How to Make a Shunt Current Sense Resistor - YouTube

Hi, I want to make a *beep*

I said shunt!

I want to make a *shunt* current sense resistor.

Basically the way we measure current is by converting it into voltage

passing it through a known resistor

which we call a shunt and so the current is the measured voltage divided by the known resistance.

To measure the current accurately you need a relatively accurate shunt resistor, for example here

I have a 1V 5% resistor and I'm gonna run two amp through it and measure the voltage across it

See now the voltage reading here is actually showing that there is 1.8 amp running through it.

*beep*, it's burning now *beep*

*beep*

*beep*

Of course that means 4 watts of power that burns my resistor. You can buy higher power *shunts* that are made

for this purpose, but why pay so much when you can make a relatively accurate *shunt* yourself

you just need some length of wire. Every wire has a known resistance. If you search Wikipedia for

AWG or American Wire Gauge, you will find a table showing

information for every wire size such as its resistance per length and diameter and such and of course there are other

standards too, like British standard wire gauge. So basically using a *shunt* resistor, you're generating a

voltage drop that's not very

desirable. Why? Because then you are dropping your supply voltage level by that much, which will affect the

current running through the load.

But if you keep that voltage drop small or compensate for it by increasing the supply voltage, then you are fine

I like to keep my *shunt* maximum voltage small. To start,

I like to have only 100 millivolt for a maximum of 1 amp, which means around hundred milliohms.

Regular multimeters are no good to measure such low resistances,

but they can measure voltage and current much more accurately.

What I have here is a 22 gauge wire, which is around

53 milliohms per meter, and so I would need around 2 meters of this wire. Now I set my power supply to

output 1 amp accurately, and I run the 1 amp of current through my 2 meter length of wire

and I measure the voltage across it and I have to read hundred millivolts,

but you see I'm reading a bit off. So I stripped some lengths of wire, and I'm gonna slide my probe across it

and you see as I slide, the voltage starts dropping

I'll keep going until I get exactly 100 millivolts right there

This means that to have exactly 100 milliohms

I have to measure exactly between these two probe points so I marked them

I'll solder wires to the two spots

I marked on my *shunt* wire

And I measure the voltage between these two wires.

The length of the measurement wire is not important because there is no current running through it.

Now I'll wind my wire like this so it takes less space.

This is bad because I'm making an inductor. If I try to measure higher frequency

AC spikes or transients, this will filter them.

It's easy to solve, though.

First, I bend the wire at the center

and then I wind it like this. This will remove the inductance.

You think I'm joking? No! If I bend the wire like this and send the current in on the top wire it creates

fields like this and when it returns on the bottom wire it creates fields the opposite way

and the fields cancel each other, so no inductance is created. Elementary!

So here's the *shunt*.

Now let's do a comparison between readings of the supplies and meter and our *shunt*.

*beep*

Who set everything at maximum?! I guess I did. What can I say, I like extreme *beep*

So here we are. I set the current to 100 milliamps, and you can see that the supply can't show it accurately, but

the shunt can. That 10 millivolt voltage times 10 is the current running through the shunt which is 100

milliamps.

And if I change the current,

our reading is still pretty accurate.

I'd like to be able to measure higher currents like 10 amps,

but using this, the voltage drop would be 1 volt, which is too high for lower supply voltages,

so in order to get the same 100 millivolt drop,

I have to change the resistance of my shunt to 10 milliohms.

Simple, I can still use the same *shunt* and just inject the current and read it at 1/10 length of this wire.

Let's just see if it can handle 10 amps, or it will burn. I'll run 10 amp through a short piece and

see if it gets warm.

Well, see ther mouth and the lip area is more sensitive to heat

Slowly getting warm, but it's not too bad,

so you should be fine. And if you are wondering how I'm running 10 amps through my mouth without dying,

there is no current running through my body

because the voltage across my body is zero so the entire 10 amp is running through the wire.

First, I strip this wire at one-tenth of the length with some margin for a calibration,

which is around 182 millimeters in this case.

Then I solder a wire here, and that's where I'll inject my 10 amp current.

Now I'm running 10 amps through my wire and have to see exactly at which point I'll read exactly

hundred millivolts.

Right around here, and I'll mark it and I solder my measuring wire to that spots

And this time to get rid of inductance, instead of winding it, I'll just fold the wires side-by-side

Something like this and now we can tape it together in a bundle and label the wire,

so we know which one is for which and also don't short the bare wires

inside the shunt otherwise it will change the resistance and throws your reading off.

Now let's step up the game.

I want to be able to measure up to a hundred amps, but my supply here

can't output more than 10 amps, so I'm gonna use my auto transformer that can output some high AC currents.

I have a piece of 22 gauge wire here,

which I'll measure the voltage across. If I connect it to the output of my auto transformer, I...

*beep* *beep* *beep* *beep*

And burn my fingers again.

The piece of wire simply shorts the AC output,

so I start at 0 volt and slowly raise the voltage while I'm measuring the current with my clamp meter. Let's see...

*beep*

*beep*

I'm starting to think coming up with the idea of a light bulb didn't require much genius. in fact the dumber

you are the more likely you are to invent a light bulb.

And remember, if your *shunt* gets too hot, its resistance rises so much that its current reading won't

be accurate anymore

We definitely need thicker wire.

I have this 18 gauge wire which is around 21 milliohms per meter,

so 400 amps and 0.1 volts drop, I would need around 4.75 centimeters of this

Okay, let's see if we can do the hundred amp now

Yeah

Seventy-seven, starting to smoke

Come on

Eighty...

*beep*

Melted again. We need thickness! And fortunately, I'm in luck,

I have some 8 gauge wire which is around 2.1 milliohm per meter

so I would need around half a meter of this for one melliohm or

0.1. Volt per hundred amps. If this can't do 100 amp, nothing can.

There we go

Let's see

There we go, 100 amps and it's holding on pretty...

*beep*

*beep*

My auto transformer can't take that much power

*beep*

So I can't calibrate at 100 amps because my auto transformer blows up.

Anyways, I don't really trust this current clamp multimeters because they're fine, but they are not very precise.

I'm thinking just to calibrate the length of this wire at 10 amps using my precision multimeter,

which will give me enough resolution, and then I can use this up to 100 amps.

There we go, I'm running 10 amps through the wire, and I'll be looking for...

10 millivolts exactly.

Right around there.

Now let's do a comparison between the clamp current meter and my shunt.

Of course I trust my shunt a bit more.

Just remember that whatever voltage you read there in millivolts times a thousand is the current.

Okay, let's do it.

There, 50 amps...

47, 48...

Wow, actually the clamp is not very bad either, which seems like the current is rising.

*beep*

Because the auto transformer is burning

There, now we have a precise 100 amp shunt resistor. Now some of you might ask

"Why don't I just use a multimeter or a clamp current meter to measure current?"

No, no, that's fine, there are no stupid questions.

Three reasons: one is that the current might be outside the range of your multimeter. For example, if I want to

measure 10 microamps, I have to use a shunt of around 10 kiloohms and measure 100 millivolt across it.

In that case I can measure accurately down to nanoamps.

Secondly, clamp current meters are fine, especially since you don't have to cut through your contacts, and you just clamp around them.

But they are not very accurate at low currents and their precision can be as bad as five to ten percent,

where a shunt can easily be as accurate as 1%

Third and most importantly, the meters only measure DC and RMS,

but with a shunt you can connect it to a scope and see all the current waveforms and transients.

I have an example set up here: I have my hundred milliohm shunt resistor

I've made in series with the primary of my microwave transformer, and there is no load on it,

and I'm gonna measure the voltage and current of the AC line on my scope.

Let's see...

You can clearly see that the yellow current line is leading the voltage green line

as should happen in an inductor.

What? Why the *beep* the current is not a sine wave

There. It says the current should look like that if you are using an iron core because the iron core doesn't have a linear magnetization

The *beep* you see and learn using a *shunt*

Thanks to the circuit specialists for providing the essential tools I need to make my videos.

They will provide five of these awesome LCR meters

and 5 of these USB oscilloscopes to my viewers and patrons.

I mean their website is filled with this great low-cost tools that you can buy like this thing that you just plug

your component into and it will tell you that yes,

It's a NPN transistor, and here are the parameters.

As usual the patrons, are automatically in the draw,

but for the viewers, if you need these tools, please leave a comment under the video like the usual