Introduction

I wanted to get some data on how much the struts have to extend/retract during the day. So I made a cheap rig with a mirror supported by three M6 threads, A, B, and C.

Methods

The threads are threaded into a wooden base with some legs to keep the wood off the ground and the platform more stable. The wooden base is slightly smaller than the mirror, because that was what I had lying around. The mirror surface is 385x385 mm.

Threads A and B were about 500 mm long, and thread C was 1000 mm long (how I got it from the shop). I kept the mirror corner of strut C constant throughout the measurement, because that one is supposed to become my fixed pillar. I've added an image of the rig, so it will probably make more sense to you now.

Around noon I aligned the mirror so that it aimed at a fixed point on the wall, and then several times later, I re-aligned the mirror and noted how much extension each of the struts needed.

Results.

Strut A had the largest change of approximately 25 cm. I suspect that had I started taking measurements earlier, then I would have seen even more change. Strut B moved comparatively less with approx 15 cm.

The increase of strut B seems to be roughly linear, whereas the increase of strut B starts off more or less linearly but then becomes nonlinear.

The average increase of A per hour is approximately 5 cm/h until 14:00 and then increases to approx 14 cm/h. However, ideally I need to take more data points to verify this.

The average increase of B is roughly 6.5 mm/h.

At 15:50 the mirror was in the shade. I didn't get any early morning measurements, because I was still making the rig then.

Discussion

The length of the struts is slightly too short. Ideally for strut A, the strut length should be 10 cm longer. So for the next experiment I'll swap struts C and A around.

Strut A has a large change in length. That is good in regards to needing a less accurate position sensor. However, bad in regards to needing a larger sensor. Perhaps I'll have to consider a stepper motor or a position encoder on a linear motor instead of a parallel capacitive sensor. I could then measure the extension by counting shaft rotations.

Overall I was pleased to find that I could indeed keep C at a constant length.

M6 thread may be a bit too flexible. However, until I've got some universal or ball joints at the mirror corner points, that is quite helpful.

Conclusion

All in all, not a bad result for a first measurement.