Yesterday Sibille and I brought the Phyllo structure with the faulty motor (see this post) to the school mechanic, Mr Croullebois. We disassembled the structure, and it turns out the motor shaft goes slightly lower than the motor base, so when we screwed it tight it would scrape against the aluminum plate underneath.
Mr Croullebois bored a depression in the middle of the plate so we wouldn’t have this problem anymore, and then we reassembled everything with the other motor that we used for tests and knew to be working. The motor that was in the structure spins more freely now, and we managed to make it run, but there is a weird creaking sound from time to time. It feels like the ball bearings got damaged, so while we can use it for tests, we’ll probably order a new motor to replace it.
We were then finally able to test the Phyllo structure out and make the PVC disk spin 🙂
This gave us a chance to see the power consumption under partial load. During acceleration phases it got as high as 1.5 A. That’s still acceptable, especially as it is much lower at steady speeds. We’ll see what happens once we have a fuller load.
However we soon noticed another issue : when we tried dropping the generator voltage, we noticed that the maximum speed we can achieve drops sharply around 16.5 V. Below that, no matter what throttle command we send the ESC, the speed never goes faster than a certain – noticeably slow – point. After 16.5 V, this limit rises sharply and a higher throttle command results in a higher speed. We can’t really measure the speed yet, but we think the limit below 16.5 V is too slow for our needs. This would mean that we actually need a 6S LiPo battery for the motor (its fully discharged voltage would be 16.2 V). We’ll do more tests to confirm.
Something else also came up : we noticed that once per rotation there was an increased friction sound. So we brought the structure back to the school mechanic, and we realised that the shaft doesn’t turn quite straight. We tracked the problem back to the point where the shaft attaches to the moto. After carefully sanding the contact surfaces and fine tuning how everything is screwed together, we managed to reduce the problem. It’s still not perfect, but it’s the best we’re likely to get.
We borrowed an optical interruption sensor from Litspin (we also ordered some but it turns out they got lost on the way, so we’ll give theirs back to Litspin when we receive ours) and started making a setup to generate an interruption once per rotation and measure the motor speed. The sensor consists an emitter facing two receptors, allowing us to detect if something blocks the view between the emitter and the receptors. Using a Saleae logic analyzer we first confirmed that the optical sensor works. Then we laser-cut discs that fit between the sensor’s emitter and receptors. If we cut a small hole somewhere on the outside of the disc and spin it with the motor, once per rotation the hole will pass between the receptor and the emitter and the receptor output will go from low to high for a short moment.
That’s all we had time to do before we called it a night, but we should be able to hit the ground running next week 🙂
Stay tuned !