SPID RAS/HR Az/El rotor

Hard technical data on the SPID rotors is somewhat lacking. There are spec sheets (some of which sometimes contain errors), but there's little detail on exactly how the system works. The information here is based on my testing of a SPID RAS/HR rotor with MD-02 controller (vintage 2017).

Like all DC motor speed controllers, SPID use PWM (Pulse Width Modulation) to control how fast the motors rotate (and to some extent how much torque they can provide. While the motors are nominally 12v, the SPID literature suggests they can be run at higher voltages (to get faster or higher torque operation). The pulse frequency appears to be around 1kHz and the controller allows power ramps and a maximum power to be set.

Using a laser diode attached to the dish and pointing at a distant screen, small angular rotations of the AZ/EL tracking system can be followed. In the images below, each division is 0.05 degrees. The small circle is 0.2 degrees in diameter, the next larger circle is 0.5 degrees in diameter and the largest circle is 1 degree in diameter.

The following series of images show the progression of the laser spot across the target as a SPID RAS/HR Az/El rotor is instructed to move in 0.2 degree steps from the PstRotator program. The resolution of the SPID RAS/HR is given in the spec sheet as 0.2 degrees. The RAS/HR uses a hall effect sensor system. One of the internal gears generates 32 pulses per revolution and one revolution of this gear turns the rotor by 4 degrees. This means that a pulse is generated every 0.125 degrees of rotation.

Mathematically inclined readers will see a small problem here. The controller tells the rotor to rotate and then counts pulses to tell how far it has rotated. You can't move by 0.2 degrees is the pulse occurs every 0.125 degrees. You can move 1 degree by counting 8 pulses, but you can't rotate exactly 0.2, 0.4, 0.6 or 0.8 degrees, so some sort of approximation has to be going on. You can see what happens in the following images. Each successive image is taken after a command to rotate 0.2 degrees has been given and (most of the time) this results in an increase in the display angle of 0.2 degrees.

Starting from the top and moving down, so can see that there is first a small rotation step, then a larger one, then another larger one, then a smaller one, follower by a larger step and a smaller step. Rounding off the readings these appear to be steps of 0.125, 0.25, 0.25, 0.125, 0.25 and 0.125 degrees. The first 5 steps add up to 1 degree and that corresponds to the 1 degree expect from 5 0.2 degree steps. To everything adds up right at 1 degree intervals, but the steps are uneven. The rotation angles are 0.125, 0.375, 0.625, 0.75 and 1.0 rather than 0, 0.2, 0.4, 0.6, 0.8 and 1 degree. It's a small difference and below the resolution limit of the display, so you don't see the odd steps, though occasionally the display will increase by jut 0.1 degrees, or 0.3 degrees, rather than the expected 0.2 degrees as the numbers are rounded. The numbers are never more than 0.05 degrees from where they should be so everything is good to within the "0.2 degree resolution" specification.

Backlash

All geared rotor systems (in fact just about any geared system) has backlash. This is the free play between the gear teeth. If they were always in perfect contact they would rapidly wear. There's always a little space which allows the final gear to move back and forth a little. The image below shows the azimuth difference of a given azimuth setting approached from either side. So if the setting was 180 degrees, one image would be when going from 170-> 180 degrees and the other image would be going from 190->180 degrees. The difference between where the rotor points in these two cases is the backlash.

This image shows a difference of about 7 divisions, which corresponds to about 0.35 degrees. Azimuth backlash is difficult to compensate for, but can be minimized you always approaching a setting from the same direction. For EME this is always the case since the moon moves continuously in only one direction. In high wind the dish may move back and forth over the backlash angle unless some sort of "bias" force is applied to it on one direction.

Backlash in elevation is less of a problem because the load is always on one direction which depends on how the counter weight is adjusted. If a dish is perfectly balanced it can move up and down over the range allowed by backlash, so it's best to keep it either slightly front or rear heavy.