What sensors do I need for altitude+azimuth measurement ?

Hi guys :slight_smile:
I want to build a small device which will be placed on a telescope tube, and which will show the alt/az the scope is pointing to. I googled a bit and found out the best option would be to use optical encoders. This seems however quite complicated, and since I'm a beginner I'll try it with sensors(at first).

I was thinking at two methods of measuring:

  • Absolute measurement, using the magnetic north to determinate the azimuth
  • Pointing the scope to a star with known alt/az,"calibrating", and doing relative measurement afterwards

Do you believe any of these two methods would work ? I guess I need a magnetometer for the first option, and a gyroscope / accelerometer for the second method ? Or maybe combine somehow magnetometer + acc/gyr for better accuracy. I've never worked with such sensors before so hopefully I'm not talking total nonsense.
Any help is greatly appreciated :smiley:

cpper:
I googled a bit and found out the best option would be to use optical encoders.

Probably true. What makes you think your methods will be simpler?

I suppose that wiring the encoders on the base of the telescope would be difficult. This would also make the base more impractical and difficult to carry.
This is the scope:

You can use a magnetometer and accelerometer to measure altitude and azimuth, and you can find many tutorials on line for how to do that for yaw and tilt (or pitch and roll) angles. The same principles apply for both but the angle definitions are slightly different. Using this method requires that the telescope be leveled and initially aligned with true North (with a correction for your magnetic declination).

The catch is that this is not a beginner project and you will need to learn how to calibrate the magnetometer and possibly also the accelerometer; required for reasonably accurate results.

You can avoid most of that hassle by buying the BNO055 absolute orientation sensor. That sensor will also return yaw, pitch and roll, which must be transformed into alt/az.

Either way, with careful calibration, you can expect an absolute accuracy of perhaps +/- 2 degrees in each angle. Is that good enough for your purposes? If not, encoders are much more accurate.

cpper:
I suppose that wiring the encoders on the base of the telescope would be difficult. This would also make the base more impractical and difficult to carry.

I can't see what is impractical or "difficult to carry". You have fixed components and rotating components just like any other telescope, and one pair of which is of encouragingly large diameter. The encoders would be compact enough, with one practically invisible, the electronics could be fixed near that handle, and a readout on your phone. You can't tell me you are the first person to want to do this, and I bet there are dozens of examples in Sky and Telescope etc.

Altitude is easy. A cheapo ADXL345 accelerometer module can do this easily, to relatively high accuracy.

Azimuth is harder. A magnetic compass module is subject to all sorts of interference from nearby magnetic objects. The encoder plus a zeroing procedure is going to be the best option.

Nick_Pyner:
I can't see what is impractical or "difficult to carry". You have fixed components and rotating components just like any other telescope, and one pair of which is of encouragingly large diameter. The encoders would be compact enough, with one practically invisible, the electronics could be fixed near that handle, and a readout on your phone. You can't tell me you are the first person to want to do this, and I bet there are dozens of examples in Sky and Telescope etc.

I was thinking the electronics would make the base more delicate when carrying. In order to go out in the garden with my scope I have to go trough some tight spaces and make sure I don't hit anything.
I'll look more into encoders and think where I could place the, maybe I'll give it a shot.

MorganS:
Altitude is easy. A cheapo ADXL345 accelerometer module can do this easily, to relatively high accuracy.

Azimuth is harder. A magnetic compass module is subject to all sorts of interference from nearby magnetic objects. The encoder plus a zeroing procedure is going to be the best option.

Would I be able to use the accelerometer with the same relative high accuracy to measure the azimuth difference relative to a starting point ?

No. The accelerometer is measuring the gravity vector. It can't measure rotation in the vertical axis.

I dunno... a rotary encoder seems like a no brainer for this. And it wouldn't add much mass or bulk to the scope at al. Plus it would be easy to make. It's kinda expensive, but this encoder seems ideal. Put a rubber roller wheel on it, and mount it so the wheel rolls on the round glide wheels of your dob. Put one on the alt, and the other on the az wheels. That way the count from each one of them will be a function of how much you have moved the scope in that axis. I would expect sub degree precision.