[Possum]

Can an op amp or buffer IC be blown if opposite voltage is applied to the Vcc's? i.e. positive rail connected to V-, and negative rail connected to V+?

 

[livingplasma]

I have damaged the Bur-Brown chips that way when connected for even just a second, though the LM386 for my first headphone amp survived even though it was hooked up long enough to get them hot. I'm adding diodes for my next project, hooking up the power backwards is one of those stupid things I seem to keep doing and I've had more bad luck than good =(

 

[Possum]

Doh!

After you've blown those Burr Brown op amps this way, and you reconnect the power supply the correct way, do the op amps still get very very hot or are they completely cool, like a dead open-circuit component?

 

[Possum]

Is a buffer (EL2001) considered blown if it's spitting out ~7-8V DC from its output?

(I have a bad feeling about this, since the EL2001's seem hard to get now)

Update: Ok, so the buffers don't seem to be blown, since I stuck in a new and untouched EL2001 and got the same result out of that.

Any ideas about this situation: Circuit has correct voltage (18V between rails, with + and - 9V from each rail to ground. Each EL2001 buffer for each channel is outputting ~7-8V DC with nothing at the amp's input. The op amp that outputs to the input of the buffer gets very hot if the buffers are in the circuit, but remains cool if the buffer IC are pulled out of their sockets. Op amp has been replaced with the same result. The only other things in the circuit are resistors, non-polarized (is that the term?) decoupling capacitors, non-polarized input-coupling capacitors, and electrolytic capacitors. The circuit has not been changed and was working before I attempted something yesterday, so there are no shorts.

What could be causing the output DC voltage? What's making the op amp hot when the buffers are in? If a source is connected to the input, all I hear are random pops and random static at the output.

 

[ppl]

If the Input of the Buffer is untermanated as would be the case with no opamp You will get half the supply voltage upon the output because the Input Bias Current of the Buffer force the input twards one of the Supply rails.

(1) What if any DC voltage appears at the OPamp output with a resistor around just the Opamp - in To Opamp out (Buffer Removed)?

(2) With tthe opamp removed and the Buffer installed, Ground the Buffers input. Measure the DC Voltage upon the Buffer's output To Ground Did this give a DC voltage less than 10Mv ?

(3) Have you tried another opamp?

 

[Possum]

I just measured the DC at the outputs of the op amp, with no buffers in the circuit, and still got ~7V DC.

Update: Hmm, perhaps a blown op amp then. Sticking in an OPA2132 in place of the AD823, still without the buffers, produced the proper results.

Looks like I've blown both of my AD823's

I'll have to use a substitute until my next order.

Update 2: It appears this moron has blown 2 AD823's and 4 EL2001's

Not so bright eh? Until I can find a good source for the EL2001's, I'm basically running a cmoy instead of an Apheared #42. (but I still have my robust Szekeres

)

 

[ppl]

Sorry to Hear that Possum;, Yes the El-2001's are hard to get as are almost all Non Common Semi's. You might use a Diode to Protect your Amp from Rev Supply Voltage. A series Diode works and you only lose 0.7 Volts I use a Parralell Diode That is rev Biased with the Proper Polarity and Foward Bias with rev Bias. This has no Voltage Drop but will act as a short if you Connect the Supply Backwards. BTW I do the same Thing alot also.

 

[Possum]

All this because I wanted thinner and more flexible wires connecting my power supply module (with regulator and virtual ground driver) to the amp circuit. I had a hefty cable with Molex-type connectors that could handle something like 400V and a couple of amps from Radio Shack, and decided to use 2 9V battery clips. So being an ignorant fool, I forgot that while the red and black wires were positive and negative on one side, they were the opposite on the other. Unfortunately, I didn't realize this until after I tried all of my spare chips

 

[aos]

You've blown 2 AD823 and 4 EL2001?? That qualifies as a catastrophe, man!

 

[Joobu]

I reversed the voltage on a pair of buf634s. They got so hot they burned my finger. Strangely enough, the buf634s continued to work. I guess that's what they mean by over-temperature shutdown.

Don't try that at home.

 

[ppl]

after Trying the Second EL-2001 Did This second unit ever get reversed voltage or just get hot from the Bad opamp? The El-2001 Dose not have Thermal Protection in the Conventional way Thay used to 1990 vintage but that was changed later about 1995 or so. The Newer ones will Disconect one of the Supply Rails causing tha Large voltage you seen. I would try them again using an opamp you can afford to blow if indeed the Part is defective. If all it did was get hot it should be OK once Cooled Down unless you had a load connected at the time of the Mishapp and you did not have a resistor between the output of the opamp and the EL2001's input. These Buffers have Protection Diodes to protect against the opamp swinging to the supply rails but if the Current through them is more than 50 MA you might have Blowen the Part. Worth a try.

The BUF-634 has reverse Bias protect in the Form of Catch Diodes placed in rev. Biased form from each Supply Rail to the output. These are most often used to protect from the Back EMF from inductive loads. The Thermal protection of the BUF-634 is done in a way i dont like because it limits Output current at evean moderate Heat. But it is as the Data sheet says Fool Proof to use. The Diodes will not only Protect from Back EMF but will also limit any rev polarity to about 1.2 volts.

Rather than demand the Opamp to drive your phones i would use those BUF-634's as outputs untill you can get some more EL's and AD's BTW an OPA2132 + BUF-634 sounds Good to me. Ihave tried this combo several times but using the single opa-132.Rather than the dual. But i must add that nether the BUF-634 nor the OPA-132 is the last word in sonic accuracy. An Opamp that sounds verry musical with the BUF-634 operated in a Wider than Min Bandwidth Mode at least a 100 ohm resistor between Pin (1) the BW pin and V (-) and using Walt's Multiloop Method is the OPA-637. Way Better than the OPA-627 but layout is a concern as the 637 unlike the 627 has almost no phase margin so Design for stability and use High Speed layout See LTC Application Note #47.

 

[Possum]

Yeah, I tried and retried all of the EL2001's again just to be sure. I used no op amp in the circuit at first, with nothing at the amp inputs, and the bad EL2001's outputted ~7V. The good EL2001 that I had would output under 2V. One of the bad ones actually outputted 2V also, but was producing crackling noises like it was broken, but not as badly as the others. I then stuck in an OPA2132 and redid the test with the same results.

I'll probably just contact Elantec and see what they recommend as the EL2001's most recent replacement.

 

[ppl]

The EL-2001 is a Current Part as far as i Know but the CN grade may have been replaced with the ACN grade, but you might check to be sure.

 

[Possum]

I'll give Elantec a call next week. This is what I got in an e-mail reply from Avnet:

"Please keep in mind that this part number is showing as obsolete in our system. You may want to check with the manufacture to they will be providing a replacement. Currently we have about 168 peices in stock."

Avnet didn't have any of the ACN's in stock, Nu Horizons has 87, but has a minimum package of 50 (what's the difference between having a minimum package of 50 and a minimum order of 50? Nu Horizons has both fields for some products.).

 

[aos]

I've also heard from a supplier that EL2001CN is obsolete but the newest revision ACN can still be had.

AVNET seems like wholesale supplier - ACN has minimum order of 100 pieces though the price is nice. CN has 50 pieces minimum order. I'd say you guys are really desparate

.