Hall Sensor with Range > 1000mT

th

n r a

I have access to eqipment to calibrate the coil, at least for a permanent f ield from a magnet

We have made measurements with the inverter turned off, letting the motor c oast (free run), but then the effect of the current in the winding does not effect the field, and then we need to model that someway. I am toying with reducing torque ripple due to non-sinusoidal winding layout

Cheers

Klaus

Nice, that is a really high field, like that of CT scanners :-)

Thanks

Klaus

Nice part, I like that the field is "unlimited"...

My budget is not particular limited, just that I do not destroy a 1000 dollar instrument

I need to fit inside a 5mm slot, so that puts a limit on the sensor size

We are measuring in the slot, and that field will be less than the field in the stator tooth, but we are more concerned about the waveform shape and not the absolute value

Cheers

Klaus

It's for measuring the Flux inside a slot of a permanent magnet synchronous motor to evaluate the BEMF waveform during operation. We have added a coil on top of the standard coil, but the coupling between the standard coil is high so instead of monitoring the BEMF, we instead just measure more or less the applied voltage from the inverter drive

That's the reason for using a Hall sensor, that will measure only the magnetic field and won't care much about common mode noise either

Cheers

Klaus ========================================================================================

What is your volume going to be? One or two for product characterization, or lots and lots of shipped motors? If it is not too many, buy a cheap gaussmeter like this one, picked at random on ebay:

Buy spare probes, I'm sure you can get a nice volume discount, and use the meter to check your circuit calibration. Not the most elegant but it gets the job done without lots more searching. If you need lots and lots see if the distributor will sell you just the element without the housing and cabling and connector.

If you are building them permanently into a motor then the suggestion to just rotate a more sensitive sensor to get the field on scale will work since you can control and lock in the geometry. Any kind of field shaping pole pieces at 1 T in air you have to start worrying about saturation and reproducibility in the pole pieces, plus the extra volume of the pole pieces, plus the potential field perturbations in your motor, so while that could work in theory I think it will take a lot of development time and effort.

----- Regards, Carl Ijames

If the field is periodic, you could phase-lock to the rotor position or drive input, and use phase-sensitive detection to grab multiple harmonics. Your seek coil would be a few turns of copper wire, mounted 35.26 degrees skew to the probe axis. Rotate this on axis to three positions, 120 degrees apart, to get the three orthogonal components of the local field.

AlwaysWrong strikes again.

I think you should retire from the field before you get started.

Tesla, Weber, 10kG, etc...

Jamie

So, do you mean you are mainly measuring the common mode voltage, or the phase voltage? Flux = integrated voltage after removing the resistance term.

I can see now that flux weakening won't work. Good luck.

joe

P.S. I've learned that 'magnetic field' = H and 'magnetic induction' = B.

On Tue, 22 Sep 2015 20:54:10 -0400, M Philbrook Gave us:

No, you do not. That is your biggest problem.

After graduation I have been involved in the design of a pulsed 4T magnet for materials research. The combined thermal, electrical and mechanical design had a very strong emphasis on the mechanical forces that would occur during the ramp up, the plateaus of constant field during the thyristor controlled plateaus of a few milliseconds each, and during the dissipative decay of the induction. Writing a dedicated FEM program in C, that was fun. The thyristor rectifier was to be designed for some MW peak.

joe

I'm not up to date on the capabilities of FEM programs nowadays, so maybe it's a silly question to ask why you don't simulate it with FEM?

joe

I think 'T' stands for 'Tesla', or am I wrong?

joe

Sorry, but that's the A/m (H = the integral I over length along a closed curve). T is the unit of magnetic induction, and Henry of inductance.

joe

? kG?

joe

Sample! :)

joe

Which field in the slot, radial, tangential due to non-homogeneous wire distribution, and there is leakage also. If you put the sensor too deep in the slot you'll get also kicked by the tangential leakage flux, etc. etc. Seems a difficult exercise to me.

joe

+1!

joe

You are not wrong.

He is wrong (imagine my surprise...)

--

John Devereux