Close

First Video of Motor Testing

A project log for TrillSat

Texting Without Cell Towers, Power Grid, or the Internet: IoT meets AX.25 Aboard an Arboreal Space Elevator

lee-djavaherianLee Djavaherian 07/16/2018 at 06:140 Comments


I just published the first video of the two motors in operation, showing the tilting of the solar panel/capstan hoist mechanism on the tether using the wooden Ark and Test Frame and also showing the Gyrofan used for inertial stabilization.  It's the first YouTube video that I've ever uploaded, and it's pretty crude.  I was able to fully load the mass on the gyro wheel and ramped it up to a fixed RPM, which it holds fairly constant, based on hall-effect sensor feedback.  The wagon-wheel effect is pretty mesmerizing.

I also made a pass at computing the solar position using just the two CdS LDRs at rest, instead of relying on sunrise/sunset tables, but the trigonometry is surprisingly tricky.  It's easy to find the brightest part in the sky--you just move the craft until the two opposing LDR sensors are close in value to one another, but it's much more difficult to estimate solar position using just these values alone.  In other words, to have the craft "see" the sun and report its exact position is difficult. 

First you have to know the angle of the craft (which I already know via the accelerometer), and then you have account for Lambert's cosine law, which is fairly easy.  But if the angle of the craft is not horizontal (or even stationary), you don't really know which side of the sun to which one of your LDRs might be pointing.  I can use sunrise/sunset tables for my location, but this requires an accurate clock, which is also a problem (as the lack of RTCs means that I have to manually set my clocks, and auto-setting them from solar position isn't accurate to more than around 30 minutes, which brings me back to the same problem...).  So rather than estimating "where" the sun is, for now I'm going to have to just go out and find it like traditional solar trackers do, by moving the motors to find the brightest point.

I'm encountering some additional problems, too:  the paracord sheath tends to slip over time, allowing the internal fibers to bunch-up, making the mechanism less smooth as I test.  Testing it on the rough PETG snags and takes a toll on the fibers.  And the Sign-Magnitude/Lock-Antiphase orbital mechanics that work so well testing on a table, which I added to my Park command, need to be heavily tweaked for the forces on the tether.

The capstan/motor is powerful enough to drive and tilt the craft, but the PWM slow-starting torque is low, and raising the torque to overcome the static friction causes jerks.  These jerks were fairly low in early testing on the tether, but when I attached the heavy Planetary Mass, the jerking is heavily exaggerated.  So I'll have to pulse the motor at higher current to overcome the static friction, then quickly slow it down.  I might also experiment with PFM modulation in the future.

If you notice, in the video, you can hear the high-pitched whine of the PWM (since I use audible frequencies), but I have to raise the duty cycle fairly high before it starts moving (unless it is on a "downhill", and then it whips around really fast and has to be slowed down (which adds wear to the paracord).  The ball-damper that I added inside Planetary Mass helps to remove the stress of the jerks, though.

The orbital mechanics aren't going to be a graceful truck driving down a smooth, sinusoidal hill, as I illustrated in my diagram--that's just the idealized path.  In reality, it's going to be like a old truck driving down a bumpy, rock-strewn road full of potholes.  The driver will have to hit the gas hard at points, then let off and hit the brakes, etc.

Discussions