Fun with HALL EFFECT RATIOMETRC device

[kinarfi]

I was playing around with a Honeywell SS490 Series Standard Miniature Ratiometric Linear Hall-Effect Sensor, SS495A, and took a small toroid core and cut a small section out that was just barely big enough for the hall device to fit into. Glued it in place and let the glue dry. I powered it up with 10.00V and started running current through the wire I had run through the toroid and measuring the output of the Hall effect device. Pretty cool, with the set up I made, I had a center voltage of 5.02 and the voltage would change .03 volts for every amp it ran through the wire. I tested it up to 15 amps and I think the more accurate value may be .028 V/amp

 

[MrAl]

Hi,

Yes those things are pretty cool arent they?

The way to get more accuracy is to build the core construction up as a flux balance.

The way a balance scale works is you put weight on one side of the scale and then place calibrated weights on the other side of the scale until the scale balances out. The way a flux balance works is you create flux with one winding with current in one direction and then try to cancel the flux with another adjustable current in a second winding in the other direction. If the flux is not balanced, the hall device shows an offset from the zero point, but if the flux is balanced then the hall effect shows zero output.
The measurement winding is usually one turn, and the balance winding several turns. The balance winding has more turns so that a small current in the balance winding can cancel any flux developed from the measurement winding. This allows using small known currents to be used to measure large currents.

Now this can be done as simply as using a pot to adjust the the small current, and measuring the current with a small multimeter you can then set it to balance the flux (getting a zero hall effect reading). Then you can calculate the current through the measurement lead by using the turns ratio.

The way to do it automatically is to use an op amp to measure the output of the hall device and use that as feedback to adjust the output of the op amp (possibly with a higher current driver added). The op amp then automatically balances the flux and you can then measure the small current through the balance winding. If you simply leave the meter in the circuit, after the op amp adjusts for zero flux the meter will indicate the current that flows through the measurement lead divided by the turns ratio. So if the meter reads 0.1 amp for example and the turns ratio is 10 (10 turns for the balance winding and 1 turn for the measurement winding) then the true current being measured is 0.1 times 10 which is 1 amp.

The reason why this is more accurate is because the zero point output of the hall device is the hall device target operating point rather than having to depend on the linearity of the hall device itself. As long as the zero point is repeatable and does not drift this works very well.

To follow fast changing measurement currents however requires a high compliance core driver, so it's easier to do for DC current or somewhat lower frequency AC.