0
\$\begingroup\$

Electronics beginner here! I'm trying to understand some confusing output from this circuit on my scope.

I'm generating a 5V 100Hz square wave from an Arduino Mega digital pin (Vsig), which looks good on my scope. I'm then using a voltage divider to create Vin, which on the scope reads ~350mV peak-to-peak (lower than I expected, but not too perplexing).

I feed this signal into the non-inverting input of an LM3900N configured as a buffer, with a 10k resistor as a simulated load. The confusing thing though is that the output of the op amp isn't a square wave anymore - the scope reads steadily at ~800mV and doesn't fluctuate from 0V to the max voltage of Vsig, as I expected in the below configuration:

schematic

simulate this circuit – Schematic created using CircuitLab

Are there any traps I may be falling into? I thought maybe the issue is that the OpAmp isn't able to get close enough to the ground rail in order to fluctuate, but being such a noob I can't really discern from the datasheet if this is true.

How can I get this setup to work?

\$\endgroup\$
2
  • \$\begingroup\$ Post full schematic, with pin numbers. You have been mistaken somethere. \$\endgroup\$
    – user263983
    Dec 16, 2021 at 19:16
  • \$\begingroup\$ Vol is the spec you're looking for. It is 0.2 V max at 15 V supply (not specified for a 5 V supply). The recommended single supply range is 4.5 V to 32 V, so that's good too. I would guess there's a problem with the implementation. \$\endgroup\$
    – Null
    Dec 16, 2021 at 19:39

3 Answers 3

4
\$\begingroup\$

The LM3900 is a Norton 'op-amp' and the inputs are not high impedance- they look more like diodes to the negative rail. Unless you have some strong reason to be using such an oddball part, I suggest getting a more conventional part such as an LM324.

enter image description here

The way you normally use this part is to have series resistors on each input and the current difference (times gain) is what drives the output. Unlike normal op-amps where the voltage difference is what drives the output. The other quirk is that the current difference is only approximate, within 10%, which is much less accurate than the effects of voltage offset on a conventional op-amp when dealing with voltages of a volt or two.

Here (from the original NS datasheet) is how you make a gain-of-one buffer amplifier:

enter image description here

But that won't do you much good since your input voltage is so low.

\$\endgroup\$
1
  • 1
    \$\begingroup\$ Curse my ignorance! Thanks very much for the elucidating answer. \$\endgroup\$
    – jonny
    Dec 17, 2021 at 13:01
0
\$\begingroup\$

One problem (that may just be an error in you diagram) is that the inverting pin needs to be connected to the output. The power pin needs to be connected to ground (which I assume is the negative power pin).

It should look something like this:

enter image description here
Source: Reasons for a voltage follower introducing voltage gain?

I'd say Double check the wiring

\$\endgroup\$
1
  • \$\begingroup\$ Ah yes, apologies, I had an error in my diagram. I've updated it now \$\endgroup\$
    – jonny
    Dec 16, 2021 at 19:34
0
\$\begingroup\$

Things to check (apart from the IN- connection)

  1. No -ve supply.
    How well does the opamp work with only a 5V supply?
    With a single supply, how close to the 0V line can its inputs and outputs swing? (Should be in the datasheet)

  2. Why do you expect Vsig on the output? How might R1,R2 affect that? What could you do to overcome that?

  3. Does your choice of opamp have the bandwidth and output slew rate you need?

One of those may be the problem.

\$\endgroup\$

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.