Dead caps, what happens? Troubleshooting CMoy
Nov 6, 2002 at 11:42 PM Thread Starter Post #1 of 11

Born2bwire

25+ Member ;-)
Joined
Nov 11, 2001
Posts
3,700
Likes
13
I finally schlepped together something that's reminiscent of what Tangent has on his website (kudos!!!) that's a CMoy, but of course the Audio Gods aren't with me and it's not working. I was lucky enough to be able to borrow a Fluke Multimeter from the ECE department and it seems that the negative voltage side of the amp is grounded. The positive side carries the entire voltage when compared to ground. I've reheated the solder points and tried to clean them up, but this actually caused a complete grounding (before I was able to get .814 V difference). The Fluke has a nice capacitance detector and a continuity detector. The continuity confirms that the negative lead is in short with the ground and the capacitance measurer needs a good connection to get a measurement, but I think that the power capacitor on the negative side is a dud, I can't get a reading from it. Since I've been having a hard time getting the multimeter to correctly read the capacitance, I was wondering if a dead capacitor would result in a short or other behavior that would cause my problem?
 
Nov 7, 2002 at 6:36 AM Post #2 of 11
Measuring components on the board will often give confusing results. If there is another path from one DMM lead to the other lead besides the one through the component, you won't get a useful measurement. For the power supply caps in a CMoy amp, you also have the paths through the voltage divider and the LED. You will have to remove the caps from the board to test them. Actually you can cheat a bit and desolder just one leg.

But really, what's more likely than bad PS caps is a simple solder bridge. Check the board carefully. Don't just blindly reheat joints, _look_ at the board. If you can't see a problem, disconnect the power supply from the amp section and test it alone. If it works without the amp connected, the problem is in the amp section somewhere.

Also, if you're using a metal case and you've got the amp encased already, remove it -- just have the board and the panel components flopping around on their hookup wires. This minimizes variables while testing. While learning, you should either use nonconductive cases or refrain from encasing your amp until you know it's working.
 
Nov 7, 2002 at 3:30 PM Post #3 of 11
Ok, I've only been working on it outside of the enclosure for now, but I'll redo all of the solder points. I will admit that that's probably the problem, though I can't see any bridges, my soldering skills are bad and I've been using a silver solder for this one and it's been rather stubborn in affixing itself to the board, causing large globs of solder at some points where it finally took hold. I'll have to run by somewhere and get a desoldering braid, hope my ECE store has one, I hate having to spend all the money on shipping and handling for just a braid.
 
Nov 7, 2002 at 7:25 PM Post #4 of 11
RadioShack carries desoldering braid. Also consider getting a desoldering pump.

If that solder has a high silver content, and especially if it's a "no lead" type, you need a pretty hot iron to make that solder flow well. I would advise getting some simple 63/37 eutectic solder instead. Save that silver stuff for the hard jobs, and for when you have more experience.
 
Nov 8, 2002 at 5:44 PM Post #5 of 11
It's only a 2% silver, with lead and tin. It's a lil stubborn to melt, but the main prob is it doesn't like to adhere. I found out that the problem was that I heated the board too much at the cap and the metal came off of the board. Moved the cap to a different three-hole series and I got the power to be just fine
biggrin.gif
.


BUT:
Without the opamp, the voltage differences across the V+ and V- line voltages are fine. But when I put the opamp in, one side is like a +8.00 V and the other is about -.178 V. Grumble... Is this an indication that I've blown my opamp (which at this point wouldn't surprise me now)? When I measure resistances, one side has a high resistance and the other a low one, and this is the same if I remove the opamp from the circuit and measure the opamp resistances seperately.
 
Nov 8, 2002 at 8:33 PM Post #6 of 11
What's the mix of that solder? If it's 62/36/2, it should melt and flow fairly easily. If it's not flowing right, your iron isn't hot enough or your technique is bad.

On the opamp, maybe you did blow it. The simplest way to test this is to put a different chip in. If you don't have an op-amp, go to Radio Shack and get a TL082. That's a suitable stand-in for an OPA2132 for testing purposes.

The chip might also be oscillating badly, or you may have tied its output to ground or something ugly like that. In either of these cases, it can unbalance the power supply rails. Changing the chip won't help in these cases.
 
Nov 9, 2002 at 2:50 AM Post #7 of 11
Quote:

Originally posted by Born2bwire
It's only a 2% silver, with lead and tin. It's a lil stubborn to melt, but the main prob is it doesn't like to adhere. I found out that the problem was that I heated the board too much at the cap and the metal came off of the board. Moved the cap to a different three-hole series and I got the power to be just fine
biggrin.gif
.


Silver solder doesn't like solder pads very much for me. Try using soldering paste or soldering flux. It helps.
 
Nov 9, 2002 at 2:58 AM Post #8 of 11
Probably it's not the silver content of your solder, Dreamslacker, it's the rosin in its core. My main solder is Kester 44 62/36/2, and it wets pads easily. I use it with a ~500 degree F iron. (About 250 degrees C)
 
Nov 9, 2002 at 9:20 PM Post #9 of 11
Speak of the devil, that's the same solder that I picked up from my ECE store, huh, I didn't think they would carry anything worth your while
biggrin.gif
. I'll go see about picking up a Radio Shack opamp today. So, I blame my iron and my technique as you pointed out as reasons for my soldering probs.

Sigh, this is why I prefer to design logic systems than work with analog.
 
Nov 10, 2002 at 3:25 AM Post #10 of 11
Ok, I jogged out to Rat Shack and got two TL082 opamps. In addition, I removed the entire amplifier circuit and resoldered it on a new board just to be safe. Right, so, I checked the power supply, fine. I build the amplifier part, put in the opamp, test the V+ and V-, just peachy. So, I've built the entire amp, and for testing purposes, wired my input and output stereo jacks to the board. With nothing plugged in, the voltages test fine. Now, if I plug in my headphones to the output jack, the voltages shift a little. I'm getting 7.11 V and -1.56 V
confused.gif
. And I hear nothing when I run an input to it. I'm pretty sure the jacks are wired correctly because if I run a wire from point to point from the jacks, I hear fine. In fact, if I connect the input and output jacks of the IC socket, I can hear the music for both channels. I've hooked up the grounds of the jacks to the circuit's ground, is this the correct thing to do? I'll keep looking over the circuit as I go, but the thing is, the line continuity tests from the ground to the output channels show that the left and right are within .1 V tolerance in the signal, so I'm pretty sure the circuit and all it's solder points are correct (well, something must be wrong seeing as it's not working, just that I can't measure it directly). Does anybody have any suggestions? I'll try searching at Headwize in the meantime but I'll greatly appreciate any input.
 
Nov 10, 2002 at 6:17 AM Post #11 of 11
Well, it's 12:10 AM right now and I finally have a working amplifier. The culprit seems to be the input stereo jack. The thing had five pins to solder to and I thought I had the right ones, when I did direct hookups the jack worked. I just replaced it with a Radio Shack jack and I replaced the opamp with another TL082 (I bought two thankfully) and it works wonderfully now. Interestingly enough, with my cheap sony backphones, the amp plays very softly and then it will suddenly jump to very loud after I increase the volume some. But when I put in my HD580's, it worked just fine. Other interesting points is when the source is not outputing, I hear a low hum in the left channel, but it completely disappears once I start playing the source. It's then replaced with a hiss in both channels. I rigged up a Radio Shack pot that I thought was a dual pot, but it turns out it's just bypassing the second channel at full strength (maybe I just bridged one of the joints, but I doubt it, I've removed it now) and I find that if I have some resistance in the signal, the hiss is removed. So once I get my real pot in, I should be golden.

Thanks for all your help.
 

Users who are viewing this thread

Back
Top