Printable View
I came across a very interesting project on Kickstarter:
https://www.kickstarter.com/projects...m?ref=category
It's a servo stepper. The encoder is simply a magnet glued to the back of the rotor shaft. They say the resolution is .1 deg. The main drawback is that the driver can only handle 2 amps,
it's where they are starting they are doing a bigger one soon
Actually, the main drawback is that it seems to be INcompatible with Mach3. You drive the single servo motor via the USB rather than the industry-standard Step/Dir lines. A pity, as this means they have missed out on the big market.
Their claim that OTC servo systems cost thousands of dollars is equally unfortunate, and false.
Unfortunately the cost of industrial servos is prohibitive to the individual maker (thousands of dollars per motor).
DC brush motor plus Gecko 320 driver plus optical encoder is unlikely to exceed $500, unless you want far more power than they are offering.
One must allow for excessive enthusiasm I guess.
Cheers
Roger
Reading the hackaday.io page comments, they say it already supports step/dir and a number of other input types.
read the hackaday page roger
OK, I have read the hackaday page, and I am even more confused.
As far as I can see, they are putting an encoder feedback around a STEPPER motor and they are claiming 0.1 degree or 1/3600 of a rev.
But a 200 pole stepper motor (1.8 degrees) with 16-stage micro-stepping will give 1/3200 of a rev. Granted, 16-step microstepping may not be absolutely linear, but neither is their magnetic encoder. I have to ask why bother?
And if I want reasonable power with finer resolution, I would go for a DC Brush servo motor with a 1024 line encoder in quadrature. Mind you, I would probably gear the servo motor (or the stepper) down 3:1 or 4:1 via a GT2 belt as well. As it is, with a 512 line encoder, 3:1 reduction and 5 mm pitch ball screw, I am getting something like a 0.8 micron single-step resolution - and way under their quote of (thousands of dollars per motor). I suggest that is quite good enough.
To be sure, it is an interesting solution, but I am not sure what problem they are solving. Cheaper solutions already exist. And yes, I was well-aware of that magnetic encoder chip years ago. I turned it down.
Cheers
Roger
it's more for 3D printers roger
For 3D printers?
In that case you do not need the servo feedback at all. A simple inexpensive stepper motor and a Gecko203V would be perfect, and far more resolution than any printer needs.
And I still take issue with their claim of (thousands of dollars per motor) for commercial units. That sort of thing makes me suspicious.
Cheers
Roger
fanic yer hobby nar, they show them working and other people have built them from what they give so it's not a con, you have to have a working product to do a kick starter, they would be good for the Z axis on cheap printers, when I win lotto on Wednesday I send you some.
I have a controller available for $16 open source. And a lower resolution but MUCH faster magnetic encoder for $20 or use an optical for about the same price for lower RPM but higher resolution with the same controller. Use whatever motor and driver you like. Modular systems allow you to mix and match... and you don't have to wait and hope they actually deliver. There have been many Kickstarters where they had a viable design and failed to deliver at the price they promised.
the prices is a lot higher than what they where going to charge, I would like to see your open source controller, I need something to make my printer run better
Hi Daniel
Are you commercial? If so, just buy a Gecko 203V and get back to machining! Time => $$$
Cheers
Roger
no just hobby I have one machine on a G540 another on 213v I would use gecko's on any machine I fix or use I can't justify spending 5 x what the printer cost this time. A 213v cost's the same as the printer
I have the printer ruining at it's limits to go any faster I would have to go to servo's or a closed loop stepper controller 4500 mm/min, it can rapped fine at 6000 mm/min but not print.
If the boards in this thread where what they original said the prices would of been I would of got them
The benefit of a driver like this, is that it turns a stepper motor into a brushless servo, without much increase in cost. An optical encoder will often cost more than the motor, making it hard to justify, because a bigger motor is cheaper and simpler.
With this approach, you can just superglue a 50 cent magnet on to the end of the rotor shaft. A microstepping drive might have finer resolution on paper, but that will not be achieved under load, due to the "spring" effect. Another benefit of feedback is that the driver will only apply as much current as necessary to keep the rotor in position. When properly tuned, the motion is quiet, without resonance, uses less power, and runs cooler.
2 amps at 30v is rather weak for cnc, but it is easy to put in a discrete mosfet driver stage to drive big motors.
Hum ...
Treating a stepper motor as a multi-pole brushless DC motor?
Technically feasible, and I gather that this PCB does that? Novel! I will have to recant somewhat.
It's an interesting idea. True, one might need a more powerful driver, as 2 A is hardly a lot, but still - cute!
Some measured performance specs would be real nice to have published.
Cheers
Roger