1
\$\begingroup\$

I have a very simple circuit with two op-amps which is latching up, as described below. I suspect that one of the IC's is simply defective.

two op-amp circuit

Don't worry about the implied circuit continuation to the right; not all the parts are installed so that is not connected. Likewise, no input source is connected.

The power supply is +/- 15V. The op-amps IC's are NE5532. The expected behavior is that a maximum of 1000 nA of input bias current flows from through both + inputs across R1, resulting in a voltage of up to 50 mV across R1. That is tolerable in this application, which is why I don't have any additional resistors to try to to compensate for input bias current.

What actually happens is that somehow enough current flows out of the + inputs to drop -15V across R1, sticking both + inputs to the negative power rail. (This is beyond the common mode voltage range given on the datasheet). The output cannot swing that far; it only goes to -14, so a differential voltage of 1V develops. This is enough to forward-bias the back-to-back protective diodes which connect the inputs, a feature of the NE5532. Both IC's run fairly hot.

Have I done something that these op-amps don't like, or do I have a defective part?

\$\endgroup\$
  • 1
    \$\begingroup\$ Check for a short between pins 3 and 4 of both devices. That would explain everything. \$\endgroup\$ – Olin Lathrop Nov 7 '12 at 19:24
  • 1
    \$\begingroup\$ Indeed, 3 and 4 are shorted to hell. 0.1 ohms on my MM's finest scale. In-circuit, I don't know which device, of course, until I pull one of them. If you make that an answer I will accept it. \$\endgroup\$ – Kaz Nov 8 '12 at 3:19
  • \$\begingroup\$ The fix will have to be to drill out that through hole so no metal remains. This takes about a 35 mil drill bit. The capacitor can be omitted, or surface mounted. \$\endgroup\$ – Kaz Nov 8 '12 at 6:01
  • \$\begingroup\$ A 33 bit did it. \$\endgroup\$ – Kaz Nov 8 '12 at 6:16
  • 1
    \$\begingroup\$ Can someone send a comment that somehow smacks me on the head? Thanks. \$\endgroup\$ – Kaz Nov 8 '12 at 6:17
3
\$\begingroup\$

Check for a short between the positive input of either amp and the negative supply. They are on adjacent pins. If the positive input is shorted to -15V, it explains all your symptoms.

\$\endgroup\$
2
\$\begingroup\$

The NE5532 part is a dual op-amp part. What have you done with the connections at pins 5, 6 and 7?

First off make sure that one of these pins is not shorted to any of the in use pins. Secondly it is highly recommended that the second amplifier, if not in use, be biased off in a safe manner so that does not induce unwanted oscillation, leakage or power dissapation in the package.

Proper connections for unused op-amps in a package and discussions regarding the bad things that can happen if you neglect to do this can be read in this PDF from the Analog Devices Web site:

http://www.analog.com/static/imported-files/rarely_asked_questions/RAQ_unused_op-amp.pdf

Note that open unused pins on the 2nd amplifier can sometimes actually lead to latchup on the package and could couple between the two amplifiers. Latchup that extends for a long time can destroy the IC.

\$\endgroup\$
  • 1
    \$\begingroup\$ The circuit diagram is an excerpt from a larger one. Pins 5, 6 and 7 are involved in active low-pass filter stages which follow. (Not all the components are there yet to connect the signal flow.) I always connect unused op-amps properly, like connecting output to -, and grounding +. \$\endgroup\$ – Kaz Nov 8 '12 at 3:24
1
\$\begingroup\$

Is maybe R1 connected to VNEG instead of GND? Or is GND = VNEG?

Either this, or a short as suggested by Olin.

\$\endgroup\$
  • \$\begingroup\$ Thanks for the feedback. GND is definitely not VNEG, or else I might as well quit laying out PCB's and take up cross stitching. :) I will check the short hypothesis at the earliest opportunity. I may have burned one of the IC's, since I'm soldering these rather than socketing. Though I make each individual joint quickly, I may have overheated the chip by making too many joints rapidly in succession on one chip without giving it time to cool. \$\endgroup\$ – Kaz Nov 8 '12 at 0:55
  • \$\begingroup\$ And thus, I think I should take up cross-stitching after all. \$\endgroup\$ – Kaz Nov 8 '12 at 7:22

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.