I am working on current sensing for three phase inverter outputs, wherein I plan to use a shunt resistor for getting a low voltage sinusoidal current waveform. Is there any way at all that could convert 0.001V peak sine wave (50Hz) to 0-5V sine wave?
6
-
\$\begingroup\$ Instrumentation amplifiers (random example: INA821) can do this. Or do you want to do this with plain opamps? \$\endgroup\$– CL.Commented Oct 6, 2023 at 9:23
-
1\$\begingroup\$ The current waveform depends on the load and the inverter output voltage sine quality so, what prompts you to say it will be sinusoidal? There could be many harmonics present so, you might need to take this into account. 1 mV to 2.5 volt is a gain of 2500 and, this is achievable for moderate bandwidths with reasonable accuracy so, what accuracy are you requiring? \$\endgroup\$– Andy akaCommented Oct 6, 2023 at 9:31
-
\$\begingroup\$ What is the value of your shunt? Depending on the circuit you choose you might require isolation. How much current is going through the outputs? How accurate do you need to be? Inductive pickup may or may not be a better fit. You should show a diagram. \$\endgroup\$– ReinderienCommented Oct 6, 2023 at 14:03
-
\$\begingroup\$ @Reinderien I am using 100micro ohm as the value of shunt. The inverter output is designed to carry 70A of rms current. However, I have selected the shunt considering 300A (including harmonics, spikes or to assist tuning of the drives). The more accurate it is the better it would be as I am planning to use the currents in the control algorithm after sensing them. How much accurate can I really get with opamps? \$\endgroup\$– Shishir TrivediCommented Oct 9, 2023 at 7:15
-
\$\begingroup\$ @Andyaka I have considered the worst case scenario wherein 300A would be the output of the inverter. As I am driving inductive loads I am not worried much about harmonics. For accuracy, do you recommend this circuit to be used for sensing real time currents for current control algorithms? \$\endgroup\$– Shishir TrivediCommented Oct 9, 2023 at 7:18
|
Show 1 more comment
1 Answer
\$\begingroup\$
\$\endgroup\$
2
Basically you need AC gain of 2500, and add 2.5V DC offset. You can do this with opamps. You might want to spread that gain over several stages, depending on device and bandwidth, for stability. For instance, you could have two inverting stages - first AC coupled with gain of 250, second one working as a summing amp with two inputs - AC from first stage, gain of 10, and a DC voltage of -2.5V with unity gain. Something like this:
simulate this circuit – Schematic created using CircuitLab
(You probably need a second DC block cap between the two stages.)
-
\$\begingroup\$ Hi, thanks for your inputs regarding the topic. The circuit which you proposed is similar to what I found in Ti's document. However, do you think it could give a good accuracy say 1% for the sensed ac currents? \$\endgroup\$ Commented Oct 9, 2023 at 7:20
-
\$\begingroup\$ what kind of accuracy? gain? DC level? frequency response? 1% should be possible without too much problem, it's about component tolerances mostly. Either 0.1% resistors or 1% with a trim should achieve that. (if you find that my post this answers your question please mark it as an answer and upvote.) \$\endgroup\$– danmcbCommented Oct 9, 2023 at 7:24