Is it possible to increase the resolution of the HM55B compass module?

EWo_Ge

Hello all -
I want to start some experiments, to use the SX in RC modell avionics.
One of the possible sensors, I'm planing to use, is the HM55 compass module. Unfortunatly the resolution of 128 inrements is not fine enough for my project.
Is there a way to increase it to·minimum 1 deg. or less????
Or has anybody any information about other compass chips, which could do that??? (I remember playing around with a chip from Phillips, I beleve, which was analog but cant find it anywhwere anymore)
I would not need an absolute value about north/south, just about the change of direction. (an accellerometer would NOT do the job, since the changes may happen very, very·slow or very fast...)

Thanks, for any opinion,
EWo

JonnyMac

The compass returns 11-bit values for each axis; with some trig you should be able to work this to a 1-degree value.

CCraig

You could use one of the small helicopter gyros. It depends on how long 'very, very slow' is. I hear they do drift over time. (at least the $30 ones that I'm looking at do). I'm thinking I'll be able to re-sync and continue on.

Note: one degree resolution IS NOT in my design spec.

HTH, Chris

EWo_Ge

Funny thing, the gyro is the part I want to replace. I want to use·the SX to realise some extra functions for a heli. Since a good gyro is pretty expensive and still has some drift, I want to give it a shot and integrate the compass module instead.

EWo

Ben Spratling

There's lots to learn.

First, even though the HM55B does give 11 bits, on the surface of the Earth you'll only get 6 bits of useful information. (the Earth's field is 45-65 µT, and the unit measures in µT.)

Second, the Earth's magnetic field is not parallel to the ground, so when you measure the field, parts of it go into the ground. If you only have a 2-axis compass you'll miss the third component, and if you bank on turns this can throw off your direction.
I do know of a 3-axis magnetometer from Honeywell that gets 1000 times the resolution of the HM55, but it costs $200, and it's not available from Parallax (sad day).

If you want true heading, instead of just magnetic heading, then you'll also either want a GPS unit that can give it to you live, or look up the value on a plot on wikipedia or something if you're only going to be using it in a localized area.

I'm currently working on a device that will use GPS, a 3-axis accelerometer and two HM55's to tell which direction it's facing. (The accelerometer is so it knows which way is down, so it can ignore that component of the magnetic field.)

The web link on page 4 of the "Hitachi HM55B Compas Documentation" is great if this is your first technical experience with the Earth's magnetic field.

Phil Pilgrim (PhiPi)

I did a study awhile ago (http://forums.parallax.com/showthread.php?p=614618) to see if a two-axis compass reading could be corrected using a two-axis accelerometer like the MX2125. I came to the conclusion that it could — at least with moderate tilts in an inertial frame of reference. But non-inertial motion, such as turning, will render the derived corrections useless.

-Phil

Ben Spratling

And if you use a 3-axis gyro, and 2 more 3-axis accelerometers arranged non-colinear with the first, combined with an acceleration estimating algorithm based on absolute position measurements, then you might actually get the info during turns as well, as long as you take a little care to eliminate the magnetic contributions from any major power lines by using twisted pairs and not placing your magnetometer on a circuit board with a ground or power plane.

Although differentiating absolute position from a GPS at 1 second intervals can result in some pretty jumpy accelerations!