0
\$\begingroup\$

Some background: I teach circuits labs at a university and we use the 741 op-amp chip. I am aware that these chips are horrendously obsolete, but they are what we use. We specifically use the LM741 from TI datasheet here. The students are tasked with building a non-inverting amplifier using a solderless breadboard.

Out of a class of 100 students, about 95 are able to get a working circuit by the end of the lab. The remaining 5 students, though, just get a bad break. It is possible that the circuits have a subtle connection error for sure, but usually they get every TA and lab instructor to look at and often rebuild their circuit. No matter what, the op-amp just saturates. This is usually fixed by swapping the chip, but not so rarely the circuit still saturates.

Now, let us suspend disbelief for a second and assume that the circuits really are connected correctly. There appears to be a systematic problem that I'd love to be able to either detect early or solve completely. What might be causing this?

\$\endgroup\$
5
  • 2
    \$\begingroup\$ Are students using solderless breadboards? These take a beating in an undergrad lab environment, so it may not be a chip-fault. \$\endgroup\$
    – glen_geek
    Feb 17, 2022 at 15:52
  • \$\begingroup\$ Yes, solderless, and I do agree that it might be the board. Any idea why though? Like what is it about a bad board that causes it, and can it be detected before we give thhem out? \$\endgroup\$ Feb 17, 2022 at 15:54
  • \$\begingroup\$ @MichaelStachowsky The board may not be making good contact to the pins of all components. I would also ask what you're doing with the offset null pins on the amplifiers, because while they're not CMOS chips that are particularly sensitive to noise, some of the chips may need that offset nulling function to work, just based on the usual offset spread and other component variations in the circuit, so if you aren't using it, you may have too large an input offset for the circuit to work properly. I would suggest investing in some more modern op amps. \$\endgroup\$
    – Hearth
    Feb 17, 2022 at 15:57
  • \$\begingroup\$ @Hearth thanks! We are investigating the use of offset null but so far we haven't used it since we are trying to simplify the circuit to its essence for the students. However, I am starting to suspect that part of our problem is a) obsolete components and b) we simplified too much and things like offset null are essential \$\endgroup\$ Feb 17, 2022 at 16:04
  • 1
    \$\begingroup\$ @MichaelStachowsky With so very many contact-points requiring electrical testing, a test-bed is not a likely solution. Since these are spring-contacts, mechanical deformation is a common fault. Debris remaining buried can also spread pins, reducing contact pressure. Although tedious, magnified visual inspection might be your only option. Look for consistency, where all spring fingers actually meet near the centre of the plastic shells' guideway. \$\endgroup\$
    – glen_geek
    Feb 17, 2022 at 16:09

1 Answer 1

3
\$\begingroup\$

It's hard to say without even a schematic, but given the assumptions:

(1) The op amp itself is functional (2) The problem occurs sporadically (3) You're using solderless breadboards

I'd say the most likely reason is that not all connections to the op amps are working properly. Specifically, I'd take a look at the circuitry associated with the inputs, as a floating input will cause the output to drift to one rail or the other. (They don't have to float completely...for instance, if the feedback path is broken, or only one resistor of a bias voltage divider is connected, you'll often see the same saturation behavior.)

It's not as hard to troubleshoot as people would have you believe. Look at the inputs and see if the levels there justify the output (or if the output changes when you probe the input). If the output is correct according to the inputs, chances are you're missing biasing or feedback paths. This is a good opportunity to teach troubleshooting skills as well...IRL, nothing ever works the first time. :-)

\$\endgroup\$
3
  • \$\begingroup\$ Thanks! I think that our next step then should be to build a bunch of the circuits ourselves and try to replicate the problem, then do this kind of investigation. \$\endgroup\$ Feb 17, 2022 at 16:07
  • 1
    \$\begingroup\$ A couple of things you could try. Swap the 741 for another one, if it starts working, swap it back, if it still works then there was a bad connection if not maybe it is an 741 at the edge of the specification as already suggested. Also try building the circuit on a different part of the board, possibly an area that has been less used. \$\endgroup\$ Feb 17, 2022 at 16:12
  • \$\begingroup\$ Keep a record of the breadboards that have this problem, and discard those that frequently show a problem. Also, suggest that the students move their circuit to another part of the breadboard - the problem may only occur in a frequently-used part of the breadboard. \$\endgroup\$ Feb 17, 2022 at 16:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.