Magnetic stirrer but with a twist

Let’s say you have a chemical flask as a part of an apparatus in which a reaction is going on. You are adding reactants to the flask and want to mix them good, what options do you have to perform that mixing? If it was only a flask, you could shake it by hand, but the flask is already attached to other glassware and to a stand so this is not an option anymore. The next solution is to push a glass rod through the flask’s neck and move the rod around. If only the neck was not already used for other purposes. There are, of course, sophisticated adaptors and mechanical mixers for that job, but by far the most convenient way is to use a magnetic stirrer.

Magnetic stirrer is a piece of equipment that rotates a magnet underneath the flask, while inside it a smaller magnetic bar (in PTFE or glass enclosure) catches on with this rotation.

One of the important parameters of such a stirrer is how far can you place the driving piece from the stir bar, because it is likely that some other equipment (like a heating mantle) might be placed between them. If we are to design a magnetic stirrer (and as you’ve probably guessed, we are), then we are going to optimise configuration of the driving magnet’s field for this “farthest reach” criteria.

Ok, what if we just place a larger magnet like that:

Nah, a wider magnet produces most of the field around its rim and the field’s reach does not extend much. On other hand, if the magnet is longer, then the reach increases indeed:

Ok, that’s one solution — massive long neodymium magnet. On downsides: it creates a field below as well as above and that’s not something desirable. Let’s place some iron shielding (core) below the magnet and to achieve two distant prominent poles, split the magnet in two halves.

Now the problem is that the middle of the core provides a low impedance path for the pole’s field to sink. As a result we get a close-to-zero horizontal field component on the axis. How to solve this trouble? Remove the naughty core middle then!

Does it look familiar? Sure — that’s a classical horseshoe magnet shape. If we had a whole shape made of NdFeB, that would create a stronger field, but the reach would be quite similar. The nice feature of this design is that we may use readily-available flat magnets in combination with a custom-shaped steel part rather then require a whole custom-shape magnet. On the other hand, using a big solid ferrite magnet (better with a custom magnetization pattern) is also a viable option.

Ok, enought theory, let’s get hands dirty. Here is what will become the steel core with recess in the middle:

On the back side another bore is provided for a pair of bearings and a groove for the drive belt.

Now a bit of milling. The function of the side flats is to assist the center cavity in preventing the flux leakage.

Turned central shaft and bearings are here:

And now the module is assembled with 4 pieces of HDD magnets attached:

You may notice that something is off here: the core looks about twice wider than the magnets and the magnets are snapped in halves. This is an unintentional outcome indeed. The original idea was to use HDD magnets in their original shape (circular segments) and they would cover the core’s faces precisely. But for that the segments must have uniform magnetization in one direction and the HDD magnets are not created like that. Their magnetization switches direction going from from one half to the other:

That’s why I had to split them in halves. I did not bother redoing the core to optimize for the reduced magnet width but rather planned for refitting the device with new magnets that will be remagnetized properly. This intent has also started the (Re)Magnetizer project. But now back to our device.

It is indicative of this magnetic design how high is it possible to elevate the stir bar above the drive assembly while not losing the coupling.

And here how it looks assembled with enclosure (non-magnetic stainless ashtray) and installed on the lab stand:

In the background you may find an old factory-made heater/stirrer with a solid ferrite magnet inside. What I found is that the factory version provided about a twice lower locking range than my design. Also I find it very useful that the stirrer is slim and is easily mounted on the lab stand on a standard steel rod. This makes it super-easy to fit it after the rest of the apparatus has been assembled and adjust its height/position without additional means like a lab-jack. Here how it couples with a heating mantle:

And for the case of a more complex apparatus it is easy appreciate how a smaller device footprint may come handy:

And what is this nice fume hood on the photos? — Glad you’ve asked! That’s another project you can read about.