Hi, is this the right place to discuss troubleshooting with HiTechnic sensors?
Thanks.

Sure, fire away!

= Xander

I recently bought a bunch of sensors, im new to all this but loving it. Anyway, I'm having problems with a HiTechnic Compass. Luckily I have 2 so I've been able to test one against another.

I've run the calibrate application from the drivers download and one sensor seems ok after about 3 attempts, one of them just doesn't want to play, it just reads between 220-350 in complete 360 degrees rotation, after the calibration program completes and says sucessful.

(I have to ammend that, i've just tried again and this time it seems to have calibrated ok) Hmm.

This brings me on to the next question, interference.

I'm testing with a really simple program that turns a basic 3 wheeled bot faster the further away from north it starts, right down to snail pace the closer it gets, for accuracy, with a 2 degree window to stop at. Some simple maths would allow me to choose any degree to stop at, blah blah. However, at one end of the room I'm testing in, North appears to be 20 degrees further west that at the other, and in the middle of the room the bot gets so confused it can't find north at all. There is nothing within 4 feet of the bot in any of the locations, there are a couple of PC's around and the normal stuff you find in a room. The only thing I can't check is what's under the floor, but it's concrete anyway. So I'm left wodnering how sensitive these are to surrounding conditions, and therefore how trustworthy they are.

Any info or things to try/test would be gratefully received.
Cheers.

Ah, concrete you say? You have no idea how much re-bars can mess with your compass! Compasses are very trustworthy when it comes to picking up magnetic fields, but they don't discriminate between man made ones and Earth's magnetic field.

Is the compass mounted on a high "pole", to be as far from the motors and NXT as possible?

= Xander

Ah, I didn't think of the re-bar in the crete! It is a newish build block of flats as well, so there will be loads of the stuff.

Yes, I have the sensor mounted on a frame over 20cm away from the brick and motors.

Having just done a bit more testing I've found it is absolutely reliable on one table in the corner of the room! I've now re-calibrated both sonsors in this location and they both perform great.

Thank you for knocking some sense into this old head of mine!
Cheers.

Glad it is working for you!

= Xander

Just want to confirm here. If I only care about repeatability, calibration is not necessary, is it? When considering the sensor is mounted 6inches away from the brick the motors.

Calibrate once for your specific robot, but certainly not every time, no

= Xander