I have a camera tracking project where I need to pan a camera to a target bearing. I am basically tracking a person from 100m to 1.5km away. I have tried a few different motor types for the movement and now have a few questions. I will eventually want the camera panning/tracking system to be run from a battery source so power consumption will need to be a consideration.
Servo:
I initially used a cheap hobby servo to control the angle. This worked well as it is a closed loop system so I can just tell it go to the required position. The problem was I couldn't achieve the resolution (I need about 0.3degrees resolution to give about 10m resolution at 1km) and I couldn't control the speed easily.
Geared (100:1) DC Motor:
Next I tried a DC motor with 100:1 planetary gearbox. I couldn't run this slow enough (before it stalled) to position correctly.
Stepper motor:
Running this 200steps/rev with 1/16 micro stepping gives me 0.11 degrees resolution. And I have good control over the speed. This is ideal except I'm sure this will consume alot of power @1.8A per coil (2 coils). This is for a NEMA17 I don't require this much torque so could downsize to reduce power consumption.
Questions:
Does anyone have any ideas/suggestions for the motor mechanism?
Is it possible to get about 0.3degree accuracy with a hobby servo? Do digital servos have more precision?
Is it feasible to run a stepper off NIMH or LIPO battery packs?
Riva:
What about a small 64:1 stepper with 64:1 gearing like this. 5V so may last longer on batteries and the gearbox will give more torque.
I would suggest the same motor, because you can switch off the power to the motor without loosing psoition. The gearbox has a double benefit here. Be aware this stepper needs a bit different code, but thats all well documented when you search for 28BYJ.
Those little 5V motors probably don't have the torque alas. Also the position repeatability
is very poor as the gears have much backlash. However something similar but a bit larger
would be ideal, but might take a some finding.
A NEMA11 or similar stepper with some reduction gearing might be a possibility -
note that reducing current when a stepper is stationary is a good way to reduce
overall current consumption.
Sail winch servo is interesting. Although this will still have the same resolution as a standard servo?
The bigger version of the stepper you posted could do the job. They don't show torque rating though unfortunately?
I haven't done the calcs to work out how much torque I'll need. But it will be small, I'll be mounting a handheld HD video camera 1.5kg max and it'll be center over the axis of rotation. I'll probably be using 47/11 ratio plastic hearing one gears (from 3d printer extruder) or similar. As this provides a base for camera mount and I can add bearings etc.
MarkT how would you normally reduce current when stationary. Pwm? Don't know if this is possible with the A4988 pololu stepper driver I have, step + dir inputs. This could get tricky if stationary at a micro step. I guess it would be key to always try stop at a full step where only one coil is powered?
Yes I need absolute positioning. I was going to use a electronic compass that I have but I'm getting too much interference from the stepper. Even if disabling the stepper. So I think now ill just have to setup at a know bearing and keep track of position. Unless someone knows some awesome tricks to reduce magnetic interference for compasses??
If your camera mount is well-balanced and you use use a stepper motor, then you should need little or no current (and hence torque) to hold the camera in the position you have moved it to. You could hack a cheap A4988-based stepper driver board so that you can control the input to the Vref pin (and hence the current) from the Arduino.
g_gregory8:
Unless someone knows some awesome tricks to reduce magnetic interference for compasses??
Position the compass well away from the stepper motor, and the wires connecting it to the driver, and the wires supplying power to the driver.
Use twisted-pair wires to connect the stepper motor windings to the driver, and to connect power to the driver.
Only take compass readings when the stepper is not moving, and reduce the stepper motor current to a low value before taking a reading.
I've tried positioning the compass away from motor but it needs to be at least 400mm otherwise it gets affected. This isn't really feasible with my project. I've tried disabling the stepper entirely (no current through coils) and it still has the same effect.
I can try the twisted pairs to reduce interference.
g_gregory8:
MarkT how would you normally reduce current when stationary. Pwm? Don't know if this is possible with the A4988 pololu stepper driver I have, step + dir inputs. This could get tricky if stationary at a micro step. I guess it would be key to always try stop at a full step where only one coil is powered?
Not sure the Pololu board allows it - isn't the current set by a trimmer pot? You want to
be able to set the current programmatically, but note there are other driver boards out there.
Most commercial stepper drivers have explicitly configurable current reduction on idle.
Yes I need absolute positioning. I was going to use a electronic compass that I have but I'm getting too much interference from the stepper. Even if disabling the stepper. So I think now ill just have to setup at a know bearing and keep track of position. Unless someone knows some awesome tricks to reduce magnetic interference for compasses??
Then you need an (absolute) shaft encoder.... Compasses will never work near an electric motor, and MEMS compasses are very noisy indeed. Such encoders tend to be expensive alas.