Tuesday I witnessed a strange phenomena. When measuring the current going from Vcc to pin 11, I noticed that every time I contacted the probes, the current would rise to about 500mA, making the chip hot and having it motorboating. If I disconnected the probes and connected them again immediately after (like a reset), the chip would function normally, being the current limit to about 50mA, although producing sound with some distortion. If I waited for some time before connecting the DVM, the chip would overheat again. Replacing the DVM by a direct connection would fix the distortion completely and solve the problem (tested several times by turning the power on and off).

I fed the circuit with 12VDC and used the the same schematic shown in the datasheet, except that the pin 7 was directly connected to pin 17 and I was using 32Ohm headphones connected to the output.

The datasheet is here:
http://www.national.com/ds.cgi/LM/LM1036.pdf

Is the DVM interfering with the chip?

 

Silly question but did you have the leads plugged in to the correct sockets in the DVM?

 

It is not a power amplifier. Its output load should be line level, about 10k ohms, not 32 ohm headphones.

 

Silly question but did you have the leads plugged in to the correct sockets in the DVM?

Yes. I used the unprotected 10A range and used the correct socket for that.

It is not a power amplifier. Its output load should be line level, about 10k ohms, not 32 ohm headphones.

Indeed. The LM1036 has an output impedance of 20Ohm, so the output with headphones will be reduced. But I think the cause here is the DVM. It was already happening before adding the headphones. I used 470nF capacitors and 10Ohm resistors on the outputs, in order not to overload the chip (although any current going to the headphones is small). My intention was to measure the current going to the chip. The headphones were used to confirm that indeed some kind of interference was happening.

Actually, I remember that a similar problem happened when I disconnected accidentally a TDA2003 (the Vcc pin was surely disconnected). When I connected it with the power on, it started to motorboat and overheat. The solution was to turn of the power and turn it on again. So, might be the cause the same or a similar one?