0
\$\begingroup\$

Circuit Schematic

Hello, I'm implementing a current sensing circuit as shown in the schematic. The simulation works but the physical circuit does not. On the breadboard if I measure between the shunt and the load to the ground on the power supply I get 67.3mV, however if I measure across the shunt I get 27mV (expected value since current @ 100ohm load is 110mA). As a result the Op-Amp amplifies the 67.3mV and I get the wrong amplified voltage. I've connected all the ground rails together and to the power supply. Can anyone say what's the possible problem ?

EDIT: I changed the shunt resistor to 1ohm and it works as expected, so why doesn't it work with the 0.25 ohm shunt ?

\$\endgroup\$
5
  • \$\begingroup\$ In all likelihood your problem is that you're trying to use a breadboard for "precision" measurements. You'll probably find that the contact resistance of the breadboard connections is throwing everything off. \$\endgroup\$
    – brhans
    Commented Dec 13, 2016 at 17:47
  • \$\begingroup\$ @brhans If I solder the circuit to a copper board, would that help the situation ? \$\endgroup\$
    – SRR
    Commented Dec 13, 2016 at 17:55
  • \$\begingroup\$ What are the tolerances of your components? What is the accuracy/precision of the OA? \$\endgroup\$
    – vofa
    Commented Dec 13, 2016 at 18:04
  • \$\begingroup\$ @vofa 5% resistor tolerance and I don't know what you mean by accuracy/precision of the OA \$\endgroup\$
    – SRR
    Commented Dec 13, 2016 at 18:06
  • \$\begingroup\$ @S.Ramjit Measure the resistors to see their actual values, then calculate what you think you should see on the output. All real devices are non-ideal. Understanding the limitations of your specific device is critical to interpreting measurement results. www.ti.com/lit/an/sloa011/sloa011.pdf \$\endgroup\$
    – vofa
    Commented Dec 13, 2016 at 18:27

2 Answers 2

1
\$\begingroup\$

In all probability you are using an op-amp that just cannot work with the inputs close to its most negative power rail (0 volts in your example). Most "regular" op-amps need a couple of volts headroom within the power rails for inputs. This also applies to outputs.

If you are using an LM324 you might get this to work by loading the output with a 1 kohm resistor. It says something to these ends in the data sheet.

\$\endgroup\$
5
  • \$\begingroup\$ Is there any other op-amp I can use ? \$\endgroup\$
    – SRR
    Commented Dec 13, 2016 at 20:05
  • 1
    \$\begingroup\$ How about telling me what you are currently using.? \$\endgroup\$
    – Andy aka
    Commented Dec 13, 2016 at 20:41
  • \$\begingroup\$ I'm currently using the LM324N Op-Amp in a Non-Inverting Configuration \$\endgroup\$
    – SRR
    Commented Dec 14, 2016 at 2:57
  • \$\begingroup\$ try loading the output to ground with a 1 kohm resistor. Come back and tell me what happens please. \$\endgroup\$
    – Andy aka
    Commented Dec 14, 2016 at 9:26
  • \$\begingroup\$ Voltage across shunt: 27mV Voltage input to Op-Amp: 202mV Voltage Output from Op-Amp: 1.96V I'm still having the same issue where the voltage from the shunt is not the voltage input to the Op-Amp. When making the measurement I simply move the negative lead from the side of the shunt connected to ground to the ground on the power supply and the voltage changes from 26mV to 202mV. Even with the 1kOhm on the output, I'm still not getting a proper reading \$\endgroup\$
    – SRR
    Commented Dec 14, 2016 at 16:03
0
\$\begingroup\$

Place a resistor with an approximate value of 80% of R2 at the non-inverting input (commercial value 8K2) to see if the sensing is correct. In case of adequate reading I will try to explain the reasons for this.

\$\endgroup\$
1
  • \$\begingroup\$ You are responding to a nearly 2 year old post, do not expect the OP to return with an answer. \$\endgroup\$ Commented Sep 27, 2018 at 20:12

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.