Hello, it is interesting. To reduce costs, replacing the IC ATSAMD21 will be effective. If replaced stm32f405 would be better with a project micropython:

https://www.kickstarter.com/projects/214379695/micro-python-python-for-microcontrollers

Might be useful to have unpopulated boards available at a discount for the more diy types...it might also help out Kris and other more cost-sensitive hackers like him out there. I'd probably be in for one if it were cheap enough. A $5 pricetag might get a significant percentage of the backers to want to include one...just a thought.

Awesome! I'm probably going to pitch in for one assembled with the stepper regardless, but do you have any more info about the stepper?

Here it is:
http://www.omc-stepperonline.com/3d-printer-nema-17-stepper-motor-2a-45ncm64ozin-17hs162004s-p-16.html

Thanks for the super fast reply!

I like Mechaduino, great idea. $45 seems high. Is that what the components cost plus some profit? Please clarify.

This is pretty much the cost of components, pcb, assembly. We didn't want to cut any corners as far as performance, so there are a couple fairly expensive components...specifically the 14 bit encoder. We think it is worth it though.

We're taking very little off the top. Really. We're much more interested in getting the hardware into the hands of makers and developers.

As a comparison, this is still cheaper than a plain Arduino Zero.

Thanks for your interest. Feel free to dig into the BOM. If you can find ways to bring the cost down, we're all ears!

Good response, thank you. I'm getting at least 1 with stepper.

For what it's worth, my hackspace is building a bunch of these from the source files, and yeah, material costs alone are in the $35 range. That's without assembly, extras in case of screwups, or profitability. So $45 for this assembled and shipped is a killer deal.

The $8 encoder IC is the big item.

Looks good.
Have you done any tests to quantify the heatup of the driver and current sense resistors during operation?

Just qualitative tests. Current sense resistors are operating well within their ratings, so the concern is more with the driver. We've been pushing them fairly hard in our testing and haven't had any problems...they get warm to the touch, but not really hot.

We thought carefully about thermal management when we laid out the board. You can always put a little heatsink on the driver if you plan on pushing them. It would pretty straight-forward to wire up a thermistor to an analog in pin if you wanted to monitor temperature.

In closed loop operation, current is applied only to correct position errors, so they run a lot cooler than standard stepper drivers.