NE5532 preamp one channel cuts out after 5 minute

[drdanrss@gmail.]

It's a simple ne5532 preamp.
It sounds great but after 5 min channel a fizzles and stops working.
9-0-9 volts battery powered. I checked all offset volts ... nothing unusual.
I removed the opamp and tested it on another board and no damage to the ne5532?

 

[drdanrss@gmail.]

https://sound-au.com/project02.htm
I'm using the final stage of Elliott's preamp

 

[rayma]

If the DC output offset stays ok, then it's likely related to a connection on the signal input or output.
Or a bad coupling capacitor. You do have supply bypass capacitors?

 

[drdanrss@gmail.]

I'm using two 9 volt batteries.
My 10uf is Audyn Q4 from my 2way speaker Cross over.
Audiocap 3uf
Should I resolder all my joints?

 

[drdanrss@gmail.]

DC offset was 143mv chanel a then went down to 12mv plugged it in and was fine for five minutes

 

[rayma]

I'm using two 9 volt batteries.
My 10uf is Audyn Q4 from my 2way speaker Cross over.
Audiocap 3uf
Should I resolder all my joints?

Yes, and also monitor the signal after the input cap, and before the output cap.

 

[kouiky]

It may be ESD damage and thermal related intermittence. Do you have another NE5532 to try in the same circuit?

 

[drdanrss@gmail.]

I'm going to resolder tomorrow

 

[drdanrss@gmail.]

I used the ne5532 in another preamp and alls fine with the opamp
I have 153mv at A out, .23mv at B out .
I'm going to check my solder joints tomorrow

 

[drdanrss@gmail.]

Thanks guys

 

[BSST]

As a diagnostic experiment, tie the opamp + inputs of the two channels together--- they will peacefully coexist. See if a problem appears after 5 minutes in ethier channel.

Good luck!

 

[MarcelvdG]

DC offset was 143mv chanel a then went down to 12mv plugged it in and was fine for five minutes

I have 153mv at A out, .23mv at B out .

Assuming A and B are the op-amp output pins, the offset at A is too large, the offset at B is suspiciously small. 12 mV is in the right ballpark, considering the voltage drop that the input bias current of an NE5532 causes across 15 kohm and the gain.

Correction: actually -12 mV would be in the right ballpark. As the bias current flows into the NE5532 input pins, it causes a systematic negative offset when the + input sees a higher resistance than the - input.

 

[Mooly]

As it is battery powered have you actually checked both supplies are OK when it fizzles and fades?

The 5532 has a relatively high current draw and will deplete batteries quicker than you might think. Measure the rails when the fault occurs. Also add some 100uF caps across each rail.

 

[Mark Tillotson]

8mA for 5 minutes is 0.7mAh, not a problem for 9V batteries unless they are very dead - and why would a drained battery affect only one channel?

Where is/are the decoupling cap(s) in that circuit? They are not optional components.

 

[abraxalito]

The higher than expected offset could be a sign of oscillation which might be induced by lack of decoupling on the 9V rails. What @Mooly said^.

 

[huggygood]

+1

 

[JMFahey]

Bot

DC offset was 143mv chanel a then went down to 12mv plugged it in and was fine for five minutes

Both are high if ground reference resistor is only 15k
143mV is gross
I suspect path +IN>15k>Gnd is open, has terrible soldering, etc. leaving that OpAmp input floating and drifting.

Of course, this can/should be confirmed by measuring DC offset, both at +IN and output pin.

Ungrounded input will still take a few minutes to charge input cap so it may temporarily work confusing the symptoms.

 

[MarcelvdG]

Clipping will happen when the input goes to roughly -1.4 V.

1.4 V times 10 uF is 14 uC, the typical input bias current is 200 nA, the ratio is 70 seconds.

One should see a gradually increasing offset if this is it.

 

[Osvaldo de Banfield]

Do you have an oscilloscope or orher means to see if the amp starts to oscillate when souds stops? Temperature of the chip?

 

[avtech]

As others have said, make sure you decouple the op-amp properly. Even if running on batteries, there is so much RF junk floating around that can get your op-amp oscillating that you need to deal with it properly.

Follow what Rod suggests in that project, putting the 0.1uF as close to the op-amp as possible:

Each op-amp should be bypassed with a 10µF/25V electrolytic from each supply lead to ground, and a 100nF capacitor between supply leads (not shown in diagrams). The latter should be as close as practicable to the op-amp supply pins, and the 10µF caps can be almost anywhere you like, but are best placed at the DC inputs.