I've been experimenting with tubular neodymium magnets for use in DIY air-core movements. I use a piece of 0.033" stainless steel wire for the rotor shaft and epoxy the magnet onto it. The movement body is made from 0.25" OD brass hobby tubing (K&S Metals). I use a short length of 0.0625" OD brass tubing as the rotor shaft bushing. I reamed the bushing to size using a 0.0335" (0.85 mm) drill bit. The bushing is held in place with yet more epoxy. I made a simple alignment tool to hold the bushing centered as the epoxy cured.

Here's a picture of an experimental air-core that's ready for the field windings.



There are more pictures and information on my site.

You've updated your site since I last saw it, Mike - looking good! Your instrument book is still on my list of things to buy...

I have to admit the only thing that is putting me off building my own instruments is the programming. I think I can handle the engineering side of things, but as soon as I get to the code bit, I get cold feet.

The site was originally built using FrontPage and was increasingly non-compliant with current day web standards. Finally had to bite the bullet and drag it into the present. Thanks for checking it out.



Software can be a frustrating handful. At least the compilers and IDEs are free.

A brief update:

I got the field windings added to two DIY air-core movements. The larger has a 10mm long tubular magnet for the rotor. The smaller unit has a 5mm long magnet. Both work well enought to swing a pointer. The larger unit has a bit more torque, but the smaller one has less weight on the rotor shaft so should see less friction. So, one might be a bit faster while the other might be a bit more accurate. I'm not sure if either difference is big enough to make a difference.

What I wanna know is; how much did you end up paying the neighbor kid to spool all that wire on there for you.

This begs for some kind of little gadget to wind the bobbins without losing ones mind in the process.

g.

I actually have a winder made specifically for winding air-core movements. Sadly, it's for an earlier design and I've yet to make an adapter. Fortunately, adding the windings is not TOO mind numbing, or maybe I'm already so far gone I don't notice.

And you're right about the "kids" here, they do want to be paid, even my own daughter.

Hehehehe. Didn't you know that "Dad" sounds like "ATM" when it's pronounced properly?

A winder & glue-up jig would make a great cnc project...

g.

Another update: It works.



It responds rapidly to 5 volts applied to a winding and shows no indication that the neodymium rotor magnet is interacting with the shield. I had some concern that it might due to the strength of the magnet and the fact that I used mild steel (galvanized electrical conduit) for the shield. The one thing I'm not happy with is the design of the wire guides that do a poor job of holding the field windings in place. There's an easy fix (I think), but I'll have to build another prototype to be sure.

Do you think it would be possible to drive these with two PWM signals? I think it could work if a virtual ground was used somewhere halfway between logic 0 and logic 1 voltages, but would the AC current negatively affect the magnet? Otherwise this would really reduce the amount of components necessary to drive the instrument.

In fact, I'm planning to drive the movement with a pair of PWM signals. It won't hurt the magnet.

The circuit is pretty simple. There is a DS3695 chip used as a RS-485 data receiver that connects to a PIC16F648A micro controller. The PIC controls an L293DD H-bridge motor driver. The micro controller generates the two PWM signals in firmware.

I'm implementing the circuitry (mostly) using surface mount components. The small components allowed me to put just about everything on a single 2.4" by 2.4" circuit board. You can see a picture of an early version of the board artwork in the "DIY Air-Core Gauge Electronics" thread.

The prototype circuit boards were supposed to be delivered last Friday, but UPS neglected to put them on the delivery truck. With luck they will be delivered today. I'll post pictures on the other thread.

I'm making another prototype. I'm not happy with the way the field windings turned out. The second winding tends to slip off. I revised the wire guides, and have lengthened the shield to better protect the windings.

I remember a design for a two phase micro brushless motor, using the magnet from a coreless pager motor (which might be simpler to come by and should have a suitable shaft already fitted). Essentially it had two cross shaped plates at either end with four "pillars" linking them at the corners. The windings went between these pretty much the same way as in your first prototype. I did try to find the pictures, but it's most likely buried deep in the archives of rcgroups. The design has long been superseded by the single phase hall controlled motor (essentially an actuator with no rotation limits), and is mostly used for sub-gram planes. There's also brushless conversions of coreless pager motors that use the original windings, but these are three phase motors, so not really suitable to work as an instrument.

edit: Oops, I just realized I made a mistake. Coreless motors have a static magnet, so the shaft won't really be fitted to the magnet. Sorry about that. OTOH, the magnet will be a nice cylinder polarized transversally, so it ought to be still good, but the shaft will have to be fitted to it somehow.

edit2: this thread in particular: http://www.rcgroups.com/forums/showthread.php?t=596219&pp=100&page=12 might be interesting for building and winding techniques.

Wow! That's some impressive miniture work.

Got the last prototype together.




Putting 5 volts on a field winding causes the rotor to respond nicely. The next step is to get the electronics working. I've more or less completed the hardware. I'm working on revising the firmware. I started with the firmware listing from an earlier air-core gauge project which uses a PIC16F628 operating at 12 MHz. The current hardware includes a PIC16F648A operating at 20 MHz. A few tweaks are in order. I also have to re-learn how to use the Microchip IDE... It's been awhile.

I have the electronics working well enough to spin the air-core shaft. Since the shaft is indeed turning smoothly, I consider the air-core movement functional and will shift my efforts to further testing of the electronics.