DIY Magnetic Stirrer

This is true for someone building a single stirrer for the purpose of getting a functional stirrer (cheap). However, given the advantages of no moving parts, the switched pole solution might be cheaper for mass production.

If its an excuse to tinker around with electronics, then it will need a complete set of H-bridge drivers, a microcontroller and some other expensive parts or its just not worth discussing in this n.g. ;-)

Would the coils be easily removed? I never taken apart a stepper motor..

D from BC

Paul Hovnanian P.E. snipped-for-privacy@seanet.com posted to sci.electronics.design:

I think that if i were trying this i would custom wind an about 12 pole stator and use a RC modeling motor (brushless AC) controller. Reasonably high power capability and easy control voltage.

I can feel the pull in that direction :)

D from BC

Folks, many moons ago there was an article on the very subject using two coils in quadrature and a quite rough 0-90º sine generator from a clock plus dividers plus a resistor-based D-A converter: Magnetic stirrer with no external moving parts P.S. Engel, Rev. Sci. Instrum. 49 (10), 1495-1496 (1978). The voltage fed to the two coils in quadrature (220 turns, #22, on 7cm form, 4.9 mH) was +/- 12 volts. This is probably the only way to move large stirring bars (8-10 cm long) on big tanks. If you go the normal route, a magnet on top of a motor, it gets very hard to find large magnets and very complicated to keep them well balanced and with enough coupling for dragging the stirring bar. Hope it helps. jon

Ahh..the ol fake a sine wave trick.. :)

And driving 2 coils only.. mmmmm

I'm trying to stay away from coil construction.. I'd rather hack up some existing consumer electronics or hack a transformer and then whip up some crude switchmode timing electronics.

But I will consider other methods if I determine if a "from scratch" idea is quicker than hacking up some junk.

Thanks

D from BC

D from BC snipped-for-privacy@comic.com posted to sci.electronics.design:

I'll try....my magnetic's knowledge is iffy..

Some Hammond xformers are on [] ..O core ??...that be hacked off at the end to make a C core. That or, slice some toroids in half.

Then perhaps do an electromagnet arrangement like:

(3) S N NbarS *Top View* S S N N (1) (2)

Then try to get the bar magnet spin with a primitive 1,2,3 switching of each C coil. Ex: To get the permanent magnet N to follow the electromagnet S from 1 then 2..then 3.

Hopefully the bar magnet doesn't zoom away from the center.

D from BC

D from BC snipped-for-privacy@comic.com posted to sci.electronics.design:

The electromagnet relative polarity is dependant on the direction of the current in the coils. So if you have a consistent winding orientation the start and finish become more important. Then use waveforms more like this:

------ ------ ------- + A: --- --- --- --- --- --- --- 0 ------ ------ ------ -

------ ------ ------ + B: --- --- --- --- --- --- 0 ------ ------ ------- --- -

--- ------ ------ ------ + C: --- --- --- --- --- --- 0 ------ ------ ------ -

Preferably active at all levels but open for 0 will still work.

mmmm... bipolar drive.. Field reverses. Heyyy...are those stepper motor waveforms?? If so, I guess I can just browse the stepper motor IC's. Maybe something with built in H bridge outputs.

At first I thought unipolar drive. No field reversal. I was thinking it might simplify the electronics. But...I'll guess there's something wrong with that... low bar torque or poor/no function.

D from BC

Had to use google, my ISP is misconfigured currently.

The object is a rotating magnetic field. You could run it in a unipolar way, just shift the waveforms up. Three steps is not actually neccessary, there are three phase hobby motor controllers (like used for RC cars, planes, boats, and heli's). Not too expensive and would do the job nicely. You may have to wind a lot more poles to get the speeds you want though.