Measuring 100A DC current

[gaminn]

Hi,
occasionally I need to measure current up to 100A DC, which is obviously too much for any handheld multimeter. I found e.g. this shunt https://www.aliexpress.com/item/1005005721758936.html. It promises +- 0.1% accuracy. I would use digital multimeter to read voltage drop accross the shunt (75 mV @ 100A) and then calculate the current. It would be nice if the shunt was 100 mV, then no calculation would be needed....

Is there something better for max 100 USD that can measure 100A +- 200 mA?

[Martin72]

Hi,

It should not be more than 100mV.
At 100A, this is already 10W, which heats up well and influences the measurement result, which has a corresponding effect on the accuracy.
Here is a shunt that has 0.25% tolerance and 1mV/A:
https://de.rs-online.com/web/p/shunts/8233589

[BeBuLamar]

For such a high current I would use a clamp meter which should have sufficient accuracy and much easier to use.

[Martin72]

But that is precisely the problem.
Finding a current clamp that is that accurate.
If you're lucky, you'll find one with 1%, which is already very accurate for a current clamp, but costs much more than a current shunt, which is even more accurate.

[Stray Electron]

    Even the best clamp on DC amp meters (Fluke, AEMC, etc) are only rated for 3% accuracy.  Forget the cheap made Chinese meters.  Clamp on amp meters are pretty trivial but DC current meters are very tricky and expensive.

   I would use a good quality American made current shunt and meter and keep your measurement time short in order to avoid heating up the shunt.  American made current shunts are almost always designed to output 50 mA at the fully rated current so a 100 Amp shunt will need to dissipate 5 Watts of power (0.050 x 100) so power dissipation and heating of the shunt shouldn't be hard to deal with. 

   Good quality current shunts and the corresponding 50mA meters can be found cheaply on Ebay. Here is just one example, this is a 400 Amp shunt that outputs 50mA.  https://www.ebay.com/itm/404932876476?itmmeta=01HZQWBXFRHYMQP7DY3KJFZ527&hash=item5e47e158bc%3Ag%3AW70AAOSwB1JkLYUd&itmprp=enc%3AAQAJAAAAwIFn1Fck%2FY3gpneHseC0olXbRZuhefV3YEwPVSqu3eK714H%2BIDsGa%2FtWRJD67lBpuoJDAtWuYwJb1eMsIigKuPFBza5ZcxyyOUx3UR7Zk3L2KoTiieJOX%2FkXvAVv%2FBVoYe0hDPwBDjJdqj7J2p6ymW13DGTtMekzmSiJNCmJ9udmJKTsGW3ZRlp5%2FxcKsQbZTx9kTueS0W2QDtJUg5Oo9giGoiyZoFe9ZwkKEreNHj%2Fc2GJfjfSsFO37aUyUv3hUYw%3D%3D%7Ctkp%3ABk9SR4TYr_z9Yw&LH_ItemCondition=4

[Stray Electron]

  Here is a 150 Amp rated 50mA shunt made by Weston (the people that used to make good quality meters) for $19.95  https://www.ebay.com/itm/276027384654?itmmeta=01HZQWKWGV8E21VKAVNKFQJRAT&hash=item404483a34e%3Ag%3ATRUAAOSwTlBk8z8M&itmprp=enc%3AAQAJAAAAwL6GTXNiQgU11FXOLXudpRlvuScMCN%2FCA%2BUqNyNLt%2B1IWEiEPASvKY6sP1jIw3QSIKks6ZB3u%2FNJgxQlO%2FKKr9Y94tPs73AlTfA%2FQI0jkY6ATJ%2B43Tg3sVRUMQssyz2H8A9tQatR4C9zD2%2F9ahhAG9TD8LlR%2FGEaC0tvHh6KFZ1cW%2FWLS3V5r2BYtF3pMsobB2x74RXONa860ITg%2FnpZ2mXBgO2oV37eDahGy6FvbW0Cn29Z%2FDgxRE80Xl%2BTCFbhow%3D%3D%7Ctkp%3ABk9SR8bIz_z9Yw&LH_ItemCondition=4

[johansen]

look for closed loop LEM hall effect sensors on ebay.


this is a random datasheet i found https://www.lem.com/sites/default/files/products_datasheets/lah_100-p.pdf

-1db out to 200KHZ and the accuracy is .25%

[Fungus]

Is there something better for max 100 USD that can measure 100A +- 200 mA?

Is there a better method? Not really.

Is that shunt good? I'd vote "probably".

It has a free 90-day return policy so what have you got to lose? 

If it were me I'd spend $7 more and go for the 300A version. It'll heat up three times less and 25mV is an easier number to work with in your head than 75mV.

(multiply by 4 instead of multiplying by 1.3333) 

https://www.aliexpress.com/item/1005005748456772.html

[gaminn]

Thanks, I will buy that Murata 100A 100mV shunt https://www.digikey.com/en/products/detail/murata-power-solutions-inc/3020-01108-0/3438619, it is quite cheap on Digikey. Hopefully I will be able to screw in some banana sockets (and they will not melt...) instead of that 8-32 screws to make my measurements easier.

[trobbins]

Perhaps best to indicate what your voltmeter is, and its accuracy, and resolution if you want to use the shunt for say a 5A 1% level measurement, as you will be down below 10mV and sort of need 100uV sensitivity at least.

[gaminn]

I have Keithley 2000 or Agilent 34411A. I'm not sure what is their accuracy, also they haven't been calibrated for a long time, but I hope it will be an order of magnitude better than the shunt.

[tooki]

Hi,
occasionally I need to measure current up to 100A DC, which is obviously too much for any handheld multimeter. I found e.g. this shunt https://www.aliexpress.com/item/1005005721758936.html. It promises +- 0.1% accuracy. I would use digital multimeter to read voltage drop accross the shunt (75 mV @ 100A) and then calculate the current. It would be nice if the shunt was 100 mV, then no calculation would be needed....

Is there something better for max 100 USD that can measure 100A +- 200 mA?

If you really care about accuracy, do you really think you should be buying from AliExpress?

[gaminn]

If you really care about accuracy, do you really think you should be buying from AliExpress?

That's why I expect 0.2% accuracy from 0.1% Chinese shunt.

[nctnico]

Thanks, I will buy that Murata 100A 100mV shunt https://www.digikey.com/en/products/detail/murata-power-solutions-inc/3020-01108-0/3438619, it is quite cheap on Digikey. Hopefully I will be able to screw in some banana sockets (and they will not melt...) instead of that 8-32 screws to make my measurements easier.

Personally I'd buy a bigger shunt (like one rated for 200A) if a current of 100A is flowing for longer than a couple of minutes. Otherwise you'll operating it at its limits and it will get quite hot. Also the temperature coefficient may add to much error when a shunt is operated near its power limit.

[tooki]

If you really care about accuracy, do you really think you should be buying from AliExpress?

That's why I expect 0.2% accuracy from 0.1% Chinese shunt.

I like your optimism! :p

[coromonadalix]

 

even with  34410a 34411a  or k2000   even uncalibrated  you get around 0.003% precision   but  your shunt will never be as precise  loll

and yes  take an 200 amp minimum to avoid drift

and please be realistic of your needs and measurement capacity   

[ArdWar]

I don't think even the best reputed shunt manufacturers offer 0.2% accuracy, much less 0.1%.

It's current shunt, if you need accuracy you're expected to calibrate out for gain errors. The sought after qualities here is usually the temperature coefficient instead.

[gaminn]

I'm probably OK with that 100A Murata shunt, most often I will be in 50 - 60 A range. If I need it I will buy 200A 50mV version later.

I didn't know max. shunt accuracy offered by reputable manufacturers is 0.2%. Then the 0.1% chinese accuracy is bullshit. However I still like their shunt design with banana sockets.

[gaminn]

even with  34410a 34411a  or k2000   even uncalibrated  you get around 0.003% precision   but  your shunt will never be as precise  loll

Yes, that's what I need - no more error added by the multimeter.

[mariush]

Why not just parallel a bunch of resistors to get a low resistance ...

ex parallel 10 of these 10mOhm 4W resistors with 20 ppm/C to get a 1mOhm current shunt : https://www.digikey.com/en/products/detail/isabellenhuette-usa/VMS-R010-1-0-U/16836578

[Martin72]

Hopefully I will be able to screw in some banana sockets (and they will not melt...) instead of that 8-32 screws to make my measurements easier.

Banana sockets are not designed for such high currents, which is why most power supplies that deliver 20A and more also have regular terminals.
At higher currents, contact resistances can no longer be neglected.
I really hope that if you “drive” currents over 10A, you no longer do this with normal banana cables...

[Psi]

A little 10x/50x/100x gain current sense amp IC and a good quality precision current-sense resistor, or array of them, is how i would do this.
You need some gain to keep the resistor power loss manageable at 100A
The current sense amps output 0-5V.  So with some careful selection of the right resistors you can make the output match current.
1V = 100A for example.

[coppice]

If you really care about accuracy, do you really think you should be buying from AliExpress?

That's why I expect 0.2% accuracy from 0.1% Chinese shunt.

The problem isn't with China, its with AliExpress. It could be E-Bay or any other unregulated source, and your chances of good results are just as bad. Get things through a well regulated supply chain from a Chinese maker and they will most likely perform to spec.

[coppice]

I don't think even the best reputed shunt manufacturers offer 0.2% accuracy, much less 0.1%.

It's current shunt, if you need accuracy you're expected to calibrate out for gain errors. The sought after qualities here is usually the temperature coefficient instead.

For 99% of uses a high accuracy shunt makes no sense, as its cheaper to get a shunt with a great temperature coefficient, and calibrate out its errors. However, there are certainly 0.1% accuracy shunts on the market.

[trobbins]

Both those meters have the same 1yr 100mV range accuracy, and only have a 3Adc current range (the Keithley seems much better spec).  You could compare both those meters at say 50-100mV to get some confidence that they haven't strayed far from their original accuracy, but be mindful that unless you keep the meters on full-time and keep them close to 23C ambient, the meter may have a significant tolerance itself, and especially for readings well below 100mV FS from the shunt.

Even with the +/-0.25% shunt, your uncertainty could still exceed 1% when used with your meters if you don't take some effort to cross-check your meters, and importantly use a fan on the shunt to keep its temp rise as low as practical (even at 50-60A where the accuracy of the meter gets worse).  But I think that is the best and easiest path to take.

Martin72, I think the comment about banana plugs was for the voltage sensing connection to the voltmeter - perhaps look at the link to the 0.1% shunt, and to the Murata 0.25% shunt for context.