Hello, from snowy Michigan... I'm making a polargraph as a jumping block for my architecture thesis project. I ended up using your instructable to build my own... using timing belt pulleys and timing belts instead of the cords which I could not get here...I've done 1.5 iterations (the second one is busy drawing beside me now) and I've attached images of what I've gotten so far... I've got quite a few questions... The first drawing I did is in purple, and the second is currently in process which is green.

First, as you can see in my second drawing (the green one on mylar) I'm getting some skewing off to the side, whereas the first one didn't really have that problem at all... I'm considering putting more weight on the gondola to see if that helps, this is probably just a minor issue.

Second, I am having a huge problem getting the software understand the machine size // paper size... When I input all of my measurements (Which I have double and triple checked) the bounds of the drawing is vastly different than what I've got on my screen... The image I input has to be scaled way down in for the machine to try and draw within viable bounds. If I put in the paper size values the same as my actual paper, for some reason the image is way too large for the machine to print... I've also measured and updated the sprocket circumference and everything to match my timing belt pulleys which I thought might have lended itself to this problem...

Third, I got interested in this project because I was looking into mobile 2 axis positioning systems... I eventually want to use this polargraph technique to hold much heavier objects and map much larger areas than the motors in the instructable are able to do... Is it possible to say beef up the motors to NEMA 34 or 42 and still use the same methods for control as this polargraph uses? Or would there have to be some massive recoding...?

Hallo! Tidy machine, nice work.

Swings to the left or right are to do with mishoming - at first glance it looks like the centreline might be a little far right? Could just be the angle of the photo. Or my eyes. Is it actually going to one side, or just expanding generally? Remember you can do something like "draw outline" to check out the extent of your selection area, without having to fill in all the pixels.

Your scaling problems are weird though, but I can't really see anything wrong without seeing your setup page. Off the top of my head, arduino with adafruit and 200 step motors should be:
stepsPerRev - 400.
mmPerRev - circumference of your pulley. Or rather, the length of belt that is extended with one revolution.
stepmultiplier - 1

No recoding required at all for different motors, but a beefier motor requires beefier drivers, and that might need a little tinkering. Nothing big though. The adafruit shield will only supply 0.6A per motor. The stepsticks used on the polarshield do about twice that, but bigger motors will pull 8A or more. Most stepper drivers take the same input signals though, a pulse for step and a high/low for direction, so I don't see that it'd cause a significant issue code-wise.

Have a look at Seb Lee-Delisle's Lunar Launcher (http://www.flickr.com/photos/sebleedelisle/8203136688/in/set-72157632054699823) machine for how to do it right. Not stepper motors at all.

cheers
sn

Ah and also, I rechecked my measurments... the line is... unless I'm crazy, in the dead center of the board.

I believe I've found something pretty interesting... So, the 'drifting' off to the right as seen in the other images only happens on the 'scribble' command... not on the basic first option... I get no drift on this setting... strange.