Looking for a wheel speed sensor...

Ed Lansinger elansi01 at mpg.gmpt.gmeds.com
Mon Oct 23 13:48:55 GMT 1995


Chris Howard wrote:

>I'm currently building a data logger for a thesis project and I need a wheel
>speed sensor.

Adding to many good suggestions, I thought I'd contribute the sensor I worked
up for a Formula SAE car.

Get a Hall-effect sensor from Allegro #UGS3140 (TTL-compatible open-
collector output, three pins).  Get the Radio Shack rare-earth magnets
#64-1895 (very small, very powerful).  Get a couple packs of the magnets
because some of them are duds.  Epoxy (good epoxy, that is, on a well-
prepared surface) will hold the magnets in place on a 6" diameter wheel up
to 11,000 RPM or so as long as it doesn't get hot.  If heat will be a problem,
a simple and effective way to hold them in is to drill a hole the diameter
of the magnet (snug friction fit), then take a sharp metal punch and
punch 4-8 divots in the metal around the perimeter of the hole, close
enough so that each divot creates a small bump that pushes in on the
magnet.  This holds the magnet at 11,000+ RPM on a 6" diameter
wheel even if it gets hot (which, according to my calculations, is
over 10,000 g's acceleration!).

The UGS3140 takes +5 and ground and pulls the output line low when
it sees a north- (or south-, depending on orientation) pole.  Tie
the output line to +5 with a 10K resistor and you are ready to read
it with your system.  I made a small PCB containing the UGS3140,
resistor, and a modular phone jack which I then screwed onto mounting
holes I had drilled in the upright.

This will work with air gaps up to about 1.5mm and could probably be
tweaked to more.

If you mount more than one magnet and you are looking for precise
velocity measurement magnet-to-magnet, you will have to mount the
magnets (or drill the holes) using a drill press and an indexing table.
Careful measuring by hand will not be accurate enough.

Don't mount magnets in, on, or too close to a brake rotor.  The rotor
is likely to get hotter than the Curie point of the magnetic material,
causing it to lose its magnetism.



Ed Lansinger
General Motors Powertrain
Premium V (Northstar/Aurora) Software & Calibration Group
Rensselaer Formula Lightning Project




More information about the Diy_efi mailing list