Freezing or heating up components , Transistors and FETS . etc. for troubleshooting

[danny davis]

When a transistor is an amplifier it will drift the voltage up or down when you freeze or heat it up, does this mean its bad or good?

When a transistor is a switch , it will drift the voltage up or down when you freeze or heat it up , does this mean its bad or good?

When resistors are temp failure, the resistance changes up or down depending on the freezing or heating it up

At work i had to do freezing spray and heating up these OP AMPS that were in the shape of a metal can transistor looking types like a IC chip package type, I had to freeze spray and heat them up ONE by ONE

But I can't tell if the transistor or these Op amps are good because the voltage will go up or down

 

[chuckey]

Well designed circuits will drift with temperature but only by, say 10mVs, any more then this means that there is a fault with the circuit which could be that the design is faulty. Also consumer electronics are not specified to a very low temperature - who watches a TV set at -10° C ?. The freezer spray is usually used to temperature contract the leads which show up bad joints.
Frank

 

[danny davis]

I used a freeze spray that was -40C , -40F , and spray each transistor and Op amp , it they drifted up or down way more than 10mV

How far can they drift up or down before you know the transistor or Op-amp is bad?

 

[betwixt]

Small and gradual changes are quite normal. The electrical properties of semiconductor materials is temperature dependant. 10mV may be nothing to worry about but it really depends on the individual circuit. It isn't a good idea to shoot freezer spray before you have localized a fault, the thermal shock and possible 'out of operational parameters' may do more harm than good. If a part is faulty with a heat related problem it will generally be catastrophic, the freezer will make the fault appear or disappear suddenly, a slow drift is usually nothing to worry about.

Brian.

 

[FvM]

It should be noted that the temperature gradient achieved by freeze spray is different from normal operation by several orders of magnitude. It might even damage some sensitive components (not simple transistors and ICs, but e.g. optical sensors). So it should be used carefully.

For the same reason I would refuse to specify a good or bad drift amount for a specific circuit. Freeze spray (and hot air) are good tools to indentify sensitive regions of a circuit and also to locate faults, once you have an idea about normal behaviour of the particular circuit.

 

[danny davis]

a good or bad drift amount for a specific circuit. Freeze spray (and hot air)

What is a GOOD drift and what is a Bad drift?

Is -40c/-40F freeze spray to strong? will it damage the transistors and op amps metal can type package?

 

[chuckey]

As said a sudden jump by 1V would be considered indicating a faulty item. If you look at the failure rates of components modern silicon devices are the most reliable, in general they only fail due to some transient overload (static voltage, transient voltage spike or over dissipation to to a current overload etc.). Suspect connectors, solder joints, electrolytic capacitors, switches, wound capacitors, resistors then silicon devices last.
Frank