good records and the old parts. I verified these old part numbers with the actual physical parts themselves. I have each board's old parts in individual ziplock bags for easy ID and retrieval.

Q137, Q138, Q139 were 2SC2575 (old original part)
(I was NOT able to find a Data Sheet online for these old 2SC2575 transistors, so could not verify the emitter lead and their overall arrangements) I could test them and determine the lead arrangements if necessary. Let me know.)

Q140 was 2SC1384 (old original part)
(I found a Data Sheet online for this old transistor and the lead arrangements and emitter lead are the same as the one I replaced it with)

All of these transistors, Q137, Q138, Q139 and Q140 were replaced with KSC2383 (512-KSC2383YBU) per your recommendations.

_______________________________________________
_______________________________________________

I posted three photos at the links below. They show the old and the new power amp boards separately, then one picture with them side-by-side, zoomed in and marked up for the Q137, Q138, Q139 & Q140 transistors. (As well as a couple of other notables)

It is obviously clear that the 2SC2575 either had a reversed lead arrangement from the KSC2383, or they were all originally installed opposite to the PCB markings...???

I verified the E markings on the board with the schematic. They correspond for the Q137, Q138, Q139, Q140 and Q125, Q126.

http://www.solutionroom.com/SX3900/PwrAmpOld.jpg
http://www.solutionroom.com/SX3900/PwrAmpNew.jpg
http://www.solutionroom.com/SX3900/PwrAmpOldandNew.jpg

_______________________________________________________________
_______________________________________________________________

There are a total of 12 transistors that are on the Power Amp board that are 180 degrees flipped from the original. I have not been able to verify ALL of them, but will. Here is a picture showing the ones (pink dots) that are flipped. I also marked 2 fried resistors on the old board and put arrows indicating the Q125 and Q126 that were broken/sheared off. (The image is quite large, at 2.2mb)

http://www.solutionroom.com/SX3900/PwrAmpFlippedQs.jpg

THIS POST IS COMPLETE. ALL FLIPPED TRANSISTORS IN THE PICTURE, AND EVERY OTHER TRANSISTOR ON THE BOARD HAS BEEN FULLY VERIFIED. SEE POST NUMBER 162 FOR DETAILS.
 
"he says as gently as he can":

Q103 & Q104 are 2sa979 the dual transistor with 5 leads, and a common emitter with b1 c1 e1+2 c2 b2 layout
Q105 & Q106 are 2sc2291 the dual transistor with 5 leads and a common base with e1 c1 b1+2 c2 e2 layout. I found it in the sx-d7000's manual.

We kept the originals. I just didn't go through all the pictures to see if it showed somewhere. If we replaced them each with TWO transistors that we made some attempt to match:
Q103 & Q104's 2sa979's would be replaced with four KSA992's
Q105 & Q106's 2sc2291's would be replaced with four KSC1845's

Also for clarity's sake, the 2sc2575's pinout is BCE with the flat facing the observer and the leads pointing down. It is hiding in the 3900 manual I downloaded.

Now on to happier notes: it's very good that we have +53v at pin 22, AND the soft start relay has activated!!

My next post will be a list of voltage readings. And I do remember that the protection circuit has been de-energized by the cut wire jumper next to R102.

As always, when you turn it off after a session, look for hot or warm components on the amp board, although that 470 ohm resistor will now be the recipient of most of the heat (3.57 watts). 41 volts across 470 ohms is 87mA and 3.57 watts.
The remaining 12.24 volts at 87mA is 1.07 watts. Most of that 1.07 watt is being dissipated by something.
 
markthefixer said:
"he says as gently as he can":

Q103 & Q104 are 2sa979 the dual transistor with 5 leads, and a common emitter with b1 c1 e1+2 c2 b2 layout
Q105 & Q106 are 2sc2291 the dual transistor with 5 leads and a common base with e1 c1 b1+2 c2 e2 layout. I found it in the sx-d7000's manual.

We kept the originals. I just didn't go through all the pictures to see if it showed somewhere. If we replaced them each with TWO transistors that we made some attempt to match:
Q103 & Q104's 2sa979's would be replaced with four KSA992's
Q105 & Q106's 2sc2291's would be replaced with four KSC1845's

Also for clarity's sake, the 2sc2575's pinout is BCE with the flat facing the observer and the leads pointing down. It is hiding in the 3900 manual I downloaded.

Now on to happier notes: it's very good that we have +53v at pin 22, AND the soft start relay has activated!!

My next post will be a list of voltage readings. And I do remember that the protection circuit has been de-energized by the cut wire jumper next to R102.

As always, when you turn it off after a session, look for hot or warm components on the amp board, although that 470 ohm resistor will now be the recipient of most of the heat (3.57 watts). 41 volts across 470 ohms is 87mA and 3.57 watts.
The remaining 12.24 volts at 87mA is 1.07 watts. Most of that 1.07 watt is being dissipated by something.
Click to expand...

So that is good that the SC2575 has BCE, that means they are opposite of the KSC2383 that were put in, so explains the flipping.

I know you mentioned the 5 lead Q103-Q106, and we did keep the originals in place... Are you saying to go to Plan B and do the paired new ones in their place???

I will be offline for a bit till later this afternoon. Doing some drum tracking sessions... Ah work.....

Please let me know what you come up with from all the research I have done..... Thanks
 
OK, back in the saddle. Anxious to get to the next step. Once again, please let me know what I can do to troubleshoot further. I will verify Q orientations. ALL Q TRANSISTOR ORIENTATIONS HAVE BEEN VERIFIED, SEE POST NUMBER 162 FOR ALL DETAILS.

This SX-3900 is waking up!
 
SolutionRoom said:
I know you mentioned the 5 lead Q103-Q106, and we did keep the originals in place... Are you saying to go to Plan B and do the paired new ones in their place???
Click to expand...

No, all I wanted was to be sure that the transistors installed in the Q103 & Q104 positions are 2sa979's and that the transistors installed in the Q105 & Q106 positions are 2sc2291. It's impossible to put those dual transistors in backwards - so I wasn't worried about that.

Across the amplifier's printed circuit board, the 2sc2291's are the closest to the OUTSIDE edges of the p.c.b. and the 2sa979's are closest to the CENTER of the p.c.b.
 
You must log in or register to reply here.
Back
Top Bottom