EEVblog Electronics Community Forum

Electronics => Projects, Designs, and Technical Stuff => Topic started by: SArepairman on April 08, 2014, 04:41:42 pm

Title: op amp failure modes?
Post by: SArepairman on April 08, 2014, 04:41:42 pm
What failure modes have you seen for an op amp? positive/negative power rail on output? high impedance? short circuit across Vin to vout? (that would be interesting to find!)


What could cause the various scenarios? Like a overload to +- terminals, or Vpp/Vee or output will have a different method of destroying the IC with different symptoms right?
Title: Re: op amp failure modes?
Post by: David_AVD on April 08, 2014, 09:10:02 pm
The most common I've seen in audio gear is the output pinned to V+ or V-.  Others have gone short from V+ to V-.
Title: Re: op amp failure modes?
Post by: SArepairman on April 09, 2014, 11:27:25 am
The most common I've seen in audio gear is the output pinned to V+ or V-.  Others have gone short from V+ to V-.

Well thats not so bad, you could even make a broken op-amp detector (peak detector with a big capacitor, or just intelligent MCU). And V+ to V- short is easily discovered with power rail monitoring, as that should drag the rail down.

How about the op-amp going high impedance? or shorting from input to output? (I imagine this might be more common with a instrumentation amplifier or something interfacing to the outside world vs an op-amp in the middle of a bunch of other op-amps.
Title: Re: op amp failure modes?
Post by: SeanB on April 09, 2014, 06:41:40 pm
Offset errors, gain dropping and nonlinearity are common errors I have found. Excessive supply current, not shorted to output but just enough to run at blistering hot. Inputs shorted either to offset pins or each other. Inputs that have massive leakage current eiher source or sink. Offset pins that short to the substrate. Totally open not so common, but all pins shorted ( generally after getting hot enough that it melts the solder and blisters the package, or melts the glass seal on a can or cracks the cerdip package) is common. Only if the power supply is capable enough does it blow the package top off, which is generally what happens to power opamps.
Title: Re: op amp failure modes?
Post by: SArepairman on April 09, 2014, 07:21:33 pm
Offset errors, gain dropping and nonlinearity are common errors I have found. Excessive supply current, not shorted to output but just enough to run at blistering hot. Inputs shorted either to offset pins or each other. Inputs that have massive leakage current eiher source or sink. Offset pins that short to the substrate. Totally open not so common, but all pins shorted ( generally after getting hot enough that it melts the solder and blisters the package, or melts the glass seal on a can or cracks the cerdip package) is common. Only if the power supply is capable enough does it blow the package top off, which is generally what happens to power opamps.

:(

thats alot of things to account for.
Title: Re: op amp failure modes?
Post by: David Hess on April 09, 2014, 11:05:09 pm
The common failures I have seen including with comparators involve either the output being shorted or open to one supply or the input differential pair or input protection circuits being damaged causing excessive input bias current and/or input offset voltage which usually ends up pinning the undamaged output.

The subtle failure which is likely to be missed involves the input differential pair or input protection being damaged but not enough to matter.  Operational amplifiers with low differential input voltage specifications usually have back to back diodes across their inputs for protection now but the diodes themselves can be damaged by overload.

JFET and CMOS amplifiers are extra susceptible to ESD damage but I have only seen this cause excessive input leakage or a shorted input in extreme cases where the protection diodes sacrificed themselves.  Electrometer devices may omit the protection diodes.

I have also seen CMOS amplifiers crowbar their power supply pins in the traditional CMOS way.  Watch their absolute maximum voltage ratings carefully.