Robert Hedan wrote:
> All I'd need would be the assembled PCB driver boards, since you appear to
> have quite a lot of inventory on hand. :) I can take care of fabricating a
> box. I guess I'd start with 3 axis, and get that working first. All I'd
> need to know is exactly what input the drivers expect, and I should be able
> to feed it something, somehow, someday, from a PIC.
>
Got my board up and running tonight, it's been probably more than a year
since I stopped drilling and have done almost completely top side boards for
testing. So I'm now remembering a few more changes to make..
50% or even 25% blind chopping on hold. Right now it just sits there baking the
motors, since I only used it intermittently I never cared. 100% torque
stationary is overkill since you don't have that much moving.
Blind chopping for everything. If you have say a 5V supply and 3V motors, this
would let you set it to 3/5ths duty and assume you're getting around the right
current. Since it's usually the heat that kills this should be ok even if it's
sloppy.
Microstepping. Right now it's just dead half stepping. Lots of resonating and
vibration if you're not at a good speed. Even just an 8 or 16 progressive
transfer instead of full on/off would give a tremendous improvement. A good
acceleration profile and not running at speeds that vibrate badly would also be
ok and easy to do. May just have it do full or half steps, and the micro is
just internal for smoothness, but really the supported range is rather high,
would be easy enough to give full access to say 16 position microstepping and
use 4 bits.
IR remote. I have good IR code already built up, very easy to add. Instant
keyboard input, for manual jogging and more for and only a single pin, and not
stuck in one place.
There are a few other things I'm not thinking of right now no doubt. While
there's nothing too special besides the IR and math code parts, I'm not sure I
actually want to give out the source. May just make it easily available as a
programmed chip and send it with the board etc, which may be easier anyway since
it's going to need the SMT version. Won't be a real limitation since it's made
as dumb as it can be made, only goes point to point etc, with the real code work
being all on the PC side.
Was really just designed to be a basic motor running unit. While it could be
made into something very advanced without too much effort I don't think most
people would need much beyond these improvements. I'd be willing to stick about
anything anyone wants in, but it'd be on a fairly long timetable for things
beyond this. Proper feedback based chopping is about the only other thing I'd
use so it too may go in eventually.
Alan