How do tell if a op amp is bad or woking right?

[Jam]

How do tell if a IC op amp is bad or woking right? I dont understand how they work? Thanks for any info you have

[Prairie Dawg]

You need to know what the input and output voltages are supposed to be and where they come in and go out, have a good multimeter capable of making the measurements, a set of test leads with needles soldered onto them to poke through the lacquer covering the traces, and measure them and be able to know to a gnat's ass certainty that you did it right.


There are any number of good articles on the web that describe the basic function of an op amp. This is a good one:

http://talkingelectronics.com/projec.../OP-AMP-1.html

[soundguruman]

How do tell if a IC op amp is bad or woking right? I dont understand how they work? Thanks for any info you have

Generally in an audio amp, you test the output(s) of the op amp(s).
If the output has DC on it, it's probably bad.
In other op amp circuits, it's not unusual for DC to be on the outputs. But not in audio applications.

Another great test is just let the amp run for a while. If the op amp is bad, a lot of times it will become HOT. Better test quick, it might BURN your finger.

[Enzo]

OP amps in the signal chain normally do not have DC on the output, and if they do, are likely bad, indeed. In guitar amps, many now use op amps in switching circuits, Fender amps come to mind, and the output p[ins on those will normally have 15v on them. Also, op amps that drive LEDs will have voltage often. Point being, you need to know what you are looking at.

You usually can;t measure signal; at the input pins. If you have signal at the output pin, then there must have been some at the input, eh? A simple function test is to have the amp on, and touch the input pins with a meter probe. That should introduce some hum, which would be heard. Not 100% effective, but most of the time it tells me. WOn;t tell me if it is distorted, but it helps as a go or no-go test when I want to know why an amp is silent.

[Prairie Dawg]

And, if you're real lucky, they'll be in sockets so you can just change 'em out.

[km6xz]

Opamps product an output that is a function of the input, either inverting the phase or not, and uses a feedback path, often a resistor or pot, connected between the output pin and the inverting input pin. This feeback component's value will determine the gain or frequency response of the amplifier stage. So there is no need to remove an opamp to test it, the output should reflect the input. The output most often sits at 0volts when there is not signal but many portable devices, or low cost items have a single positive power source feeding the opamp. In those cases, a no signal DC output will be close to 50% of the positive supply voltage. Many newer opamps can function well using a single supply. The inverting input of the opamp, in amplifier configuration, will not measure any signal if the opamp is working, but a signal will be measured on the other side of the resistor that is connected to the junction of the feedback resistor and the inverting input pin. The non-inverting input may simply be connected to ground, or be serving as the signal input depending in if it is wired as an inverting or non-inverting amplifier stage.
The most common indications of a bad opamp are pulling high current due to a massive short in the output stage of the opamp chip, or there being a DC voltage at full positive or negative volt level in a two power supply configuration(usually the power sources are 12-15 positive and another power input connected to a 12-15v negative power source.) Portable battery powered devices are often used in single supply configuration so the only voltage supply is 3-5 volts and the opamp is expected to have a steady no signal output of 1/2 the supply voltage.
If there is a massive short in the IC, so the two supply voltage are shorted together, usually there is enough current pulled to cause the power supply to shut down of its regulators generate a lot of heat. Opamp power supplies are not needed to be high current so a short in one of the ICs will usually cause the power supply output to drop close to 0v. The faulty chip causing the low supply voltage quite often has a very low resistance so there might not be much heat from it, but the power supply voltage regular will be hot.

[R.G.]

Here's what I do.
- voltmeter with reference probe on power ground, measure voltage at opamp power pins. This will usually be bipolar (i.e. +, - X volts dc) or single ended (0V and either +x or -x but not both). Look at the schematic. Do the power supply voltages on the actual pins match the schemo reasonably well? If not, the power supply has to be fixed before the opamp has any chance to work.
- while in the schematic, look to see what the opamp does. Is it an amplifier, oscillator, integrator or misused as a comparator? If it's an oscillator or comparator, it's harder to tell if it's working, because other stuff may be telling it NOT to work.
- but amplifier opamps are easy. Measure voltage on output and both inputs. All must be within a few millivolts of the same voltage, and that voltage must be well in the middle between the power supply pins. In an opamp working as an amplifier, feedback forces all pins to the same DC value. How much "in the middle" depends on the opamp. Rail to rail opamps can be working right within 50mV of the power supplies. Non rail-to-rail opamps generally need 1 to 3 volts inside the power rails to be working right.
- for signal opamps, all three pins must follow the + input. If it's at some middle bias point and the output is not following it, either the opamp is bad or the feedback elements are telling it not to be linear. If it's off at one or the other power supply, then the outputs may be being driven into unusual voltages by the opamp being unable to follow that bias voltage.

I like the comment about heat. Hot is not good for a signal opamp.

[woodyc]

I usually test opamps on a breadboard.

But if its a typical dual type, here's something you can use to make a cheap tester:

Amazon.com: Super Ear Amplifier Kit: Home Improvement

[loudthud]

In some circuits there are multiple opamps in a giant circuit with several feedback paths. So which one is bad? Try to find the opamp where the output is not obeying the inputs. Put the minus probe of your meter on the minus input and the plus input of your meter on the plus input. If there is more than +/- a couple of millivolts, what is the polarity of the voltage? If it's positive, the output should be on the positive rail, if it's negative, the output should be on the negative rail. This method will work on the switching circuits Enzo mentioned.

In most linear (amplifying) circuits, an opamp trys to bring both inputs to very nearly the same voltage. There is a little uncertainty, called offset voltage, but the inputs should be within a couple of millivolts of eachother. When the output slams the rail, the feedback cannot accomplish this (called closing the loop). This happends when an opamp clips a signal on the output.

The inputs of an opamp only have a cetain range that they will operate correctly over called the common mode range. You'll have to look this up on the data sheet. If one or both inputs are outside that range, the output can go to the opposite rail that it otherwise would. Usually no damage will result unless the inputs are taken beyond the supply rails.

[Mick Bailey]

Here's an excellent reference for op-amp theory.

www.ti.com/lit/an/sboa092a/sboa092a.pdf

[Mick Bailey]

Just out of interest, and by no means meant to confuse the OP, but does anyone remember the Norton current-differential op amps - the LM2900 and similar? I built lots of interesting audio stuff around those, kicked off by building the MS-20 synth from MFOS (which I still have). Shame they didn't last, a really nice chip. Interesting theory document;

www.ti.com/lit/pdf/SNOA653