0
\$\begingroup\$

I was reading a discussion on this below thread regarding shunt Vs. hall effect and had a few questions in mind. I do not have enough reputations to add comment on that thread hence starting this new thread.

Also being noob, my question may seem dumb to the senior fellows in the community. Here to learn.

The link : What are the advantages of a shunt resistor vs. a hall effect sensor?

1) Am I correct to understand the current probes such as Tektronix TCP312A are also "Hall effect sensor" type devices? IF true, next question is applicable.

2) According to the Pros listed by @Vladimir Cravero, it seems Hall effect sensor measures more precisely. However in this old post Agilent claimed they are inaccurate.

Link: http://www.electronicproducts.com/Test_and_Measurement/Benchtop_Rack_Mountable/Characterizing_dc_inrush_currents.aspx

Those two statements seems contradictory.

3) It was mentioned in the thread that HS won't work beyond 100Khz. But Tektronix TCP probe series has range of DC -100 Mhz. (Again only valid if answer to Q1 is yes)

Lastly, I cannot add appropriate tags such as "current-probe", TCP300, TCP312, to this thread as I do not have enough reputation. Can admin pls apply it? Appreciate the help.

\$\endgroup\$
  • 1
    \$\begingroup\$ There is more than one way to use a Hall sensor- if you use a feedback loop to cancel out the field caused by the current you are measuring with a feedback coil, the linearity can be improved greatly, however the bandwidth will be less. \$\endgroup\$ – Spehro Pefhany Sep 2 '16 at 16:58
  • \$\begingroup\$ Please change the title to your question. The question seems to have nothing to do with the current title. \$\endgroup\$ – user2943160 Sep 2 '16 at 17:41
  • \$\begingroup\$ Well, I had some question related to the title but sure, I will change it for now. \$\endgroup\$ – Jigish Parikh Sep 2 '16 at 19:19
1
\$\begingroup\$

1) Current probes and non-contact current sensors are not necessarily just Hall-effect devices. Some are hybrid Hall-effect/current transformer devices or even flux-gate/CT hybrids. So the Hall sensor works for DC and low frequencies, then the current transformer takes over for higher frequencies. Sometimes these are closed-loop schemes to eliminate gain error in the Hall-sensor.

2) Hall sensors are subject to many errors. Residual field requiring degaussing, offset and gain errors, positional errors, stray field errors. If all of these things are controlled or calibrated they can be very accurate. My opinion is that a shunt resistor is typically capable of being more accurate, though it can dissipate power, drop voltage and of course is not non-contact.

3) See 1.

\$\endgroup\$
  • \$\begingroup\$ Thank you @john D. So, for inrush current measurements on FPGA power supplies during bring-up, is it better to use shunt resistor? \$\endgroup\$ – Jigish Parikh Sep 2 '16 at 19:23
  • \$\begingroup\$ How accurate do you need to be? I'm fine with using a current probe for measurements like this, but if you need very high accuracy you could consider a shunt. \$\endgroup\$ – John D Sep 2 '16 at 21:54
  • \$\begingroup\$ Readings would be under 30 mA, so accuracy is important. I can afford up to 1-2% margin. \$\endgroup\$ – Jigish Parikh Sep 3 '16 at 0:00
1
\$\begingroup\$

The main benefit using Hall probe is a galvanic isolation. Special care must be taken using shunt resistors, which is easiest method and also most reliable and cheap.

A hall effect probe may be closed-loop or open-loop.

  • A closed loop eliminates all non-linearity of the sensor itself, since there is a secondary coil used to bring the total flux to zero. The hall sensor in a closed loop is used to detect zero flux. No high bandwidths are possible because the secondary winding has inductance, therefore the current can't change very fast.

  • In open loop the bandwidth is greater, but the precision is smaller due to non-linearity, although hall element has excellent linearity, so it is necessary for the user to determine the precision.

\$\endgroup\$

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.