All the common magnet materials have a temperature coefficient much greater than your spec.

Do you have a link to the Thermoflux Sheets. Google didin't turn up anything that made sense to me.

One technique to reduce the effect of the magnet changing with temperature would be to saturate your circuit. So that as the magnet dropped off the field would change less.

Is that what the thermoflux sheets do?

Mike

Basically flux density of the material decreases with increase in temperature. Heres the link to the Thermoflux sheets.

http://www.vacuumschmelze.de/dynamic/en/home/products/materialsampparts/softmagnetic/nickeliron/thermoflux.php

The technique is called shunt temperature compensation if i'm not wrong.

You mentioned about saturating the magnet. Will it have any effect on its stability?

I'm looking at a temperature change from 0 degC - 40 degC.

Another company Electron Energy Corporation boasts of temperature compensated magnets with a coeff of <0.001%
Any idea about this?

http://www.electronenergy.com/products/samarium_cobalt_magnetic_properties.htm

Look at EEC 1:5TC-9, EEC 2:17TC-18, EEC 2:17TC-16 & EEC 2:17TC-15 (TC is for Temperature Compensated Magnets).

Thermoflux is a soft magnetic material not a PM.

Not saturating the magnet but saturating the steel circuit. If the steel is saturated then as temperature effects the PM the resultant field changed is muted. In other words the steel carries only so much flux. And changes in the generated flux do not change the result as much if initally saturated.

ECC stuff as always looks interesting.  Perhaps they run it through some temperature cycles prior to it's use. Never noticed it before.

Your tolerance run well beyond what I normally have to worry about.

Mike

Well I dint understand how do you saturate the steel? I mean when I'm magnetizing the PM it is completely assembled with the steel. So how do I know the steel is saturated?

The magnet i'm working on is used in very high precision weighing balances with accuracy of 0.1mg and less. In this case even the slightest change in the temperature affects the stability of the scale. So we're trying to make it as stable as possible...

Your steel saturates from the flux generated by the permanent magnet. I believe you can see that in the FEMM solution you posted. Although I don't know the scale value of the colors shown.

A curve for 1018 steel will begin to show saturation at maybe 16000 gauss and start getting serious at 18000 gauss.

I describe saturation as like what happens to a sponge as one tries to pick up water with it. Once steel becomes saturated it begins to behave much like air.  

Mike

Most of this is done by shunting the part of the field with a magnetically soft material that has an negative temp coef.  In this way as the temp rises you get less shunting, and your field stays more stable.
I don't think that even the radar guys are working to the levels that you are.  I suspect that you will need an active electromagnetic circuit to both measure and trim the field in order to meet your requirement.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube