I think it’s more likely that the drift is caused by poor tolerances in the foam board construction, movement of the base of the arm, movement of the writing surface, or some combination of those. Otherwise, I think it’s a good design and good instructions for adding a 2nd pivot-point to the servos.

Yea that and people tend to not have a reasonable idea of accuracy, like the guy I saw yesterday wanting to achieve 0.001 degree steps

Hey there,

Just seen this :) The reason why I blame the servos is even without the foamboard they fail to measure 1 degree properly – they tend to sweep around until they settle on the correct measurement. very odd and if you set the servo to 90 degrees that point will change a few degrees each time.

The foamboard is actually suprisingly solid providing there is no significant weight on the end actuator – the weight of the arm is used to force the pen to the paper. However I’ve not used the arm alot and with repeat use the foamboard wont last, so now I know its working (kind of) I’ll rebuild with better materials.

The base did move around alot, but because i’d taped down several sheets of paper but not all that well and as the pen caught in the paper edges, the base moved. (also the video is after its been runnign for a while)

Other thing I think you can see is only 2 joints move, the 3rd one nearest the pen can’t draw enough power.

heh.. now where did i put that linear hall effect sensor?

seems to me that a relatively simple circuit could replace the (temperamental) servo pot with a robust and reliable non contact sensor based on two magnets at the ends of the servo positioner with the poles opposing (i.e. north at one end and south at the other)

for small cheap mini magnets harvest them from broken cdrom drives (laptop ones are best) and superglue to the relevant points.

hope this helps someone..

I know a guy doing the hall effect sensor + motor. It’s good for knowing where it is positioned (which is what we want), but not perfect for putting it in a special place.

Not sure why really.