Well, let's see...
Q6 and Q7 are voltage limiters so if the circuit is working are they limiting the current through RN35 1 & 2? This would explain why the working board has a lower voltage, perhaps? And it would suggest that Q6 and Q7 are not operating in the non-working boards -- for whatever reason.
Any thoughts? Am speculating, obviously.
Q6 and Q7 are voltage limiters so if the circuit is working are they limiting the current through RN35 1 & 2? This would explain why the working board has a lower voltage, perhaps? And it would suggest that Q6 and Q7 are not operating in the non-working boards -- for whatever reason.
Any thoughts? Am speculating, obviously.
I also measured the LM385s (again installed on board) and pins 2-3 with probes either way gives me 0.083 on diode, so very little voltage drop. Not sure if this means anything by itself, but have you checked to see if they measure open? Looking at front of lm385, pins would be 1-2-3 from left to right.
Measured 2 non-working boards and the working one they are all 0.118 for both Q6 and 7. Don't know if that is a sufficient difference to indicate a problem but am learning that the road to hearing a M2x is a long one...(LOL). Thanks, bullittstang. BTW, measured the working RN35 1 and 2 in both directions and, like yours, they read the same both ways although my current is a bit higher.
Last edited:
I would think, out of circuit, that your measurements for the 4N35 are OK, have a look at the data sheet.
You measured 1.1V for 1-2, that's forward voltage for the Infrared Opto LED - sounds reasonable and according to data sheet page 3 that I attached a few pages back.
Also, according to data sheet, 6-4 would be Basis-Emitter of the opto transistor - 0.71V sounds OK as well.
If you look at the schematic, in-circuit the measurements will be different because of the outside loop R10-R13-R14-R12-R11.
And I don't know how the LM385 behave in-circuit.
In short: I think your un-mounted 4N35, which you measured, are measuring as expected.
Regards, Claas
You measured 1.1V for 1-2, that's forward voltage for the Infrared Opto LED - sounds reasonable and according to data sheet page 3 that I attached a few pages back.
Also, according to data sheet, 6-4 would be Basis-Emitter of the opto transistor - 0.71V sounds OK as well.
If you look at the schematic, in-circuit the measurements will be different because of the outside loop R10-R13-R14-R12-R11.
And I don't know how the LM385 behave in-circuit.
In short: I think your un-mounted 4N35, which you measured, are measuring as expected.
Regards, Claas
Claas: Let's play "what do we know."
1. We know the circuit board itself is probably correct -- ordered from DIY and others have successfully built from it.
2. The power supply has been thoroughly checked and reported on above and is providing stable +/- 25.8 vdc
3. I know that the build was very carefully done and shows no problem signs -- the pictures above indicate the soldering accuracy and all parts are correctly applied.
4. We know that the resistors are correct and have been checked carefully 3 times or more.
5. We know that the diodes check out: 0.62 vdc in the positive polarity and OL in the negative.
6. We know that the voltage limiters Q6 and Q7 have the same 0.118 diode reading as the working board -- so they are likely correct.
7. We know that the 4N35 has an incorrect R value of 220 from R14 to Pin 2 and that the working circuit shows a correct 332 R -- and that this is a "mystery."
8. We know that the diode value for pins 1-2 is 0.440 vdc in the working board and 1.07 in the non-working ones. And that the working board has the same vdc in both 1-2 directions and the non-working one has 1.07 vdc positive and OL in the reverse direction.
9. We know that the diode values for non-working 4N35s are the same for a unconnected Opto as they are for the in-circuit, non-working ones. This suggests that, somehow, the circuit is not connecting correctly to the 4N35.
10. The only other regulation loop parts are Q1 and Q2. Q2 is probably the one that is starting to smoke through excessive voltage. And the voltage through R13 and R14 just keeps increasing when turned on and, therefore, 4N35 is not doing its job.
11. The two 4N35s that worked are from an August Mouser order and the 4 that were ordered in September all seem to be non-working. You suggest that the 1.07 value indicates the Opto is within spec.
And from this information, I draw the following conclusions.
If 4N35 is a good part, it is not operating within the circuit. Since the soldering is correct, the only other option seems to be Q1 and Q2. Do you agree?
All 3 non-working boards are using the same 4N35s. Have ordered a bunch more today after talking with Mouser and have also added substitutes by Broadcom and Onsemi. Will try these in the board to see if the correct 1-2 value of around 0.4 VDC can be achieved in both directions.
Have also ordered additional Q1s and Q2s to have as needed.
Claas & bullittstang: don't know what else to try. Why the 4N35 is not creating the regulation loop appears to be the central question but this non-function also appears to be a "mystery."
1. We know the circuit board itself is probably correct -- ordered from DIY and others have successfully built from it.
2. The power supply has been thoroughly checked and reported on above and is providing stable +/- 25.8 vdc
3. I know that the build was very carefully done and shows no problem signs -- the pictures above indicate the soldering accuracy and all parts are correctly applied.
4. We know that the resistors are correct and have been checked carefully 3 times or more.
5. We know that the diodes check out: 0.62 vdc in the positive polarity and OL in the negative.
6. We know that the voltage limiters Q6 and Q7 have the same 0.118 diode reading as the working board -- so they are likely correct.
7. We know that the 4N35 has an incorrect R value of 220 from R14 to Pin 2 and that the working circuit shows a correct 332 R -- and that this is a "mystery."
8. We know that the diode value for pins 1-2 is 0.440 vdc in the working board and 1.07 in the non-working ones. And that the working board has the same vdc in both 1-2 directions and the non-working one has 1.07 vdc positive and OL in the reverse direction.
9. We know that the diode values for non-working 4N35s are the same for a unconnected Opto as they are for the in-circuit, non-working ones. This suggests that, somehow, the circuit is not connecting correctly to the 4N35.
10. The only other regulation loop parts are Q1 and Q2. Q2 is probably the one that is starting to smoke through excessive voltage. And the voltage through R13 and R14 just keeps increasing when turned on and, therefore, 4N35 is not doing its job.
11. The two 4N35s that worked are from an August Mouser order and the 4 that were ordered in September all seem to be non-working. You suggest that the 1.07 value indicates the Opto is within spec.
And from this information, I draw the following conclusions.
If 4N35 is a good part, it is not operating within the circuit. Since the soldering is correct, the only other option seems to be Q1 and Q2. Do you agree?
All 3 non-working boards are using the same 4N35s. Have ordered a bunch more today after talking with Mouser and have also added substitutes by Broadcom and Onsemi. Will try these in the board to see if the correct 1-2 value of around 0.4 VDC can be achieved in both directions.
Have also ordered additional Q1s and Q2s to have as needed.
Claas & bullittstang: don't know what else to try. Why the 4N35 is not creating the regulation loop appears to be the central question but this non-function also appears to be a "mystery."
Signal lost: First, spent an hour going over all of the post-Edcor connections with the schematic to make sure that all was connected properly. And they are all. However...
The mystery R14 to Pin 2 pathway acts like so. R14 goes direct to R12 (100ohm) and the R after is the correct 100 ohms. Then the signal splits and goes to Q7 (pin 3) and R11 (221ohm). The in to R11-R14 is the correct 100ohms as is the in to Q7, Pin 3. But the out of R11 is 221ohms rather than the correct 331. Measured by itself, R11 has the correct 221ohms. Q7 appears to be removing 100ohms from the pathway.
Suspect this is the central problem. Please offer any explanations for this incorrect 221 ohm reading at the output of R11.
As to the Power Supply, it works fine with the working channel and can supply that board OK. But, yes, the voltage drops when the boards are connected -- but not when it is measured as it arrives at the amp splitter box. There it remains the +/- 25.8 vdc. See post 5462 above for the tests done as requested
How do you test an incorrect PSU voltage drop under load?
Oh, and there are burned traces on the original, faulty left channel board (that started all the problems). But the new board with the voltage increase problem has none and is pristine.
The mystery R14 to Pin 2 pathway acts like so. R14 goes direct to R12 (100ohm) and the R after is the correct 100 ohms. Then the signal splits and goes to Q7 (pin 3) and R11 (221ohm). The in to R11-R14 is the correct 100ohms as is the in to Q7, Pin 3. But the out of R11 is 221ohms rather than the correct 331. Measured by itself, R11 has the correct 221ohms. Q7 appears to be removing 100ohms from the pathway.
Suspect this is the central problem. Please offer any explanations for this incorrect 221 ohm reading at the output of R11.
As to the Power Supply, it works fine with the working channel and can supply that board OK. But, yes, the voltage drops when the boards are connected -- but not when it is measured as it arrives at the amp splitter box. There it remains the +/- 25.8 vdc. See post 5462 above for the tests done as requested
How do you test an incorrect PSU voltage drop under load?
Oh, and there are burned traces on the original, faulty left channel board (that started all the problems). But the new board with the voltage increase problem has none and is pristine.
Last edited:
Hmmm ... looking at the schematic, a shorted Q7 would lead to the "removal of 100R" between R14 and Opto Pin2 
Also, if Q7 is a short, the voltage across the Opto LED can only come from the upper half (R13). No voltage over R14 makes it to the Optocoupler. This would mean the Opto LED gets only half of the voltage it is supposed to, and regulates to a much higher current than intended, as it only starts to light up at a much higher current through R13 ...
What about just replacing Q7 once more as a try ?
Also, you do get 100R between R13 nd Pin1 of Opto ? That would at least indicate the Q6 is not shorted.
Regards, Claas
Also, if Q7 is a short, the voltage across the Opto LED can only come from the upper half (R13). No voltage over R14 makes it to the Optocoupler. This would mean the Opto LED gets only half of the voltage it is supposed to, and regulates to a much higher current than intended, as it only starts to light up at a much higher current through R13 ...
What about just replacing Q7 once more as a try ?
Also, you do get 100R between R13 nd Pin1 of Opto ? That would at least indicate the Q6 is not shorted.
Regards, Claas
rayma: don't know what a silk screen is. The parts were added according to the designations on the boards themselves (are these the silk screens?) and are in different positions between boards, sometimes. All parts were added in the marked position -- checked multiple times. All of the continuity checks as requested above were done and are all correct as noted above. Three boards show the same 100ohm drop so it is common with the parts being used. Q6 and Q7 have been checked for continuity and shorts and there are none. There is no reason to replace Q7 on one board when it is showing the same results on three, right? And, yes, you do get 100ohms between R13 and pin 1. Thanks as always, helpful souls.
Last edited:
chede: would like to stress that in my test above, tested R14 to the output of R11. It is the same 221 value as to pin 2. So that means the Q7 must be making the 100ohm reduction at the shared junction, right? Not sure if a problem with 4N35 could cause the bad reading from R14 to R11. That's why I speculated it might be Q7.
And this same problem occurs with the common parts in both left and right boards. So this problem, at least, is not related to any differences between boards.
And this same problem occurs with the common parts in both left and right boards. So this problem, at least, is not related to any differences between boards.
The silk screen contains the component outlines, values, reference designators, etc. Markings can be wrong
(+ on the wrong capacitor terminal, etc). If the pc boards themselves have been proven correct by other users,
then remaining are wrong parts, or swapped parts, or other assembly errors. This includes the wiring to the
bad board. Bad parts directly from well known vendors are very unlikely, though they can be damaged by
handling or wrong installation.
(+ on the wrong capacitor terminal, etc). If the pc boards themselves have been proven correct by other users,
then remaining are wrong parts, or swapped parts, or other assembly errors. This includes the wiring to the
bad board. Bad parts directly from well known vendors are very unlikely, though they can be damaged by
handling or wrong installation.
rayma: understand now and thanks. This is why I did all of the continuity checks today. There are no problems -- none. These three boards all share the same run of 4N35s, 385s and mosfets from Mouser. Called them today and asked about potential errors in the 4N35s and they suggested it was unlikely. Still, have more and alternate parts coming.
Will make any tests you fine people suggest. Could remove Q7 from one of the boards if that is indicative.
Will make any tests you fine people suggest. Could remove Q7 from one of the boards if that is indicative.