[nw_lton]

This site was recommended to me on the weekend, and I'm loving it so far!

I've been looking into building a lightweight Dobsonian scope for a few years now, but the thought of spending weeks grinding a primary doesn't quite appeal.

I know the issues of mechanical grinding jigs, and the repetition errors and symmetrical wear, but what if it's movements were slightly unpredictable.

(This idea has most likely been thought of already, but here goes...)

Consider a setup where:
The work top was solidly mounted, but could rotate to any degree.
The tool was mounted in an XY plotter style setup, with spring-loaded mounts to apply gentle pressure to the workpiece.

The workpiecewould do the regular W shaped movements over the edge of the tool as normal to produce the convex bottom piece, and the concave top piece, but here's the sneaky bit.

Have an AD converter take a sample from a radio station. This sample is then divided back to influence the final position of each of the four positions of the 'W' movement.

Say, for 0 input volume, and the top stroke in the middle of the 'W', it's position is X:300, Y:50, but, for 100% volume, it's X:320; Y:80.

These co-ordinates may only be a few millimeters apart, but that way, it could possibly simulate the slight randomness a human grinder gives to the piece.

Similarly, the bottom worktop would rotate 5% +- 1%/v of input signal every 20 seconds...

Any feedback on this idea?

Nige.