3
\$\begingroup\$

I am designing a small audio amp and am experiencing some weird oscillations in the first OPamp (NE5532), wich is setup as a noninverting amp. The amp is supplied with +-20V and has 100nf bypass capacitors directly at the supply-pins.

schematic

simulate this circuit – Schematic created using CircuitLab

Only the inverting input pins are oscillating at 250kHz with an amplitude of 2V.

The supply is linear regulated (LM317,LM337) and has no measureable noise/ripple.

What could be the reason?

Thanks in advance

EDIT: I noticed, that the oscillation is quite intermittent. Then the smallest amount of signal on any pin will retrigger it.

\$\endgroup\$
  • \$\begingroup\$ Could you post your layout please, including info on board stack-up? I suspect this is an usual parasitic feedback effect. \$\endgroup\$ – Ale..chenski Jul 31 '17 at 20:31
  • \$\begingroup\$ Is it being driven from a sig gen or is that a simplification? \$\endgroup\$ – Colin Jul 31 '17 at 20:40
  • \$\begingroup\$ Only inverting input is oscillating? What about the output pin - is it oscillating at 250 kHz? \$\endgroup\$ – glen_geek Jul 31 '17 at 20:42
  • \$\begingroup\$ Is it happier if you add an 8k2 resistor from the non inverting input to ground? \$\endgroup\$ – Colin Jul 31 '17 at 20:45
  • 2
    \$\begingroup\$ Measure both In- and Out at the same time. Also note that a capacitive load on In- (like a scope probe) adds some phase shift, reducing the phase margin. You may need compensation in the feedback network to correct for that. \$\endgroup\$ – Brian Drummond Jul 31 '17 at 22:11
1
\$\begingroup\$

Try it with 10pF-100pF across R1, to reduce the phase shift on the inverting input. Even with 100pF you'll still have fairly flat gain across the audio bandwidth.

And I'd strongly recommend dropping the supply voltage to +/- 15v, for thermal and lifetime reasons if nothing else. +/- 22v is a survival rating, not a recommended rating.

| improve this answer | |
\$\endgroup\$
1
\$\begingroup\$

First check that this amp is stable with a gain of 5.7. Most would be.

Then look carefully at the layout of R1 and R2.

If necessary, add a small cap directly between the output and negative input pins of the opamp. Since this is audio, you don't care about frequencies above 20 kHz. I'd set the RC rolloff frequency between this capacitor and R1 at around 40 kHz so that the effect in the desired frequency range is minimal. With R1 of 47 kΩ, that would be 85 pF, so 100 pF would be fine. If you change R1, then adjust this cap accordingly.

| improve this answer | |
\$\endgroup\$

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.