Calibrating high current shunt

Thread Starter

Gizer

Joined Aug 15, 2017
19
So im working on a basic microwave transformer converted to produce high current. I want to increase the current further by adding a second transformer in series and add power control maybe via a dimmer ok the primary.

so i need to be able to measure large currents and my clamp meter goes up to 100a only.

so my idea was to use my dc bench power supply with constant current, apply a known current (upto 2a), measure the voltage drop across my shunt and calculate the resistance.
This came to 1.3m ohm. (I did at 0.5a, 1a and 2a, all v similar)

next i attached the shunt to the high current secondary winding and measured the voltage drop here. (Primary connected to mains 240v AC supply). Voltage drop was 0.32v , which gives me a calculated current of 237a ( 0.32v/ 1.3m ohm)

seems reasonable so far, i was expecting something >100a….

now i also made my shunt with 6 lengths of thick gague wire, firstly to minimise resistance and also so i could use my clamp meter on each piece and add up the current; 32a + 43a + 67a +50a + >100a + >100a = 392a !!
(>100 since my clamp meter maxes out at 100)


so im a bit puzzled now:
- current clamp readings much higher than calculated current. > 390 vs 230 amps???
I could only calibrate using dc supply since thats all i had… some sort of rms factor i needed to apply?

- the individual readings of each wire in the shunt were surprisingly different at 32 to >100. Appreciate they might be slightly different lenghts amd better or worse connected, but still surprised to see such a difference.

being able to measure the current fairly accurately is important for the application i want to use this for in welding (spot/tig welder…), so id appreciate any thoughts/feedback on this.
 

Attachments

dendad

Joined Feb 20, 2016
4,494
Firstly, the way you have mechanically constructed the shunt it less than ideal. Clamping the wires together like that is not the best way to go. The sensing points should not be the connecting points as then you add the connection resistance to the reading. One thing you could try is to solder your sensing wires onto the main black cable a couple of feet apart, and not use the red at all.
It it was me, I'd use a current transformer, not a shunt. That would be a lot easier and it gives you isolation.

I want to increase the current further by adding a second transformer in series and add power control maybe via a dimmer ok the primary
Do you mean in parallel to increase the current?
 

Thread Starter

Gizer

Joined Aug 15, 2017
19
Ok, i take the point, i did think the connections were a bit awkward (amd caused the different amps on the diff cables), will try what you suggest. ( The seperate red cables did give me the clamp meter option thougj..)

for commecting the transformers, i believe circular currents could be a problem in parallel if they werent perfectly the same, hence series secondaries
 

Thread Starter

Gizer

Joined Aug 15, 2017
19
It it was me, I'd use a current transformer, not a shunt. That would be a lot easier and it gives you isolation.


Do you mean in parallel to increase the current?
Thanks for the tip ok current transformer, didnt know what that was but looks useful. I was trying to do this on the cheap but these are reasonably priced, certainly cheaper than a high current shunt..
 

Ian0

Joined Aug 7, 2020
10,006
Why don't you cut 6 identical pieces of cable, put a ring terminal on the end of each, bolt them all together in parallel?
Then measure the current down one of them and multiply by 6.
 

Thread Starter

Gizer

Joined Aug 15, 2017
19
Why don't you cut 6 identical pieces of cable, put a ring terminal on the end of each, bolt them all together in parallel?
Then measure the current down one of them and multiply by 6.
I have similar now but have to measure 6times since they’re different ( I just used pieces of cable i already had available). Since some max out the clamp meter i may indeed try your suggestion.. thanks

Who’d have thought a seemingly simple job as measuring current could be so tricky!!
 

Thread Starter

Gizer

Joined Aug 15, 2017
19
Why don't you cut 6 identical pieces of cable, put a ring terminal on the end of each, bolt them all together in parallel?
Then measure the current down one of them and multiply by 6.
The other issue i have is that this works for the current i have now but if i manage to increase/double it, i would need 12 strands of cable? Maybe time to buy the right gear! (Transformer/shunt or high current clamp!)
 

Ian0

Joined Aug 7, 2020
10,006
I have similar now but have to measure 6times since they’re different ( I just used pieces of cable i already had available). Since some max out the clamp meter i may indeed try your suggestion.. thanks

Who’d have thought a seemingly simple job as measuring current could be so tricky!!
Measuring high currents isn't easy.
Are you measuring AC or DC? If it's AC, then make or buy a current transformer. If you have a small toroidal transformer that would be a start.
 

Thread Starter

Gizer

Joined Aug 15, 2017
19
Measuring high currents isn't easy.
Are you measuring AC or DC? If it's AC, then make or buy a current transformer. If you have a small toroidal transformer that would be a start.
Ah ok, i didn’t appreciate that… its AC im measuring, (for now at least. Depending on success, i may also look to rectify to DC). I do have some transformers/coils lying about, so i will look into how to make a current transformer. Hadnt thought of making one, I understand the principle but surely the calibration would be just as hard as I'm finding with a simple shunt?!
 

Ian0

Joined Aug 7, 2020
10,006
Ah ok, i didn’t appreciate that… its AC im measuring, (for now at least. Depending on success, i may also look to rectify to DC). I do have some transformers/coils lying about, so i will look into how to make a current transformer. Hadnt thought of making one, I understand the principle but surely the calibration would be just as hard as I'm finding with a simple shunt?!
Maybe, maybe not.
If you can count the number of secondary turns that would be really useful. On low-voltage transformers you often can.
Then you can work out the current ratio, which is the turns ratio. Choose a burden resistor appropriate for the current.
It's not perfect, because the magnetising inductance is in parallel with the burden resistor; but you can calibrate it at a lower current.
 

Thread Starter

Gizer

Joined Aug 15, 2017
19
Maybe, maybe not.
If you can count the number of secondary turns that would be really useful. On low-voltage transformers you often can.
Then you can work out the current ratio, which is the turns ratio. Choose a burden resistor appropriate for the current.
It's not perfect, because the magnetising inductance is in parallel with the burden resistor; but you can calibrate it at a lower current.
i’ve made a current transformer of 50 turns on an iron core and connected it to my multimeter.
Id expect the current in the transformer to follow;
Ip/Is = Ns/Np
So i put 100a thru the cable (adjusted variac to hit 100a, measure with the clamp meter). I would therefore expect 2a read by the mulitmeter, instead i got 5mA!!

might it be the width of the transformer? Too wide a diameter/ theres a big gap between primary-secondary windongs/ poor efficiency??

i didnt commect a burden resistor but i think the ammeter would have one?
 

Attachments

Ian0

Joined Aug 7, 2020
10,006
That looks like an iron powder core. You need something with a lot more permeability. Ferrite would be better, but for 50Hz good old silicon-iron laminations are better still.
 

Thread Starter

Gizer

Joined Aug 15, 2017
19
Thought an iron/steel core would work, didnt know about eddy currents etc…
Will look for a proper core maybe in some old electric scrap that i have or try to buy one.

On the other hand, ive been messing about with current shunts and now current transformers for a while, whilst my real project is not progressing! i may just go out and buy one or a 1000a clamp meter!

all good, interesting stuff and learning a lot along the way, thanks for your help!:cool:
 

Thread Starter

Gizer

Joined Aug 15, 2017
19
I used a small transformer core to make this current transformer with a couple of turns.
i tested it at different primary winding currents and measured the secondary ac current:
10a -> 0.7a, 50a -> 1.5a, 100a ->1.8a

as you can see the relationship isnt linear so i wont be able to extrapolate for higher currents…

out of interest i also tested the other transformer i made with a solid steel core, it actually seems better ( but still not reliable to extrapolate…

10a-> 0.04a, 50a ->0.25a, 100a->0.88a

Neither follow anything like the Ip/Is = Ns/Np rule so no idea what im missing there :rolleyes:
 

Attachments

Top