Comment by Kadin | Hacker News Reader

Comment by Kadin | Hacker News Reader

Parent Full thread

I have an inexpensive stepper-based cutter and it has really impressed me. I had very low expectations when I bought it, but it's quite nice as a pen plotter.

With an 0.05mm fiber tip technical pen, the stepper movements are indistinguishable in the final drawing. I've also run the same drawing multiple times to check repeatability, and it's within probably 10 thou (10 * [1/1000 in]) at least at first glance with a small drawing.

Having a servo-based movement is probably important if you are cutting tough outdoor vinyl or stencil material, but if you are just moving a pen across paper, some of the advantages of servos don't come into play.

The only problem I've had with it is when cutting 4-5mil mylar for stencils. There, because of variations in the thickness of the mylar (I think), the machine sometimes fails to cut completely. It doesn't lose its place in the drawing, though, which makes me think that they are running the stepper motors in a closed-loop mode similar to how you'd use a servo. I am not 100% on that, it's just a theory I've developed after watching the thing at work for a while. I'm still enjoying just playing with it enough for the moment, so I haven't bothered to dig into its innards.

Lots of fun for under 200 bucks though, IMO.

"makes me think that they are running the stepper motors in a closed-loop mode similar to how you'd use a servo. I am not 100% on that, "

Do the steppers have encoders on them? If not, the machine can't do closed loop control. It can keep track of where the rotor should be based on counting the steps of coils it has energized, but it has no way of directly keeping track of the rotor. A large enough torque resisting the rotor's rotation can cause the rotor to skip steps, but this doesn't influence the energizing of coils in the stator so the machine has no idea that anything has gone wrong.

An encoder lets you directly measure what's going on with the rotor (and therefore the shaft of the motor). You basically keep track of where you should be (based on counting the steps delivered to the stator coils) then use the encoder to validate. e.g. 'I've sent 12 steps so I should be 48 degrees from where I started. My encoder is reading 26 degrees, so I must have skipped steps somewhere along the way.'

If you use an encoder to adjust your position, then you are a servo; you probably won't use a stepper motor because the main advantage to a stepper motor is that you don't need position feedback to estimate the position accurately, but a stepper motor with feedback is just as much a servomechanism as any other motor with negative feedback.

[edit]

Wikipedia tells me that there are some mechanisms that use stepper motors that run mostly open-loop with an encoder only for detecting stepper-miss. Whether or not this qualifies as a servo-mechanism is up for debate.

In any event, I seriously doubt that a sub $200 plotter is doing this, as the encoder adds to the BOM.

If the step distance is very small then many designs would tolerate a fair amount of stepper-miss before being noticeable. Might be easiest to see if you did half of a shape, then went somewhere else and finished the shape.

What's the part number for the one you have, please?

Reply on news.ycombinator.com