Most motor drive control systems incorporate current feedback of some description.

What determines the required frequency response or bandwidth of the current transducer?

Is there any other consideration other than ensuring the sensor bandwidth is greater than the sampling/switching frequency?

  • 2
    \$\begingroup\$ There are several types so without wishing for anyone to write an essay explaining what "determines the required frequency response or bandwidth of the current transducer" I suggest you target a particular device. Also, can you explain why you would ensure the sensor bandwidth is greater than the sampling frequency? \$\endgroup\$ – Andy aka Dec 14 '15 at 14:10
  • \$\begingroup\$ For sure, your system dynamics can't be better than the bandwidth margin of any part included in the closed loop control. \$\endgroup\$ – Marko Buršič Dec 14 '15 at 17:03
  • \$\begingroup\$ Think about what would happen if the current feedback had a bandwidth equal that of the fwd (current) path, especially with a -45deg phase shift in the feedback. \$\endgroup\$ – JonRB Dec 14 '15 at 21:33
  • \$\begingroup\$ @Andyaka The fundamental component of the current ripple will be at the switching frequency and thus I assume to see the effect of this the bandwidth needs to be at least as great as this. Of course, since the load is inductive, the current will effectively be low-pass filtered and so the control system may still be stable with a lower bandwidth. \$\endgroup\$ – Dave Winterborne Dec 15 '15 at 11:25
  • \$\begingroup\$ You said "sampling/switching" and this implied you were using an ADC (sampling) tied to the switching frequency. If you are not using an ADC like this then you should remove "sampling/" from your question. \$\endgroup\$ – Andy aka Dec 15 '15 at 11:28

The required bandwidth must be great enough so faults are responded to quickly. Remember that solid state components are not as forgiving as motors when it comes to overloads.Newer devices like IGBTs have surge current ratings for about 10 microseconds and the older SCRs had surge ratings of about 10 milliseconds. When current sensing is part of the control loop which is common practice stability can be a real issue. Higher bandwidth is better from a control loop stability viewpoint.

  • 1
    \$\begingroup\$ DeSat protection is better for detecting gross overcurrent & is a lot faster and does save IGBT's. \$\endgroup\$ – JonRB Dec 14 '15 at 21:36
  • \$\begingroup\$ My question relates to the current sensors used in the control loop. Fast hardware overcurrent protection can be implemented separately. \$\endgroup\$ – Dave Winterborne Dec 15 '15 at 11:21
  • \$\begingroup\$ Yes it should be done seperately but I have seen Chiistchurch Design Engineers not do this much to my horror and then seen them wonder why it eats IGBTs \$\endgroup\$ – Autistic Dec 15 '15 at 11:29

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.