EEVblog Electronics Community Forum

Electronics => Beginners => Topic started by: Boschi on September 27, 2017, 09:58:38 am

Title: High current "probe"
Post by: Boschi on September 27, 2017, 09:58:38 am
Hi, i need to measure some low voltage (12v) high current (100-200A)  stuff.

If possible i want at least 10khz of bandwidth in order to measure a 1Khz square signal.

I dont need high accuracy, 5% would be acceptable.

I am tempted to do a DIY shunt resistor with a sheet of stainless steel with a low ohmic value (100-500 micro ohm), use a ucurrent-like circuit to amplify the voltage and calibrate the gain of the circuit until i get 1mV per A.

My limited experience says that apart from  the TC of the metal it would be pretty precise and with a bandwidth of about 100khz ( max4238, 1Mhz of bandwidt).

Im missing something?

I'll post some schematic once i get home.

About calibration, i mean making about 9-10 amps flowing in the resistor, and calibrating the gain to get 1mV/A.

Will it be linear enough?


Thanks for the answers.

Inviato dal mio A0001 utilizzando Tapatalk

Title: Re: High current "probe"
Post by: Giaime on September 27, 2017, 10:05:32 am
For those currents, I would suggest an Hall-effect sensor, something like this:
http://www.tamuracorp.com/clientuploads/pdfs/engineeringdocs/L32PXXXS05FS.pdf (http://www.tamuracorp.com/clientuploads/pdfs/engineeringdocs/L32PXXXS05FS.pdf)
Title: Re: High current "probe"
Post by: capt bullshot on September 27, 2017, 10:17:37 am
Or something similar by LEM:

http://www.lem.com/hq/en/component/option,com_catalog/task,displaymodel/id,90.13.36.000.0/ (http://www.lem.com/hq/en/component/option,com_catalog/task,displaymodel/id,90.13.36.000.0/)
Title: Re: High current "probe"
Post by: oldway on September 27, 2017, 11:33:07 am
Low cost solution is to use the voltage drop of one the cables and to amplifier and measure it.
Gain of the operational amplifier must be ajusted to calibrate the signal.
Title: Re: High current "probe"
Post by: xani on September 27, 2017, 12:57:16 pm
Current clamp ?
Title: Re: High current "probe"
Post by: Jeroen3 on September 27, 2017, 01:06:40 pm
Fluke has several current probes with BNC adapter. The i1010 for example, €400. Or the i410, but that's only 3Khz.

But DIY using a shunt is certainly possible. You can just buy the shunts in the required value.
Title: Re: High current "probe"
Post by: mk_ on September 27, 2017, 03:55:15 pm
Hi, i need to measure some low voltage (12v) high current (100-200A)  stuff.

If possible i want at least 10khz of bandwidth in order to measure a 1Khz square signal.

I dont need high accuracy, 5% would be acceptable.

I am tempted to do a DIY shunt resistor with a sheet of stainless steel with a low ohmic value (100-500 micro ohm), use a ucurrent-like circuit to amplify the voltage and calibrate the gain of the circuit until i get 1mV per A.


Take a look at sensitec-Current-sensors.  If you do the math right you don`t even to calibrate (it`s more or less some kind of opamp-gain)

Take care about the inductive part of your microohm-shunt.... this can result in some surprising measurements.... (t=L/R, so... if you expect something below 1uH you still have a long tau because R is very low too)
Title: Re: High current "probe"
Post by: Boschi on September 28, 2017, 08:42:45 am



Take care about the inductive part of your microohm-shunt.... this can result in some surprising measurements.... (t=L/R, so... if you expect something below 1uH you still have a long tau because R is very low too)

This is exactly the reason why i ask here!

Im not very used to work with inductor, so ill study the math these days.

I want to keep it as DIY as possible, as a student im pretty broke [emoji14]

Anyway, this is basically the circuit, ill just connect the shunt directly whitout using any switch.

Link:

http://www.eevblog.com/files/uCurrentRev5schematic.pdf (http://www.eevblog.com/files/uCurrentRev5schematic.pdf)

About buying the resistors, are they better? Also, here is very important using 4 wire connection, any keyword to use?

If i choose to use the DIY shunt, any way to make it better about the inductance?

Like giving a u shaping to the resistor in order to cacel out the magnetic flux.

Some reference that i think are useful: https://meettechniek.info/instruments/shunt-resistors.html (https://meettechniek.info/instruments/shunt-resistors.html)

A u shape would be like a basic bifilar winding thing, so what culd i expect from something like that?
Title: Re: High current "probe"
Post by: max_torque on September 28, 2017, 02:40:37 pm
Check out the Sensitec range of high bandwidth current sensors:

http://www.sensitec.com/products-solutions/current-measurement (http://www.sensitec.com/products-solutions/current-measurement)

can be used stand alone or shunted to increase range!
Title: Re: High current "probe"
Post by: Jeroen3 on September 28, 2017, 07:22:54 pm
Played around with the sensitec CMS3005 just now. It's spot-on, and it responds to 1 mA! An 5A LEM doesn't get out of bed for 1mA.

Update: 2018-11-05, Sensitec discontinued the CMS and CDS families current sensors.