Child care at the moment, but I will check and take pictures later. I poked around pre reading the last few posts this afternoon, ended up plugging it in again, and the board I thought was the problem was fine (at least non-pyrotechnic). Smoke poured out of the 7924 on the other board, however. I replaced it, and now that board is fine (getting proper v out of the regulator, at least) and I was able to zero out the pot. Of course, I checked the voltages on the first board, and am getting -31 out of the 7924, which means that it isn't working. At least I have one pcb working!
Unfortunately, I spoke too soon. The board that had allowed me to zero the pot did not work when I tried it again, and fried the resistors. No idea why. I am not sure what to do short of pulling everything out around the negative regulator ant trying again.
Probably the best thing would be to build the lightbulb tester!
Probably the best thing would be to build the lightbulb tester!
1. First, disconnect all power leads to one of the amp PCBs (no reason to torch BOTH PCBs, as you troubleshoot.
2. On the board with the toasted resistors, disconnect one end of R31. Same, with one end of R33 (to isolate the U2 regulator from the audio portion of the amp. (You might also want to disconnect the positive voltage rail, for now)
2. Remove the toasted R30 and R32.
3. With NO power on the unit, short out the terminals on C26 (to remove any residual charge).
4. Now measure the resistance across C26 (which also thereby checks the resistance of the U2 input to ground).
5. Let the meter "stabilize"--resistance may increase as C26 charges via your test meter.
6. If the resistance reading is zero. you have a C26 or U2 problem. If the reading is not zero (say several hundred ohms or more), continue to 7, next.
7. Reinstall a good R30 and R32. DO NOT yet reconnect R31 and R33!
8. Turn the power supply on, and check the negative rail voltages before and after U2. Let us know what you find......
(As an aside, it is ALWAYS wise, to build electronics "incrementally"---build the power supply and test. Next build the on-board regulators/filters and confirm voltage BEFORE you even populate the audio portion, etc, etc. 'Makes it a lot easier to locate a problem, and saves you from repeatedly smoking parts. Beilieve me--I've learned that.....)
2. On the board with the toasted resistors, disconnect one end of R31. Same, with one end of R33 (to isolate the U2 regulator from the audio portion of the amp. (You might also want to disconnect the positive voltage rail, for now)
2. Remove the toasted R30 and R32.
3. With NO power on the unit, short out the terminals on C26 (to remove any residual charge).
4. Now measure the resistance across C26 (which also thereby checks the resistance of the U2 input to ground).
5. Let the meter "stabilize"--resistance may increase as C26 charges via your test meter.
6. If the resistance reading is zero. you have a C26 or U2 problem. If the reading is not zero (say several hundred ohms or more), continue to 7, next.
7. Reinstall a good R30 and R32. DO NOT yet reconnect R31 and R33!
8. Turn the power supply on, and check the negative rail voltages before and after U2. Let us know what you find......
(As an aside, it is ALWAYS wise, to build electronics "incrementally"---build the power supply and test. Next build the on-board regulators/filters and confirm voltage BEFORE you even populate the audio portion, etc, etc. 'Makes it a lot easier to locate a problem, and saves you from repeatedly smoking parts. Beilieve me--I've learned that.....)
First, thanks for taking the time, I appreciate it very much. This makes a lot of sense. I followed your instructions, and the results were mixed:
Good: the resistance measured across the capacitors/U2 was 220 ohms on the pcb frying resistors.
Bad: I replaced the bad resistors, left the 31 and 33 out, and they fried again.
Weird: I checked the other pcb (which was giving -31 volts past U2), and got a resistance (same approach as above) of over 9000 ohms. ???
I am out of 10 ohm resistors, so testing will have to stop for a few days. I will buy some more u2's as well.
Thanks again for the help!
Good: the resistance measured across the capacitors/U2 was 220 ohms on the pcb frying resistors.
Bad: I replaced the bad resistors, left the 31 and 33 out, and they fried again.
Weird: I checked the other pcb (which was giving -31 volts past U2), and got a resistance (same approach as above) of over 9000 ohms. ???
I am out of 10 ohm resistors, so testing will have to stop for a few days. I will buy some more u2's as well.
Thanks again for the help!
OK. if you had R31 and R33 out, and still toasted the "new" R30 and R32, then the heavy current draw that's toasting these resistor must be from C26 (unlikely), a solder bridge from the negative rail to ground (near C26 and U2; possible), or a bad U2 (probable). The quiescent current flow for a good 7924 regulator is about 1-2 milliamps (not enough to toast R30 and R32). I'd put my money on a bad U2 (but double check your solder joints, too!).
If you are getting -31 volts "through" the U2 on the other board, you probably need to replace that one, too.
Re: soldering. I noticed a number of your pads on the PCBs had minimal solder, and poor solder flow (can create "cold solder joints"). When you are done trouble shooting and have a working amp, you might want to touch up a lot of the joints--it's just good practice to fill the pad, and make a good shiny flow. There's probably a number of good video "tutorials" on Youtube that can show you the technique.
Hint: inexpensive components like these resistors and regulators can be ordered in larger lots (say 10 or 20), and still not cost much. 'Never hurts to have a few around......
If you are getting -31 volts "through" the U2 on the other board, you probably need to replace that one, too.
Re: soldering. I noticed a number of your pads on the PCBs had minimal solder, and poor solder flow (can create "cold solder joints"). When you are done trouble shooting and have a working amp, you might want to touch up a lot of the joints--it's just good practice to fill the pad, and make a good shiny flow. There's probably a number of good video "tutorials" on Youtube that can show you the technique.
Hint: inexpensive components like these resistors and regulators can be ordered in larger lots (say 10 or 20), and still not cost much. 'Never hurts to have a few around......
Quick update: the new components came in. I had emptied the negative voltage regulator section (removing the capacitors, presumed faulty regulator, and resistors). Starting with the pcb that did not fry resistors (just gave -31v past the regulator, which is incorrect), I put them in one piece at a time and tested the voltage after each piece. No problems, everything went in fine, got -23.7V past the regulator. Was able to zero out the test pad. Good!
The second pcb was more interesting. I put in the first set of resistors intending to test the voltage, and again they burned--this is with no capacitors, including the big one, and no voltage regulator. Taking a close look at the back of the pcb, I noted that there was some stray solder around the contacts for the voltage regulator, so I removed that. I also tested the voltage going to the board. I will address this is a second. Following cleaning up the extra solder and dealing with the voltage issue noted below, the resistors did not fry, and I got the expected -33.5V at the resistors, so progress has been made there. I will put in the rest of the pieces this weekend.
The one odd thing--I tested to voltage both at the power supply and at the pearl case (which I was hooking up the pcb's to) after the resistors fried. I got +-33.5V at the psu as expected, but 0V on the positive (to ground) section, and -67V from ground to negative at the pearl case. I unplugged everything (basically, the homemade xlr cable) and plugged it in again, and got +-33.5 V at the pearl case. So, I am not sure if it was the higher than expected voltage that caused the problem, or the loose solder. Also, I am not sure why the voltage is odd. I have checked the jacks carefully, and they seem fine. I will check the xlr cable next (I didn't think to last night). Possible there is a short? I am not sure what would cause this.
Regardless, progress, and more things to look at. I will make sure to test the voltage at the pearl case before plugging things in until I figure out what happened.
Oh, and I will take your advice re: soldering--I am getting the hang of it, and will repair the worse joints once I get the boards fixed (and voltage!).
Thanks again.
The second pcb was more interesting. I put in the first set of resistors intending to test the voltage, and again they burned--this is with no capacitors, including the big one, and no voltage regulator. Taking a close look at the back of the pcb, I noted that there was some stray solder around the contacts for the voltage regulator, so I removed that. I also tested the voltage going to the board. I will address this is a second. Following cleaning up the extra solder and dealing with the voltage issue noted below, the resistors did not fry, and I got the expected -33.5V at the resistors, so progress has been made there. I will put in the rest of the pieces this weekend.
The one odd thing--I tested to voltage both at the power supply and at the pearl case (which I was hooking up the pcb's to) after the resistors fried. I got +-33.5V at the psu as expected, but 0V on the positive (to ground) section, and -67V from ground to negative at the pearl case. I unplugged everything (basically, the homemade xlr cable) and plugged it in again, and got +-33.5 V at the pearl case. So, I am not sure if it was the higher than expected voltage that caused the problem, or the loose solder. Also, I am not sure why the voltage is odd. I have checked the jacks carefully, and they seem fine. I will check the xlr cable next (I didn't think to last night). Possible there is a short? I am not sure what would cause this.
Regardless, progress, and more things to look at. I will make sure to test the voltage at the pearl case before plugging things in until I figure out what happened.
Oh, and I will take your advice re: soldering--I am getting the hang of it, and will repair the worse joints once I get the boards fixed (and voltage!).
Thanks again.
Yes, 6L6 pointed that out to me early. Re: the earlier post, the reading is stable at the PSU. I am pretty sure I had it all plugged in, but will check again.
This wins the prize. I needed to clean up the solder on one of the pins. That problem solved.
The rest was pretty anticlimactic--remaining pieces put in formerly pyrotechnic pcb, voltage tested properly, no smoke (!), zeroed out the test pad. Is there anything left but to put it together and try it?
Essentally that's it... If the PSU is behaving you can stuff the rest, double-triple check your soldering job and fire it up (one channel at a time...)
Oh, one other thing -
Lightbulb test lead.
Sound!
Well, crappy, and from only one channel, but sound. One of the boards isn't doing anything, so I will start there, see what I can find. No time to look at it today after plugging it in to the system. I will post some pictures when I get a chance to poke around a bit.
I appreciate that this isn't much, probably, but I am very pleased to have gotten this far.
Well, crappy, and from only one channel, but sound. One of the boards isn't doing anything, so I will start there, see what I can find. No time to look at it today after plugging it in to the system. I will post some pictures when I get a chance to poke around a bit.
I appreciate that this isn't much, probably, but I am very pleased to have gotten this far.
OK. Ponder this....the -67 volts you are getting is TWICE the voltage you expect! Why? Because you are measuring voltage ACROSS BOTH RAILS (ie.,
-33.5v to + 33.5v (= 67 volts). At the same time, your measurement on the amp chassis is +0v for the positive rail.
Check the pinouts on the cable from the power supply to the amp PCB. You have two wires crossed (probably the ground and the + rail wires). If you have a GOOD PCB in the other channel (sounds as though you do.....), use it as a reference.