Overheating op-amp
Jan 22, 2009 at 1:19 AM Thread Starter Post #1 of 17
smrtby123

smrtby123

100+ Head-Fier
Joined
Jan 4, 2009
Posts
397
Likes
10
So i just finished a kind of hybrid with the buffered powered section of the A47 and the amp section of the Cmoy. My amp is all set and I know that all the soldering and everything is fine, but my amp overheated and I think it is blown but I can't see why. I have it running off of 2x4 stacks of AAA and I was testing with just 2 installed. I got 1.2V between pin 4 and ground and -1.4V between pin 8 and ground.

I was looking over the diagrams again and it seems like Ben Feists power section is wrong.
How to build the Apheared 47 Headphone amplifier for Grado Headphones

In his diagram he has the common of the TLE going to the V- of the buffer, and the out pin of the TLE going to the Vin of the buffer. Shouldn't it be the other way around?
 
Jan 22, 2009 at 1:52 AM Post #3 of 17
I think perhaps you are confused. He shows a picture of a buf 634t and in the schematic he uses a buf 634p. Likewise he has a picture of a tle2426 in the TO-92 package and the schematic shows a dip8 device.

So my question to you would be what package are your devices in?

If you are using a TO-92 tle2426 then with the flat to the right looking at the top(pins down) It should be In_Common_Out from top to bottom or V+_V-_Gnd if you prefer.
 
Jan 22, 2009 at 2:23 AM Post #4 of 17
Ok my head wasn't on straight for a sec I guess I was confusing the pins of the DIP buffer with the one in the illustration. What would be a reason for the amp to overheat though? I ohm'd out all of the socket pins and they all match what they should be.

I have the 3 pin TLE and the 5 pin BUF634T. I went back and rechecked all my joints and they all look fine but the voltage is still lopsided.
 
Jan 22, 2009 at 1:46 PM Post #5 of 17
Quote:

Originally Posted by smrtby123 /img/forum/go_quote.gif
Ok my head wasn't on straight for a sec I guess I was confusing the pins of the DIP buffer with the one in the illustration. What would be a reason for the amp to overheat though? I ohm'd out all of the socket pins and they all match what they should be.

I have the 3 pin TLE and the 5 pin BUF634T. I went back and rechecked all my joints and they all look fine but the voltage is still lopsided.


Assuming you had red lead on the pin, black on ground, your power supply voltages are backwards. Pin 4 should be negative, pin 8 positive. That will make an opamp heat up real quick.
 
Jan 22, 2009 at 2:11 PM Post #6 of 17
Although I wired the power supply section backwards with positive on the right and negative on the left, everything is where it should be I think. In Tangents tutorial he has the positive going pin 4 and negative going to pin 8 and I get a positive reading for pin 4 and negative for pin 8.

I was looking over it again and do I have to have the TLE leads between the Buffer and the power connection? I have it so that looking on the bottom of the board from left to right on the same pad it goes TLE lead, power, Buffer.
 
Jan 22, 2009 at 2:27 PM Post #7 of 17
Quote:

Originally Posted by smrtby123 /img/forum/go_quote.gif
Although I wired the power supply section backwards with positive on the right and negative on the left, everything is where it should be I think. In Tangents tutorial he has the positive going pin 4 and negative going to pin 8 and I get a positive reading for pin 4 and negative for pin 8.


Uhhh, no he doesn't. You are either not counting the pins correctly, or the voltages are backwards from what they should be. The hot opamp would seem to concur.
http://www.tangentsoft.net/audio/cmo...ayout-0150.pdf
 
Jan 22, 2009 at 3:32 PM Post #8 of 17
I guess maybe I am not reading the pins correctly, op-amps were my weak point in EE, but I followed Tangents schematic. From above, I have a wire from the positive cap to the bottom-left pin of the op-amp and then a wire from the negative cap to the upper-right op-amp pin. What would account for the lopsidedness of my voltage readings though?
 
Jan 22, 2009 at 5:01 PM Post #9 of 17
Quote:

Originally Posted by smrtby123 /img/forum/go_quote.gif
I guess maybe I am not reading the pins correctly, op-amps were my weak point in EE, but I followed Tangents schematic. From above, I have a wire from the positive cap to the bottom-left pin of the op-amp


If the opamp is in correctly (dot [pin 1] to bottom right) that would be pin 8; Quote:

Originally Posted by smrtby123 /img/forum/go_quote.gif
and then a wire from the negative cap to the upper-right op-amp pin.


That would be pin 4 subject to the conditions above.
Quote:

Originally Posted by smrtby123 /img/forum/go_quote.gif
What would account for the lopsidedness of my voltage readings though?


I lost track of what ground architecture you are using... this was Feist's A47 section? The referenced link you gave is correct. The TLE output (virtual ground) should be going to the buffer. The buffer "buffers" the TLE from having to deal with much current.
 
Jan 22, 2009 at 5:41 PM Post #10 of 17
Ok so im pretty sure I put the op-amp in backwards the first time so thats a good reason why it blew.

Yes I am using Feists A47 power section with the rail splitter and buffer, and 2 470 uF power caps. I went back over and metered everything. With a AAA in each pack I get a total voltage across the power section of 2.7V, but from pin 4 to ground I get -1.5V and from pin 8 to ground I get +1.2V. This is the same if I go from each side of the battery packs to ground, and the same with from the same pins of the TLE and buffer I get the same readings. With just one battery installed the split of a total 1.35V goes -0.75 and +0.6.

I didn't think the voltages were supposed be this disparate, but is the amp still usable in this state?

btw if it is a possible reason, my BUF634 did take a little bit of abuse during the soldering because I had to take it out and resolder.
 
Jan 22, 2009 at 7:28 PM Post #11 of 17
Quote:

Originally Posted by smrtby123 /img/forum/go_quote.gif
Ok so im pretty sure I put the op-amp in backwards the first time so thats a good reason why it blew.

Yes I am using Feists A47 power section with the rail splitter and buffer, and 2 470 uF power caps. I went back over and metered everything. With a AAA in each pack I get a total voltage across the power section of 2.7V, but from pin 4 to ground I get -1.5V and from pin 8 to ground I get +1.2V. This is the same if I go from each side of the battery packs to ground, and the same with from the same pins of the TLE and buffer I get the same readings. With just one battery installed the split of a total 1.35V goes -0.75 and +0.6.

I didn't think the voltages were supposed be this disparate, but is the amp still usable in this state?

btw if it is a possible reason, my BUF634 did take a little bit of abuse during the soldering because I had to take it out and resolder.


When you are measuring the V+ and V-, what are you using as the v-ground? The output of the TLE or the output of the BUF? You might see what the voltage between these two points is (TLE out to BUF out). The TLE should be easily within a 0.1V split. I don;t have the datasheet in front of me so I don't know its minimum input voltage, etc. You might look at that, or increase the voltage and see if the split gets closer. Just as a datapoint, resistor divider Cmoys often have much greater voltage splits. Take a look at Tangent's article on improving the virtual ground.
 
Jan 23, 2009 at 8:31 AM Post #12 of 17
I can attest that putting an OPA in backwards will cause it to blow REALLY fast. I was switching out a few different ones to see the differences, it was kinda dark and I mistook the notch side. Two seconds after turning on there was magic smoke and a non-functioning OPA.
 
Jan 23, 2009 at 2:09 PM Post #13 of 17
I couldn't really tell what had happened when I did that until the AD8397 fell off of the DIP adapter it was previously soldered to. I have to wait for another set of rechargables to come in, ( I don't have a variable power supply quite yet, I am too new at this) and my sample chips from TI and AD until I can test it again. Everything seems fine with the meter just the voltages are lopsided, but the min voltage for the TLE is higher that what I have been giving it.
 
Jan 25, 2009 at 8:00 PM Post #14 of 17
I tested the amp with a 9V battery that read about 7.5V. From pin 4 to ground it measured about -6.3V and from pin 8 to ground it measured about 1.2V. For some reason it seems to be dumping all of the voltage to one side, and I have no idea why.
 
Jan 25, 2009 at 9:14 PM Post #15 of 17
Quote:

Originally Posted by smrtby123 /img/forum/go_quote.gif
I tested the amp with a 9V battery that read about 7.5V. From pin 4 to ground it measured about -6.3V and from pin 8 to ground it measured about 1.2V. For some reason it seems to be dumping all of the voltage to one side, and I have no idea why.


I guess in order to troubleshoot this problem, you need to determine where it is getting messed up. If you measure the output of the virtual ground to battery+ and then to battery-, do you get an even split? You may want to remove the BUF for that test, just to make certain that it is not influencing it. If that looks OK, I would then check and double check your wiring of the BUF (against the datasheet) and make sure that it is implemented correctly.

This setup should give a split within 0.1V and more likely within 0.01V.
 
You must log in or register to reply here.

Users who are viewing this thread

Total: 2 (members: 0, guests: 2)
Back
Top