Re: Brushless Motor Spindle with CNC desktop machine

Hey Alan,
Thanks for the great post. I have a link here for you from Model
Motors in Sydney, Au that you may peruse. Instruction manuals are on
the site as well as a lot of info for the controllers. I opted for
the rotating casing/housing and fixed coils as I had heard alot of
people had blown windings from the high speed throwing them out. The
AXi has neodymium magnets with a very strong cogging effect when
turned by hand and a 5mm shaft size which is large enough for a decent
coupling. This was the motor I opted for as the price was about what
I was willing to pay for a prototype. There are bigger more powerful
motors that go way up to 75-80 Amps. WOW!!! Way too expensive though.
here is the link.
http://www.modelflight.com.au/axi.htm
hope it helps.
Warm regards.

> Hi Richard,
> Cc Dave,
>
> The RF format follows the "servo" format used in the R/C world. A PWM
> signal designed to control an R/C servo. The pulse width varies

around a

> center of 1.5 mS (.5 - 2mS), and the pulse is repeated at a 20 mS rate.
> There are PIC controllers to run R/C servos that can do this. The robot
> guys use 'em all the time.
>
> It would be difficult to control this R/C servo with Gcode, unless

it was

> the only thing the controller did! A special PWM signal (which

would have

> to run continuous) would be needed.
>
> A PIC controller could be devised to take step/dir signals and ramp

up/down

> a continuous PWM signal (1 "slot" in about 10). In this case, the CNC
> controller would generate an "up" direction and a few pulses to set the
> speed to the desired level.
>
> I like the idea of the high speed R/C spindle, just what I need for PCB
> work. I take it I can look for an "AXi 2826/10" motor at my local hobby
> store? How much?
>
> "continuous pulses" That's an oxymoron! The pulses are no doubt TTL

(.5V

> low, 4.5V hi) signals of sorts, it's the generation of them that is

a little

> more complex. And this stream of pulses doesn't start/stop like the

axis

> moves do. The "1 in 10" may be thought of as "time slots", if one were
> running 10 channels of servos (!), then the time allotted to one

servo would

> be similar to what's being described.
>
> Chuck, an old friend of mine from the "Homebrew Robotics club" (Silicon
> Valley, circa '86) has a good write up.
>
> http://www.mcmanis.com/chuck/robotics/projects/servo.html
>
> Humm... That's a good circuit for an H-bridge, if the FETs are big

enough,

> (or half an H-bridge) ...could drive your motor. And the PIC code

could

> be modified to take an occasional step/dir command to set the spindle
> speed. COOL!
>
> Alan KM6VV