Making a Shunt Resistor

[rick.hughes6]

As part of another project, would like to take some High Current DC measurements.

I have a couple of multimeters ... but obviously not going to be any good for around 80-100A

Now going back to school physics lessons I could use a Shunt resistor ...

issue will be obtaining a suitable shunt ... assuming I use a 1 ohm shunt, anybody made one of these ? The WEB ref page: using 12guage copper wire .. based on 1.619 ohm per 1000'

Although maybe a resistance wire may be better eg Nichrome around 10.58" for 1 ohm. Welcome to any better ideas.

[Andrew Gabriel]

around 80-100A

anybody made one of these ?

page:

details using 12guage copper wire .. based on 1.619 ohm per 1000'

You probably don't want resistance wire for this sort of current, and you don;t want anything like 1 ohm resistance.

As a teenager, I made a multimeter (moving needle type back then), and I used a length of wire coat-hanger as the shunt resistor with a 50microamp meter movement. Just needed a few inches IIRC, although I don't now recall what the current was for full scale deflection.

Later, I made a car battery charger, and used a length of plain copper wire as the shunt, folded back and forth a few times.

One important thing - make sure your sense wires are connected to the shunt wire only, and not to the shunt wire terminals/ends. Otherwise variation in the terminals will generate more noise than the voltage drop measurement in the shunt wire you're looking for, and it removes the risk of the shunt wire terminal failing such that the shunt drops out and everything passes though your sense circuit (for a millisecond or so until it burns out).

[John Walliker]

for around 80-100A

> The issue will be obtaining a suitable shunt ... assuming I use a 1 ohm s= hunt, anybody made one of these ?

Using copper wire for a shunt is fine if the temperature does not vary much. The link you posted overcomplicates things a bit.

1) You don't need to worry about the resistance of the meter if you use a voltmeter and calculate the current. 2) You can calibrate the shunt by first using a long length of wire and passing a current that one of your existing meters can measure. Then shorten the wire and scale the readings proportionately.

Make sure you use 4-terminal measurements. In other words, pass the high current through the sensing wire and add a pair of connections soldered onto that wire, one near each end. The millivoltmeter is connected to those soldered tappings.

John

[Martin Bonner]

Ouch! That's a serious current.

Ahem. Let's review those school physics lessons shall we?

Volts = Ohms x Amps

So 80A through 1 ohm will generate 80V. Further, the definition of volts is: Watts = Volts x Amp

So your shunt resistor is going to be dissipating 80 x 80 = 6400 W = 6.4 kW.

That's a serious amount of energy!

(Of course, what I expect will happen is that your shunt resistor will have much higher resistance than the rest of the circuit, and bring the current way down - or possibly kill it altogether.)

I suggest you need something more like 1 milli-ohm, which will produce a voltage of 80 mV (perfectly within the range of digital volt meter), and an energy output of 6.4W

(for example) ?

[tinnews]

around 80-100A

anybody made one of these ?

page:

details using 12guage copper wire .. based on 1.619 ohm per 1000'

A much more elegant (and still DIY) method is to use a Hall Effect device, no direct connection to the circuit being measured at all. The bigger devices (from 50 amps or so upwards) are just toroids through which you feed the current carrying wire.

Available from CPC, Farnell and others. Devices such as HTFS200-P and LTFS 50-NP should take you where you need.

[Tim Streater]

And you want it is series, don't you.

[John Williamson]

Depending how accurate they need to be, you can get a cheapish DC only current meter with ranges of +- 100A and +-600 Amps from most decent motor factors. Mine's made by Durite, Part No. 0-534-00.

They're sold for checking charging and starter currents without breaking the circuit, you just put the wire into a channel on the back of the meter.

e-bay auction's ended, but you get the idea...

I bought mine a few years ago, though. And no, it's not for sale, it's too useful. :-)

[<me9]

I'd just use a hall effect clamp on ammeter. CPC do a good one for up to

400A.

[The Natural Philosopher]

Problem with resistors that low is the leads themselves start to make a difference.

This is car starter motor type territory..think of the battery leads to THAT

[geoff]

That'd be two coathangers in parallel then ... or one 6" nail

[Bill Wright]

anybody made one of these ?

page:

details using 12guage copper wire .. based on 1.619 ohm per 1000'

I made up some ammeters that read up to 30A and the shunts were pieces of copper about 30mm long x 8 x 1, with terminal holes at each end. I'm remembering this from 2001 so I might be a bit out with the dimensions.

Bill

[Gazz]

if you are to do the measurements more than once, and would like to keep a log of them.... you might want to try something like :

's a power meter for the radio controlled plane bods, they are using brushless motors and lithium polymer batteries nowadays pulling/producing upto 100 amps, (my plane's motor is a tiddler, pulling 28 amps peak, and using a battery pack that weighs next to naff all... thank bod the days of ni-cads and brushed can motors are over :)

Anyway, check the voltage range of the meter suits your application if you decide to go for one of those meters, and the time they can handle the current you want to pull,

If you really want a shunt, i bought one off ebay a while back, something like 6 quid, was sold for motorhome/boat use, to measure the current being drawn from the leisure batteries,

[harry]

for around 80-100A

> The issue will be obtaining a suitable shunt ... assuming I use a 1 ohm s= hunt, anybody made one of these ?

These shunts are very low resistance and made by the manufacturers to very exact values. So you have no chance of making your own and expecting accurate amps readings. You will have a job to even determine the value of the internal resistance of your meter and hence find out what value the shunt should be.

[Martin Brown]

around 80-100A

anybody made one of these ?

page:

details using 12guage copper wire .. based on 1.619 ohm per 1000'

A 1R shunt will drop 100V at 100A and have to dissipate 10kW! Even a 0.01R dropping 1V will need to handle 100W at 100A without getting too warm (and so altering its own resistance). Forced air ventilation on a chunky heatsink would seem to be the best bet.

You would be much better off buying a 10x high power 0.1R resistors and a suitably large aluminium heatsink.

Are you absolutely sure you want to do this ? You don't sound at all safe. 10 of these on a decent heat sink ought to just about hack it.

Zetatalk is a bunch of delusional nutters led by "Nancy" that expected the world to end around 1996 when the "Radish" planet X arrived. World ending cults go quiet for a while after it doesn't.

Yes. Indeed it is the same bunch of delusional nutters see the home page:

you want to be first in line for a Darwin Award you would do well to ignore anything they say on that site!

nothing with the sort of power handling rating you would want.

[The Natural Philosopher]

around 80-100A

...

>>> The issue will be obtaining a suitable shunt ... assuming I use a 1 ohm shunt, anybody made one of these ?

page:

> details using 12guage copper wire .. based on 1.619 ohm per 1000'

it is in fact not hard at all, and if its a moving coil meter it will be in the spec as so many ohms per volt. If its digital you dont really care. You are using it as a voltmeter anyway, measuring millivolts.

[newshound]

around 80-100A

anybody made one of these ?

page:

details using 12guage copper wire .. based on 1.619 ohm per 1000'

Other posters have covered shunt design and issues in plenty of detail.

Pico Technology used to do a "clamp" type ammeter which you connect to a normal DVM, and which does DC as well as AC amps, but it looks as though this has been discontinued. ISTR it was about £50. Occasionally very handy for car diagnostics (I think it gives 1 mV per amp).

[newshound]

page:

>>> details using 12guage copper wire .. based on 1.619 ohm per 1000' >>

Oh yes, here it is

site search facility is crap!)

[Man at B&Q]

for around 80-100A

> The issue will be obtaining a suitable shunt ... assuming I use a 1 ohm s= hunt,

Do the maths. That's at least an order of magnitude too large.

MBQ

[harry]

ood for around 80-100A

...

> >> The issue will be obtaining a suitable shunt ... assuming I use a 1 oh= m shunt, anybody made one of these ?

0.58" for 1 ohm.

So if it's measuring milivolts, the internal resistance will be miliohms. And the shunt will be smaller resistance still. Micro ohms. So, a futile pursuit trying to make one. Which is why your multimeter manufacturers don't even try.

[Dave Plowman (News)]

I have one of these, and it's not really accurate enough on DC. Or rather not for my purposes. Although one which was would be handy.