Measuring high DC amperage / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Measuring high DC amperage

1 2 next

Move Thread

LAN_403

Storken

Fri Jul 20 2012, 08:23PM

Registered Member #1960 Joined: Thu Feb 05 2009, 11:57PM
Location:
Posts: 48

Hello, after hitting a roof with my allegro ACS756 current sensors (200A bidirectional) - I'm looking to wander into the world of gapped toroids and a hall sensor mounted here.

The application is motor controll, and sampling current bandwidth should probably be in the 50kHz range. Allegro has some SIP packages that can be programmed and what not for this use. But I want some pointers as to how to approach this :)

I want to measure all the way up to 1000A, while still keeping low end current accurate. The wire size I plan to use is 0- 00, about 50-70mm^2.

Any tips would be greatly appreciated ;)

HV Enthusiast

Fri Jul 20 2012, 11:51PM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

Hmmm . . . you say you want to measure DC current but need a bandwidth of 50kHz? Something doesn't sound right.

Anyways, these are really nice current sensors and are inexpensive. I think the 1000A version is only about $80 USD.
Measures DC, Pulsed, and AC and good for 25kHz bandwidth so that should be good for your DC measurements.

Storken

Sat Jul 21 2012, 01:57PM

Registered Member #1960 Joined: Thu Feb 05 2009, 11:57PM
Location:
Posts: 48

80 dollars is waay out of the price range I was picturing. The SIP hall effect packages go for 1-2 dollars, and only need a toroid in addition to this to work.

I'm under the impression normal transformer current sensors won't work in a motor application. Almost like in a buck converter, if you're controlling current - you need a current shunt with opamps or a hall-effect sensor.

The 200A bidirectional allegro sensors are 5 dollars each, and they fullfill my needs up to this level - now i only need a more powerfull sensor.

Steve Conner

Sat Jul 21 2012, 04:30PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

If you want to build a 1000A motor controller, it'll cost a lot more than $80, so you might as well spend $80 on the right part. LEM may have a cheaper model, though.

Hall effect CTs are a feedback system: the Hall module drives an amplifier that drives a coil wound on the core to null out the field in the air gap. The high frequency response comes from the coil directly, the DC component of the output is supplied by the servo loop.

Storken

Sat Jul 21 2012, 05:17PM

Registered Member #1960 Joined: Thu Feb 05 2009, 11:57PM
Location:
Posts: 48

BOM of a 80v110A (three phase current rise goes up to 200A peak) controller is about 100 dollars, I've got working prototypes of this controller.

BOM of a 80v700A controller is about 400-500 dollars (1000A peak phase currents). I will not spend much more than 10 dollars on each of the current sensors. I do not need a finished product, I can DIY it - and I plan to.

HV Enthusiast

Sat Jul 21 2012, 07:21PM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

Are you trying to measure real DC current, or pulsed current?

Storken

Sat Jul 21 2012, 09:02PM

Registered Member #1960 Joined: Thu Feb 05 2009, 11:57PM
Location:
Posts: 48

I guess pulsed current once the motor is up to speed. The current is sine shaped once the motor crosses a threshold (foc control). At startup it uses a simple v/f routine with a squarewave current. My guess is this is defined as pulsed current.

A gapped toroid, like this
20Normal

with a hall sensor mounted in the slot was what I was picturing. Allegro's hall effect sensors can be programmed, but I'm not sure if they're the best there is. They are cheap, at 1-2 dollars a pop.

Field strength depends on distance from conductor and amps flowing, here is a hyperphysics link:

If I know the minimum or median in the toroid, is this representative ?

Some form of shielding the hall effect, toroid and conductor from outside magnetic fields - is this needed? Any DC bias would be very detrimental to controller performance.

Pinky's Brain

Sun Jul 22 2012, 03:32AM

Registered Member #2901 Joined: Thu Jun 03 2010, 01:25PM
Location:
Posts: 837

You can get 500 Amp shunts for around 20 bucks.

Mads Barnkob

Sun Jul 22 2012, 07:46AM

Registered Member #1403 Joined: Tue Mar 18 2008, 06:05PM
Location: Denmark, Odense C
Posts: 1968

Go to a local scrap yard, I have walked away with 1200A shunts bought at scrap metal price.

klugesmith

Sun Jul 22 2012, 02:49PM

Registered Member #2099 Joined: Wed Apr 29 2009, 12:22AM
Location: Los Altos, California
Posts: 1716

Storken wrote ...
A gapped toroid, like this
20Normal

with a hall sensor mounted in the slot was what I was picturing.
...
Field strength depends on distance from conductor and amps flowing, here is a hyperphysics link:

Your hyperphysics link is a good starting point. It's more complicated in circuits that include both magnetically permeable material and air gaps.

For open-loop (no pun intended) response to be linear, you want almost all the reluctance to be in the air gap. With 1000 ampere-turns across, say, a 1 mm gap, the H field in gap would be 1e6 A/m, so B in whole loop would be 1.26 T. But that's much more than ferrite can carry, so the model's no good.

If you need good linearity up to 1000 A, IMHO you need cores of a size where gap can be around 10mm. Or as Steve said, drive a compensating winding to keep the DC flux density at zero. (a 1000 turn winding would need 1 ampere in your case).

It's a challenge to get high bandwidth from regular current shunts,
because of their inductance and very low resistance.
Consider a standard 1000 A : 50 mV shunt (which dissipates 50 W at rated I).
Even if its inductance were only 50 nH, the L/R time constant would be 1 ms
for corner frequency of 159 Hz. (sensitivity goes up above that point).
Maybe you can live with that, or design a passive compensation network.

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob