Op amp voltage follower problem

[jeg223]

I'm using a TS974 quad op amp wired as a voltage follower to buffer a pot for an ADC. Vcc is connected to 9V and Vdd to 0V. When the voltage at the non-inverting terminal gets to about .7V above Vdd, the output voltage goes to Vcc. I've tested 3 chips in and out of circuit and with dual supplies at ±2.5V and get the same results.
Has anyone seen an op amp behave like this before? Am I damaging the chips some how or am I just missing something?

Here is the datasheet
https://www.st.com/stonline/books/pdf/docs/6031.pdf

Thanks for your help.

 

[ericgibbs]

I'm using a TS974 quad op amp wired as a voltage follower to buffer a pot for an ADC. Vcc is connected to 9V and Vdd to 0V. When the voltage at the non-inverting terminal gets to about .7V above Vdd, the output voltage goes to Vcc. I've tested 3 chips in and out of circuit and with dual supplies at ±2.5V and get the same results.
Has anyone seen an op amp behave like this before? Am I damaging the chips some how or am I just missing something?

Here is the datasheet
https://www.st.com/stonline/books/pdf/docs/6031.pdf

Thanks for your help.

hi,
Are you saying that with +/-2.5V rails you are setting the pins greater than +2.5v.?

EDIT:
added image
The IC is latching up.

 

[MrAl]

Hello,

Cant say for sure what may have gone wrong, but if everything is hooked up right and one input terminal was at 0.7v and the other was at 9.0v, then the differential input voltage max has been exceeded so it may be that the chips dont work anymore.

 

[jeg223]

I made sure the input pins didn't exceed the power supply voltage. The differential input voltage is exceeded when the negative feedback forces the -input to 9V, but I don't see why that would happen if the chip is operating properly to begin with. I thought with latchup you need to remove the supply power to get back to normal operation, when I raise the voltage again, the chip acts as expected. With a 9V Vcc and 0V Vdd, I adjust the + input pin from 2V to 0V. The output tracks until about 0.7V. Then the output (and the -input) go to 9V, but raising the +input above 0.7V brings the output back to normal. Shouldn't the negative feedback keep the differential voltage fairly small?

 

[ericgibbs]

I made sure the input pins didn't exceed the power supply voltage. The differential input voltage is exceeded when the negative feedback forces the -input to 9V, but I don't see why that would happen if the chip is operating properly to begin with. I thought with latchup you need to remove the supply power to get back to normal operation, when I raise the voltage again, the chip acts as expected. With a 9V Vcc and 0V Vdd, I adjust the + input pin from 2V to 0V. The output tracks until about 0.7V. Then the output (and the -input) go to 9V, but raising the +input above 0.7V brings the output back to normal. Shouldn't the negative feedback keep the differential voltage fairly small?

hi jeg,
To avoid any misunderstanding it would be helpful to post a sketch showing how you are connecting the opa.

From your original description you say...
When the voltage at the non-inverting terminal gets to about .7V above Vdd, the output voltage goes to Vcc.

This suggest you are exceeding the OPA input range.??? its not clear how you have it connected.

 

[jeg223]

This is a simple voltage follower. When the pot voltage falls below about 0.7V the output jumps to 9V. When the pot voltage is above about 0.7V, the output = the pot voltage. It does violate the differential input voltage when the output jumps to 9V, but I don't see why it jumps.
I've tried this circuit with an LM358 on a breadboard and it acts as expected, Vout follows Vpot to the lower limit of the op amp, around 0.3V. I've also tested a new TS974 with a variable power supply and Vout jumps as Vin gets to 0.7V.
The obvious answer is to use a couple LM358s, but I've never seen an op amp act like this and I would like to know why it's happening.

Thanks again for your help.

 

[audioguru]

You are exceeding the allowed input common-mode voltage range.
The inputs don't work anymore and the output might go high when the input common-mode voltage range is exceeded. It was a common problem with older Fet-input opamps like the TL08x.

The only spec's in the datasheet are when the supply is plus and minus 2.5V. The allowed input common mode voltage range is from -1.35V to +1.35V. So the inputs must be (2.5V - 1.35V)= 1.15V above the negative supply which is ground in your circuit.

 

[jeg223]

Interesting...
I think that is what Eric telling me. I just took another look at page and it explains exactly what I was seeing.

Thanks again for all your help.

 

[RWD7980]

hi jeg,
To avoid any misunderstanding it would be helpful to post a sketch showing how you are connecting the opa.

From your original description you say...
When the voltage at the non-inverting terminal gets to about .7V above Vdd, the output voltage goes to Vcc.

This suggest you are exceeding the OPA input range.??? its not clear how you have it connected.

IF VDD is = 0V and the he writes the non-inverting terminal gets to about .7V above VDD (what is the problem with that statement)...he is saying that the follower shots to VCC when the non-inverting input gets to .7V above GROUND..


Sounds to me that what every the op amp is connected to is back feeding voltage to Vout,,,add some impedance to Vout,, possible a ground problem try decoupling the op amp.. Make sure the source is not choking out the opamps input bias currents I had a problem with that one time..

 

[MrAl]

Hello again,


I checked the common mode input range (first thing i checked) and that 0.7v is well within that range with a supply voltage Vcc of 2.5v, so with 9v it is certainly well within the common mode range, unless of course the data sheet is wrong, but that would be pretty darn wrong.

It sounds to me like something very basic is wrong, possibly with the connections. My suggestion would be to connect an LM358 just as a matter of proof that everything is connected right, making sure you can get that working ok.
I would also suggest that you check the package type and pinout with the data sheet and make sure you are looking at the right package when you read off the pin functions (such as Vcc, Vdd, etc). If you would like to post a picture of the actual hookup that would help too so we can actually see for ourselves.

 

[jeg223]

I think audioguru is correct. Because of the common mode input voltage, the inputs must be within 1.15V of the rails.

I found this on allaboutcircuits.com:
One quirk of some op-amp models is that of output latch-up, usually caused by the common-mode input voltage exceeding allowable limits. If the common-mode voltage falls outside of the manufacturer's specified limits, the output may suddenly "latch" in the high mode (saturate at full output voltage). In JFET-input operational amplifiers, latch-up may occur if the common-mode input voltage approaches too closely to the negative power supply rail voltage. On the TL082 op-amp, for example, this occurs when the common-mode input voltage comes within about 0.7 volts of the negative power supply rail voltage. Such a situation may easily occur in a single-supply circuit, where the negative power supply rail is ground (0 volts), and the input signal is free to swing to 0 volts.

The LM358 has a range of 0V - Vcc-1.15, which explains why it works in this circuit and the TS974 doesn't.

 

[audioguru]

Hello again,


I checked the common mode input range (first thing i checked) and that 0.7v is well within that range with a supply voltage Vcc of 2.5v

No,the 0.7V input (above the negative supply which is 0V in this circuit) is not within the common-mode input voltage range.
The allowed input common-mode voltage range is 1.15V above the negative supply and higher.

My suggestion would be to connect an LM358 just as a matter of proof that everything is connected right, making sure you can get that working ok.

Of course an LM358 will work because its allowed input common-mode voltage range includes its negative supply which is 0V here.

 

[MrAl]

No,the 0.7V input (above the negative supply which is 0V in this circuit) is not within the common-mode input voltage range.
The allowed input common-mode voltage range is 1.15V above the negative supply and higher.


Of course an LM358 will work because its allowed input common-mode voltage range includes its negative supply which is 0V here.

Hello again,


Yes that is correct, my mistake. I was assuming that there was a negative supply voltage of at least 2.5v for some reason. Im not sure now what gave me that idea

However, the point of using an LM358 is not to prove that the LM358 can work down lower (as we all know all too well) but to prove that the operator can identify and connect all the pins properly. It's not for a test of the LM358, it's a test of the human operator.

Another test with the original op amp would be to start at +2.00v and see how the output responds to that. Hopefully the op amps were not damaged due to high differential input. The input can be varied from +2v to +7v and the output should follow that within reason.

 

[jeg223]

MrAl,

I did try the LM358 in this circuit and it works. Also, using a variable power supply with the original op amp, I adjusted the input voltage like you suggested. Below, 0.7V with a 0V negative supply, the output saturates to the positive rail. When the input is within the common mode voltage range, the output follows the input.

I may have confused you with a previous post. I mentioned that I tried a dual +/- 2.5V supply with an adjustable input voltage. The output saturates when the input gets to 0.7V above the negative supply. In this case below -1.8V.

I really have to be more careful when choosing my amps.

 

[MrAl]

Hi again,


Oh yes, that's right, now i remember. I've been working on a host of other electrical problems in between so im spreading a bit thin.

Anyway, yes, now i remember you posting that you tried plus and minus 2.5v and that is why i had said that the input common mode range was ok.
It wasnt about the 9v supply, but then i should have made that more clear anyway, which i dont think i did.

If it doesnt work with plus and minus 2.5v supplies (the +0.7v input that is) then something is really wrong. It should not work right with the minus power supply pin at 0v, but with that pin at minus 2.5 it should work ok. Since it doesnt, maybe the amplifier got damaged when you connected the 9v source (with no minus supply) and the input differential went too high for that op amp. That's my only guess, especially since you got the LM358 to work with no problems. The only other thing would be to make sure that the pinout is what you think it is, but i think you did that already.

So the thing to do now i guess is get another two or three of these op amps and try again, only this time make sure to use a minus supply, and also another idea would be to use a resistor say 10k in the feedback path rather than a short as in a normal voltage follower. That resistor may help protect the inputs if the output decides to go super high again, at least possibly.

The only other thing i can offer then is if you want to send me one of the op amps i'll test it myself with my own equipment, but it sounds like you got the tests right already.

Wait a minute...are you saying that you got it to work now with voltages above 0.7v with a 0v negative rail? If so then it sounds like it is working right?
Still, if you would like me to test one for you you could send me a chip and i'll test it and send it back, no prob.

 

[jeg223]

Thank you for the offer, but I think I have everything worked out now. The chip operates properly within its specified parameters. The problem is that I over looked the input common mode voltage specs. I don't remember any discussion of this in school and have never had an opa latch up like this one did. Hopefully the LM324 I ordered will work in my circuit.

 

[audioguru]

The lousy old LM324 is noisy and has crossover distortion. It has trouble at frequencies above only 2kHz. It should be buried.

MC3317x single, dual and quad opamps have the same low power supply current, the same low supply voltage operation and their inputs work fine at 0V when there is no negative supply. They have good response to 35kHz and have no crossover distortion.

MC3407x single, dual and quad opamps have the same spec's as the MC3317x but have good response to 100kHz and have "normal" opamp supply current.
The pin numbers on the ON Semi (Motorola) opamps are the same as the LM358 and LM324.