Measuring 100A DC current

[Njk]

Consider a welding machine Hall sensors (available as spare parts). Something like that

[geggi1]

If you want to measure DC very accurate look into fluxgate current sensors.
As far as i know it is only Lem and Danisense that sell these.
https://danisense.com/flux-gate/
https://www.lem.com/en/fluxgate-current-sensors
They are expensive but you can get precision down to hundreds of percentage. 

[Hydron]

Old ultrastab sensors (i.e. what was suggested above) go for pretty cheap on eBay. They have a current output of a specified ratio to the measured current, and you can always just use a bench meter in current mode (even with autoranging) to measure the output if you don't have some precision resistors better than the internal meter shunts. If you have a decent meter with a scale function then you can even get direct readings.

[Martin72]

I find the idea of using Danisense or Ultrastab quite charming.
We have both for our ZES LMG power analyzers, they are accurate.
And we had an LEM converter built into a case plus supply, plus huge terminals - as a current transformer for currents up to 2500Adc.....

[Stray Electron]

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...

   X2. Adding additional wires and sockets to the shunt is going to wreck the accuracy of the circuit! The resistance of a 100 Amp 50 mV shunt is only 0.0005 Ohms. And the OP going to add wires to it!  What is the resistance of your wires? And the resistance of the sockets? No where near as low as that of the shunt, I can guarantee.

   The math:  OMG!  With a 50 Mv 100Amp shunt; every millivolt represents 2 amps of current!   So millivolts x2 equals the current in amps.  How much simpler could the math be?  If you use a 150 amp 50 mV shunt then millivolts x3 equals the current in amps. 

    And if you (the OP) takes a look at the various brands of shunts being sold you'll see that the better ones have four terminal connections, two large ones for the high current circuit and two smaller terminals for the voltage metering circuit. It forms a four wire Kelvin connection and the theory behind it is exactly the same as that for very accurate power supplies and very accurate meters and very accurate electronic Standards.

    The OP is very naive is he thinks that a 200 Amp shunt from Alibaba is going to be anywhere near as accurate or the size or will  dissipate as much heat as a good quality 100 Amp American shunt!  Chinese products simply aren't made that way.

   If you look at the shunts on Ebay, you can easily find shunts that are rated for 0.25% accuracy and that are made by companies like Weston, Fluke, and Simpson, that actually build meters and understand how measurements are supposed to be taken.  Specs for many of the US and Canadian made ones are online and you can look up their tolerance and have some degree of faith that they will actually meet their spec, unlike anything made in China today.

[trobbins]

StrayElectron, the OP is proposing to use a 'standard' current shunt - that is a 4-point measurement.  That is also quite an acceptable method for 1% uncertainty if appropriately managed, and the cheapest and most convenient, given the two meters available.

Also you just provide a stereotyped response to anything made in China.  A large number of top rated companies set up shop in China - their decision, and have sold worldwide.

[Fungus]

    The OP is very naive is he thinks that a 200 Amp shunt from Alibaba is going to be anywhere near as accurate or the size or will  dissipate as much heat as a good quality 100 Amp American shunt!  Chinese products simply aren't made that way.

Complete rubbish.

The Chinese are perfectly capable of making quality products if you're prepared to pay for them.

The $10 shunts are probably not much use but a $75 solid copper one? What exactly could go wrong? A chunk of metal is a chunk of metal no matter where it's made.

Shunt calibration isn't difficult - manufacture them to be a bit on the low side then use a file to increase the resistance to spec. The shunts inside every multimeter you own will have a few little nips in them if you look.

I've bought a lot of really nice stuff on Aliexpress. The only things I definitely wouldn't buy are power MOSFETS, 16850 batteries, and multimeter fuses. They're almost guaranteed to be fake.

Everything else? I'd give them a chance. Aliexpress returns are easy peasy, lemon squeezy - just drop the thing off at a collection point (usually the same places as take Amazon returns take Aliexpress returns too).

[BillyO]

    The OP is very naive is he thinks that a 200 Amp shunt from Alibaba is going to be anywhere near as accurate or the size or will  dissipate as much heat as a good quality 100 Amp American shunt!  Chinese products simply aren't made that way.

Complete rubbish.

I agree completely.  I get sooo tired of all the China bashing that goes on.

If your willing to pay $27,000 for a scope, the Chinese will make you a$27,000 scope that is probably better than any $27,000 US made scope.  However, if you only have $40 to spend on a scope the Chinese can help you out and I guarantee it will be better than every $40 US made scope.

There is no reason to believe those AliExpress shunts are garbage.  None at all.

[Martin72]

The Chinese are perfectly capable of making quality products if you're prepared to pay for them.

Absolutely.
But it's also true that you get what you were prepared to pay for.
Because even the Chinese can't do magic.
If you pay €10 for a shunt, you get a shunt for €10, but don't expect the properties of a €100 shunt (which may have a different material property so as not to drift too far away at higher temperatures).
If you are aware of this, everything is fine.
If you ignore the circumstances that make it so cheap in the first place, all the better.

[Stray Electron]

    The OP is very naive is he thinks that a 200 Amp shunt from Alibaba is going to be anywhere near as accurate or the size or will  dissipate as much heat as a good quality 100 Amp American shunt!  Chinese products simply aren't made that way.

Complete rubbish.

I agree completely.  I get sooo tired of all the China bashing that goes on.

If your willing to pay $27,000 for a scope, the Chinese will make you a$27,000 scope that is probably better than any $27,000 US made scope.  However, if you only have $40 to spend on a scope the Chinese can help you out and I guarantee it will be better than every $40 US made scope.

There is no reason to believe those AliExpress shunts are garbage.  None at all.

     Everyone here seems to overlook the fact that the OP said that he wanted something cheap. He's not going to buy any top of the line product, Chinese or otherwise.   It's not a question of whether the Chinese can build a good product but more a question of what kind of quality when you buy from an unknown seller on Alibaba at the cheapest price that you can find!

    FWIW I have no reason to bash the Chinese products, there are plenty of postings on this forum and many others, complaining about the quality of every Chinese electronic product that you can name!  Fake transistors, counterfeit ICs, solder that won't stick to anything, "copper" wire that can't be soldered to; tiny capacitors inside of large cans with much higher ratings, batteries with physically impossible amp hour ratings, a water heater that uses the outer metal body as a current carrying conductor;  the list goes on and on!  I know from experience that even when Chinese products to work, most of them are rated right on the edge of destruction and there is no margin for error. And that's why with everyone here worried about the heat dissipation of the OP's shunt, I would not recommend buying one from China, regardless of the price. 

[johansen]

    The OP is very naive is he thinks that a 200 Amp shunt from Alibaba is going to be anywhere near as accurate or the size or will  dissipate as much heat as a good quality 100 Amp American shunt!  Chinese products simply aren't made that way.

Complete rubbish.

I agree completely.  I get sooo tired of all the China bashing that goes on.

If your willing to pay $27,000 for a scope, the Chinese will make you a$27,000 scope that is probably better than any $27,000 US made scope.  However, if you only have $40 to spend on a scope the Chinese can help you out and I guarantee it will be better than every $40 US made scope.

There is no reason to believe those AliExpress shunts are garbage.  None at all.

     Everyone here seems to overlook the fact that the OP said that he wanted something cheap. He's not going to buy any top of the line product, Chinese or otherwise.   It's not a question of whether the Chinese can build a good product but more a question of what kind of quality when you buy from an unknown seller on Alibaba at the cheapest price that you can find!

    FWIW I have no reason to bash the Chinese products, there are plenty of postings on this forum and many others, complaining about the quality of every Chinese electronic product that you can name!  Fake transistors, counterfeit ICs, solder that won't stick to anything, "copper" wire that can't be soldered to; tiny capacitors inside of large cans with much higher ratings, batteries with physically impossible amp hour ratings, a water heater that uses the outer metal body as a current carrying conductor;  the list goes on and on!  I know from experience that even when Chinese products to work, most of them are rated right on the edge of destruction and there is no margin for error. And that's why with everyone here worried about the heat dissipation of the OP's shunt, I would not recommend buying one from China, regardless of the price.

Yep that why i said look on ebay for LEM closed loop sensors.

I bought 3, 100 amp units for maybe 30$ last year.

[pdenisowski]

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.

How about a current probe?  We (R&S) make current probes that measure over 100 Amps DC with 1% accuracy.

https://www.rohde-schwarz.com/us/products/test-and-measurement/oscilloscope-probes/current-probes-for-oscilloscopes_63493-73797.html

[Fungus]

    FWIW I have no reason to bash the Chinese products, there are plenty of postings on this forum and many others, complaining about the quality of every Chinese electronic product that you can name!  Fake transistors, counterfeit ICs, solder that won't stick to anything, "copper" wire that can't be soldered to; tiny capacitors inside of large cans with much higher ratings, batteries with physically impossible amp hour ratings

In other news: People don't read seller ratings, buy things at prices that are too good to be true, and are predisposed to hate on "China".

If it doesn't work, send it back. 

PS: I can buy Fluke multimeters on Aliexpress, Rigol/Siglent oscilloscopes, etc.

[coromonadalix]

the problem is :    when you say  "cheap"  and put 6.5 digit meters on them,  what do you expect ?

i have 6.5 digit dmm's  and 1 amp of resolution is enough for 200 amp shunts

[Calvin]

Hi,

similar as #6 and #28 I´d throw Hall- or Fluxgate sensors in the ring 
See for example the N4646 series from Vacuumschmelze sourceable from Mouser (say the T6044-N4646-X412) for less than 17€/pcs.
I have KiCad PCB-designs for the X400, X410 and X412 and the voltage-output 15A type X662.

regards
Calvin

[AVGresponding]

If "cheap" is the primary operand, I'd buy a couple of hundred of something like these and bin them to get a hundred with the correct values to make a 500W 10mΩ 1% shunt. Also you'd have a hundred spare resistors for your stock afterwards.

[nctnico]

If "cheap" is the primary operand, I'd buy a couple of hundred of something like these and bin them to get a hundred with the correct values to make a 500W 10mΩ 1% shunt. Also you'd have a hundred spare resistors for your stock afterwards.

The self inductance will be too high when using wirewound resistors. Especially if the current is switched off quickly you can get quite high voltage swings. This is already a problem with low inductance, planar resistors at 15A.

Another advantage of using a current shunt over a hall-based current probe is having a high pulse handling ability. So you can measure short pulses to hundreds of amps without needing a current probe which is rated for hundreds of amps (which may also have a higher offset / noise floor).

[AVGresponding]

So use a snubber diode or TVS   

[trobbins]

AVGresponding, how much is the OP going to pay himself to sort those resistors, and then to construct end terminals and then fit the resistors, then test the one off special ??

A shunt manufacturer has the facilities to test.  The OP doesn't.

[nctnico]

So use a snubber diode or TVS   

That is not the point. The added inductance can influence the circuit you are measuring and/or affect the measurement results.

AVGresponding, how much is the OP going to pay himself to sort those resistors, and then to construct end terminals and then fit the resistors, then test the one off special ??

A shunt manufacturer has the facilities to test.  The OP doesn't.

Indeed. And 1 Ohm is low enough for the leads to also have a significant contribution to the resistance. Not to mention the leads won't be made from a low tempco. material.

That is the reason why current shunts have sense terminals as close as possible to the resistive material; to make sure the leads / connections to the load have a minimum impact on the accuracy of a shunt.

[HighVoltage]

We discussed this before:
https://www.eevblog.com/forum/metrology/calibrating-a-100a-dc-shunt/

Not in your price range but this is what I have used very successfully for 100A DC:
The RUG-Z-R001-0.1-TK1

https://www.distrelec.de/de/leistungswiderstand-1mohm-250w-isabellenhuette-rug-r001-tk1/p/16057580?queryFromSuggest=true&itemList=suggested_search

[coppice]

The self inductance will be too high when using wirewound resistors.

Too high for what? Has there been any indication of how much inductance the OP's application can tolerate?

[johansen]

If "cheap" is the primary operand, I'd buy a couple of hundred of something like these and bin them to get a hundred with the correct values to make a 500W 10mΩ 1% shunt. Also you'd have a hundred spare resistors for your stock afterwards.

The self inductance will be too high when using wirewound resistors. Especially if the current is switched off quickly you can get quite high voltage swings. This is already a problem with low inductance, planar resistors at 15A.

Another advantage of using a current shunt over a hall-based current probe is having a high pulse handling ability. So you can measure short pulses to hundreds of amps without needing a current probe which is rated for hundreds of amps (which may also have a higher offset / noise floor).

You seem weirdly confident that 100 resistors in parallel will have too much inductance.

Does a 3 inch long shunt have too much inductance for you?

[BillyO]

You seem weirdly confident that 100 resistors in parallel will have too much inductance.

100 of those resistors in parallel would not have very much inductance, depending on the frequencies being applied.

However, the final device would be kind of large and unwieldy.  The resistors are about 1cm square and if you put 1cm between each for cooling then put them in a 10 x 10 matrix it would be square 19cm on a side!

[johansen]

You seem weirdly confident that 100 resistors in parallel will have too much inductance.

100 of those resistors in parallel would not have very much inductance, depending on the frequencies being applied.

However, the final device would be kind of large and unwieldy.  The resistors are about 1cm square and if you put 1cm between each for cooling then put them in a 10 x 10 matrix it would be square 19cm on a side!

didn't look at them till now. yeah would not use those.

6$ on amazon - 100 1/2 watt resistors.
https://www.amazon.com/EDGELEC-Resistor-Tolerance-Multiple-Resistance/dp/B07XLBY5RF