3
\$\begingroup\$

I have this circuit. The goal is that it going to control the current loop by using the NPN-transistor. My problem here is that 0V results that the output from the Op-amp is near 24V. Why?

enter image description here

If I'm increasing the voltage a little bit, then I got this.

enter image description here

And even better, If I implement a schottky diode, then I got this.

enter image description here

So why does the output goes to near 24V when the non inverting input is 0V?

\$\endgroup\$
7
  • \$\begingroup\$ There should be a pull-down resistor at the base of the NPN transistor. Most op amps can not output a voltage equal to the bias voltages (24V and GND in this case), so the output can't go to zero. Probably better to reduce 10k to lower value. \$\endgroup\$
    – Ernesto
    Mar 16 at 17:59
  • \$\begingroup\$ @Ernesto Thank you! I will try that! \$\endgroup\$
    – MrYui
    Mar 16 at 18:05
  • \$\begingroup\$ @Ernesto Did not work. Are you sure about there should be a pull-down resistor (e.g 10K) before the base of the NPN transistor? \$\endgroup\$
    – MrYui
    Mar 16 at 18:09
  • 1
    \$\begingroup\$ There doesn't need to be a pull down @DanielMårtensson - always check someone's reputation before starting a wild goose chase LOL. Falstad isn't that good. \$\endgroup\$
    – Andy aka
    Mar 16 at 18:10
  • \$\begingroup\$ @DanielMårtensson I should have qualified my statement as I think this might fix it. I think now that 0V is not a voltage source but a open connection, maybe try adding 0V voltage source? \$\endgroup\$
    – Ernesto
    Mar 16 at 18:36
6
\$\begingroup\$

It looks OK in my simulator: -

enter image description here

0 volts in produces near enough 0 volts out (well, 7.847 μV) with a drive level from the LM324 of around 200 mV.

If I increase V1 to 100 mV, it all makes sense too: -

enter image description here

I suspect that Falstad (is that your sim?) is having problems (after all, it's pretty much a toy when it comes to sim tools) but, importantly, the LM324 will never have problems with inputs or outputs close to 0 volts because it was designed to encompass the negative supply rail, especially so the inputs.

\$\endgroup\$
9
  • \$\begingroup\$ Great! So Flastad have some issues here. What software are you using? \$\endgroup\$
    – MrYui
    Mar 16 at 18:11
  • 1
    \$\begingroup\$ Microcap 12 - it is currently free. \$\endgroup\$
    – Andy aka
    Mar 16 at 18:11
  • \$\begingroup\$ Thank you very much! \$\endgroup\$
    – MrYui
    Mar 16 at 18:12
  • 1
    \$\begingroup\$ @jusaca I've been using microcap since circa 2005 and LTSpice has a few anomalies I don't like (differential time measurement for one and not as well-rounded as microcap). \$\endgroup\$
    – Andy aka
    Mar 17 at 10:31
  • 1
    \$\begingroup\$ @Andyaka never heard about Micro Cap. Only LTSpice. Thanks! \$\endgroup\$
    – MrYui
    Mar 17 at 23:39
10
\$\begingroup\$

The LM324 model (or the simulation) is seriously broken.

LM324 input common mode range includes the negative (and a bit more) so this should never happen for non-inverting input within 100mV of 0V. (If you drive the non-inverting input well below 0V, the output can be driven high, but that's not happening here).

enter image description here

Raise the non-inverting input to 500mV and it outputs a negative voltage. That is never going to happen either.

enter image description here

Suggest not wasting time with this software. LTspice is another serious option.

\$\endgroup\$
1
  • 1
    \$\begingroup\$ +1 for LTSpice since it runs very well with wine, like he mentioned \$\endgroup\$
    – devnull
    Mar 16 at 18:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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