I am building an amplifier with multiple op-amps, and it seems that some are overheating. Here is the schematic:

U62 heats up the most (too hot to touch), U63 also heats up but a little less, U64 heats up the least but still significantly. U65/66 don't heat up at all. DAC0_OUT1/2 are current outputs from a DAC with a maximum output of 20mA. The amplifiers that are heating up are all LM7372, powered with a split supply of +/-15V. (From a 7815 and 7915). I've confirmed the voltages on the supply pins and they are correct. I'm not even sure where to begin debugging this. Strangely enough, the circuit actually functions - I see the expected voltages at the outputs of the amps that heat up. However, the amp should only be consuming a few mA, so no reason it should be heating up - any ideas?

If I've left out any information that could be helpful let me know! Thanks!

  • \$\begingroup\$ You have U62B wired /w positive feed back. Is that intentional? \$\endgroup\$ – pgvoorhees Dec 9 '17 at 15:33

Since the LM7372 is a high speed op-amp your first suspicion should be that they are oscillating at high frequency.

Your U62A/B can be expected to oscillate since it is decompensated for high speed- the minimum gain to assure stability is +2 and you have +1. Bingo, it oscillates like a banshee.

The LM7372 is stable for gains as low as +2 or −1.

The following amplifiers are inverting and have higher gain (minimum gain is -1 for inverting), so they'll just attenuate the many-MHz oscillation from the first ampifier and not oscillate themselves.

  • \$\begingroup\$ That makes sense - is there any way to fix the oscillation without changing the gain? \$\endgroup\$ – Billy Kalfus Dec 9 '17 at 5:31
  • 1
    \$\begingroup\$ Use a different op-amp or increase the gain to at least +2 with two resistors added on. \$\endgroup\$ – Spehro Pefhany Dec 9 '17 at 5:32

The output resistance of 499 Ohms is too low in value for the opamp to drive if the supply voltage is +/-15V. I suggest increasing the resistances by ten times.

  • 3
    \$\begingroup\$ The 499 ohm resistors do not connect to ground. Effective load is about 1.4k (well, 1.387 k nominally) so the load is well within spec. \$\endgroup\$ – WhatRoughBeast Dec 9 '17 at 15:51

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.