High Power DC current sensing - How to?

[schwarz-brot]

Hi Folks,
right now I am working on a project that requiers current sensing in a higher power environment. This means exactly:
I need to sense the power in a circuit with about 550V at a maximum of about 200A.
I would choose standard Hall Effect current transducers but I am not sure about their abilities of sensing about 10A... Have you some suggestions?
I do not like the idea of a shunt resistor with that a high power.

Any Ideas?

No transformer possible - it is a pure DC circuit.

(We are speaking about electric cars...)

[Dago]

Hi Folks,
right now I am working on a project that requiers current sensing in a higher power environment. This means exactly:
I need to sense the power in a circuit with about 550V at a maximum of about 200A.
I would choose standard Hall Effect current transducers but I am not sure about their abilities of sensing about 10A... Have you some suggestions?
I do not like the idea of a shunt resistor with that a high power.

Any Ideas?

No transformer possible - it is a pure DC circuit.

(We are speaking about electric cars...)

What do you mean "sensing about 10A"? Something like ACS770 from Allegro http://www.allegromicro.com/en/Products/Current-Sensor-ICs/Fifty-To-Two-Hundred-Amp-Integrated-Conductor-Sensor-ICs/ACS770.aspx has a specified maximum error over the whole lifetime of 1.9% which equates to under 4A for the 200A model. Typical noise is 20mV and the sensitivity is 80mV/A so the noise limits your resolution to "sensing" about 250mA.

[David Hess]

Given the voltage and current requirements and safety considerations, I would use a DC current clamp.

[max_torque]

I'd just use the standard hall effect sensors from companies like LEM. ie:

LEM automotive current sensors

[Fank1]

The easy way is with a shunt resistor.
I've used them for up to 2000 A.
The output is usually 50 MV or 100 MV full scale.

[max_torque]

If i were messing around with 550v DC i'd want to know the isolation requirement was met by my current sensing solution!  Using an off-the-shelf LEM or similar gives you that for (almost) no work on your behalf.  For a low volume (one off) requirement, this is a no-brainer. 

(i'd also recommend you consider installing an isolation/insulation monitoring system too. An ISOMETER gives you early indication of any issue (which for a one off EV certainly can occur) and gives you a suitable shut down signal as you approach such a condition)

[Neverther]

Hall effect sensors arent bad.
I got CEM DT-3351, which can sense up to 1500DCA, but at 400A range I can still be pretty sure at 100mA currents. Below 100mA it can still detect and show current, but the accuracy and trustworthyness is somewhere eround the corner.

More info needed: is the current sensing for permanent control or just for measuring/monitoring stuff during the project developement.

[David Hess]

Some hall effect based current sensors are better than others.

The straightforward way to measure DC current is to simply use the output from the hall effect sensor but their offset and gain are sensitive to temperature limiting accuracy.  A more complicated way is to introduce a counter current and use the hall effect sensor to measure balance so its linearity and temperature coefficient of gain become irrelevant.

I hope newer ones are better because my old Fluke DC current clamp has a pretty limited temperature range.

I see that Allegro is now hiding some of their datasheets. :/

[T3sl4co1l]

I can vouch for LEMs, being used frequently in industrial equipment.

If you need higher accuracy, a special purpose CT (current transducer, that is) might be suitable, maybe one of those that works by magnetic saturation (VAC and a few others make them).  Otherwise, for best results, you'll need a shunt, and an analog isolator or full on digitization (e.g., an ADC on the high side with SPI over optos).

Tim