I Have finally got my Polargraph built and seem to be having some issues getting the settings correct. It seems that the motors move way too far. I am using a mega with Motor shield v1. My steppers are random but managed to find the model number online and these are the specs I can find...

Steps per Revolution: 200
Step Angle: 1.8°
http://cncsuperstore.com/index.php?route=product/product&product_id=64

The only way I managed to get this to print was to set the mm per rev to 1000, not good I know!

So my steps per rev should be 200
mm per rev should be at 95?
What should my step multiplier be?

Please if anyone can help I would be much appreciative!

Thanks Sandy for the quick response! I will give it a go tomorrow and see what happens.

That's good news. The step multiplier is a bit of a funny anomaly, I'll try to explain why it's there and what it means.

All motors have a basic number of steps per rev, usually 200.

With motor drivers that run in full steps, there are 200 steps per rev, and so that is your basic resolution. One set of coils turn off, and another set turn on, and the rotor jumps from one position to another.

Can't move in half steps or quarter steps can only move in full steps. This is great for reliability. When the machine stops, it's on a full step and only one set of coils is energised. So it's a positive lock, and the machine can keep in that position indefinitely.

This is addressed simply: move to position 2000, simply steps the motor 2000 times, because there is no step multiplier.

The adafruit motorshield v1 drivers has a step style called "interleaved", which creates a kind of half-step that is also pretty reliable. This allows us to double the resolution, and that's why the steps per rev is doubled in your case (400 half-steps). So moving to the same position as before would be 4000 steps, and because all those steps are reliable, then you can actually go to each one of those steps, and stop there if you want to.

Microstepping is a way of driving the motor that slides the current from one set of coils to the other in a series of steps, rather than suddenly. It makes for much smoother transitions from step to step. So the motor drivers that the PolargraphSD uses is configured to use 8x microstepping, so each full step of the motor actually consists of 8 small steps. To move to position 2000, you'd actually need to step 16,000 times.

The problem is that the positions in between full steps are not reliable. If I step once, then one of the coils goes down to 87.5% of full current, and the other coil comes up to 12.5% of full current. This pulls the rotor one eighth of the way towards the next position.

If you don't step again, it stays in this "in between" steps, and warms up, and eventually snaps back to the original position. Electromagnetic physics collude to make it so, and it's true even if it's exactly in the middle - it'll snap either back to the last step, or forward to the next.

So you've lost track of where you started, and you now don't know where the motor really is. If every time you stop, you lose half a step because you pause and it snaps back, then eventually you're a couple of steps out.

So long story right? The purpose of step multiplier is to state how much microstepping to assume. We address the machine in full steps (200 steps per rev), because those are the steps that we know it's safe to stop on, but we tell it to multiply all of the coordinates by stepMultiplier to get the real position.

We say "go to 2000", it says, "ok, stepMultiplier is 8, so I'm going to 16,000 (8x2000)".

In the case of the adafruit shield, it doesn't microstep, so we say "go to 2000", it says, "ok, stepMultiplier is 1, so I'm going to 2000 (1x2000)".

That was a long winded way of saying "no, nothing to do with resolution", ha!

sn