After running for 24 hours I stopped the SA running and measured the distance from the finishing point to my initial mark:
Looks like it’s gone a bit too far. However, that’s in a solar day, not a sidereal day, so let’s do some calculations…
The radius of the circle described by the end of the counterweight bar was 305mm, so in 24 hours it had travelled 1916mm (circumference of the circle) plus 2mm over as measured off the wall. 1918mm in all. A sidereal day is 236 seconds shorter than a solar day, so in a sidereal day it would have travelled ( 86400 – 236 ) / 86400 x 1918 = 1913mm — 3mm short of where it ought to have got to.
That 3mm is equivalent to about 2191 arcseconds over the course of a sidereal day, or about one arcsecond too slow every 39 seconds. Imaging at a scale of, say, ten arcseconds per pixel, it’s clearly going to be a while before any error is apparent due to the tracking and quite possibly polar alignment will become a problem first.
I am slightly tempted to do attempt a more accurate measurement using a laser pointer to mark the position with more distance between the mark and the mount, but perhaps that can wait until I’m really bored 😀