Re: [CAD_CAM_EDM_DRO] Re: Geckodrive
Posted by
art
on 2006-04-29 08:15:23 UTC
Hi Chris:
tightly the motor "stalls.", however, it can be called unstallable, because
it really isnt stalled at all, the hardware at that point is calling for no
movement. (You cant stall a motor that isnt moving..). In actual fact the
motor is holding itself at the maximum magnetic moment it can achieve
without "clicking" into the next position at which it will lose position.
The Lead/Lag register in the G100 can measure up to 7 counts off target.
So imagine your moving at speed, the lead/lag is zero indicating the motor
is not going too fast, nor too slow, the necode agrees 100% with position.
SO a person puts their hand out and pushes against the axis's movement. The
Lag register at that point counts up to one, indicating that the encoder is
now one pulse behind. So the axis output stage , slows slightly to see if
the motor then catches up. This catchup is likely because the torque of the
motor has just increased a bit due to going slower. This is additive in
nature, the more you push, the more the output slows down and the higher the
motors torque. The trick here is speed. If the G100 can quickly enough
monitor (on a step by step basis) the lead/lag, then it can reduce the
ourtput speed quickly anough to produce an output speed correction that is
smooth enough to allow the motors torque curve to compensate in a curve that
assures fast enough correction time. This produces, for all intents and
purposes, an unstallable motor.
It is planned, very shortly, for MachIV to start experimentally using
the lead-lag to have "Stall prevention" , since it cannot deal fast enough
as a program to produce the unstallable motor, a hardware feature, I was
thinking even under normal software control, you can have Lag determined
speed vectors. SO as the program sees up to 4 times a secnd the lag
registers, it should be possible during cutting to adjust the feedrate to
ensure the lag stays as close to zero as possible, in the odd event a person
has programmed a feedrate too high, or is hitting denser material in wood
and such. This "slow adjustment" has a possibility fo allowing for a small
correction in feed values in what I would call "Stall prevention" on the
application side. Not near as good a solution as Mariss's , but it may allow
one addition level of tool bit load smoothing. It'll be an interesting
experiment anyway. Im adding in the Lag registers use soon, cuz I did want
to play a little bit in that regard.. The functions of the MachIV trajectory
planner allow me some flexability n that area to apply corrections in
feedraet at intervals of 64ms or so feedrate variations.. A lower term
solution than an unstallable stepper, but one that also bears examination, a
corralary to a spindle load sensor of sorts..In fact , my plan was to simply
use it on the spindle output frequency, and see how it may work on axis
movement after that. In other words I intend to use "Lead/Lag" to regulate
the load on the spindle
output.. Im just waiting for the Spindle partition to be ready..
Thanks,
Art
www.artofcnc.ca
Videos And Support Forums http://www.machsupport.com
Users Map: http://www.frappr.com/machsupport
>>> Perhaps I don't understand it based on what you are calling it. ButA matter of definition. It is stallable, if the pliers are held very
> as far as I know, every motor of every kind can be stalled with
> enough load on it.
tightly the motor "stalls.", however, it can be called unstallable, because
it really isnt stalled at all, the hardware at that point is calling for no
movement. (You cant stall a motor that isnt moving..). In actual fact the
motor is holding itself at the maximum magnetic moment it can achieve
without "clicking" into the next position at which it will lose position.
The Lead/Lag register in the G100 can measure up to 7 counts off target.
So imagine your moving at speed, the lead/lag is zero indicating the motor
is not going too fast, nor too slow, the necode agrees 100% with position.
SO a person puts their hand out and pushes against the axis's movement. The
Lag register at that point counts up to one, indicating that the encoder is
now one pulse behind. So the axis output stage , slows slightly to see if
the motor then catches up. This catchup is likely because the torque of the
motor has just increased a bit due to going slower. This is additive in
nature, the more you push, the more the output slows down and the higher the
motors torque. The trick here is speed. If the G100 can quickly enough
monitor (on a step by step basis) the lead/lag, then it can reduce the
ourtput speed quickly anough to produce an output speed correction that is
smooth enough to allow the motors torque curve to compensate in a curve that
assures fast enough correction time. This produces, for all intents and
purposes, an unstallable motor.
It is planned, very shortly, for MachIV to start experimentally using
the lead-lag to have "Stall prevention" , since it cannot deal fast enough
as a program to produce the unstallable motor, a hardware feature, I was
thinking even under normal software control, you can have Lag determined
speed vectors. SO as the program sees up to 4 times a secnd the lag
registers, it should be possible during cutting to adjust the feedrate to
ensure the lag stays as close to zero as possible, in the odd event a person
has programmed a feedrate too high, or is hitting denser material in wood
and such. This "slow adjustment" has a possibility fo allowing for a small
correction in feed values in what I would call "Stall prevention" on the
application side. Not near as good a solution as Mariss's , but it may allow
one addition level of tool bit load smoothing. It'll be an interesting
experiment anyway. Im adding in the Lag registers use soon, cuz I did want
to play a little bit in that regard.. The functions of the MachIV trajectory
planner allow me some flexability n that area to apply corrections in
feedraet at intervals of 64ms or so feedrate variations.. A lower term
solution than an unstallable stepper, but one that also bears examination, a
corralary to a spindle load sensor of sorts..In fact , my plan was to simply
use it on the spindle output frequency, and see how it may work on axis
movement after that. In other words I intend to use "Lead/Lag" to regulate
the load on the spindle
output.. Im just waiting for the Spindle partition to be ready..
Thanks,
Art
www.artofcnc.ca
Videos And Support Forums http://www.machsupport.com
Users Map: http://www.frappr.com/machsupport
----- Original Message -----
From: "lcdpublishing" <lcdpublishing@...>
To: <CAD_CAM_EDM_DRO@yahoogroups.com>
Sent: Saturday, April 29, 2006 8:34 AM
Subject: [CAD_CAM_EDM_DRO] Re: Geckodrive
> Perhaps I don't understand it based on what you are calling it. But
> as far as I know, every motor of every kind can be stalled with
> enough load on it. So, is it "Unstallable" in the sense the motor
> cannot be stopped?
>
> Or is it just closing the loop?
>
>
>
>
>
>
>
>
>
>
>
> --- In CAD_CAM_EDM_DRO@yahoogroups.com, "Mariss Freimanis"
> <mariss92705@...> wrote:
>>
>> Sorry for the cross posting but it's been a stellar day here. This
>> has to do with the "Unstallable Step Motor Project" that may be of
>> interest to some:
>> -----------------------------
>> Hi,
>>
>> The very crux of the G100 project is to have PID servoed step
> motors
>> that cannot be stalled under load. The G100 hardware was designed
> with
>> this goal in mind; that is why it has 6 encoder input channels, one
>> for
>> each axis. I have recently made/found the time to work towards this
>> goal.
>>
>> Before anyone asks "When will it be done?", don't. It will be done
>> when
>> it's done. A lot of steps, work and time stands between what has
> been
>> recently accomplished and a finished product.
>>
>> The goal: Imagine you are doing 3-D contouring (X,Y,Z) when someone
>> decides to sit on your gantry. An ordinary open-loop stepper or
> servo
>> motor would stall/fault-out and that would be that.
>>
>> Now imagine the overloaded X-axis servo-stepper would just slow
> down
>> under this load. The other axis (Y,Z) would slow down as well,
> keeping
>> their speed ratios with the overburdened X-axis. The 3-D contouring
>> path would be strictly maintained albeit at a much reduced speed.
> Once
>> your friend got off of the gantry, speed would pick up to where it
> was
>> before. Your work would not be ruined.
>>
>> How would this be done? Your step motors would be mounted with
>> encoders. The G100 would sense 1 or more motors are about to stall
>> (develop a phase lag of 1.8 degrees). The G100 would slow them down
>> and
>> back up the speed-torque curve to where more "grunt" (torque) was
>> available.
>>
>> Steppers have the unique characteristic of having a speed-torque
> curve
>> where the slower you go, the more torque that is available. Can't
> do
>> that with a regular servomotor.
>>
>> OK. That's the goal; that's the picture in my mind. Here's what has
>> been done by today:
>>
>> I have a G100 driving a G201 running a step motor at 3,000 RPM
> (NEMA
>> 23, 3A/phase, 24VDC). Open-loop, you so much as look at it and it
> will
>> stall. A flea landing on the shaft would stall it open-loop. I
> take a
>> pair of pliers and clamp down on a wet paper towel folded over the
>> motor shaft and all I get is steam. The motor will not stall. It
> slows
>> down under load but it delivers 50W to the paper towel/pliers
> combo to
>> make said steam. It goes all the way down to zero RPM without
> losing
>> position.
>>
>> I ease up on the pliers and the motor is back up to 3,000 RPM
>> lickedly-
>> split, gives a chance for the boiling-hot paper towel to cool off.
> Not
>> a step has been lost; you just can't kill the motor (stall it).
>>
>> This is just the first step (though a very important one) to
> integrate
>> the unstallable feature into a working CNC firmware. There are many
>> problems this feature presents.
>>
>> Just right now, having it do what it does, has me totally stoked.
> It,s
>> so cool, jerk down on the pliers, motor slows down. Instruments say
>> it's delivering 95%+ of all it has to give at all speeds. Let up
> and
>> bing-boom-bam, 3,000 RPM again. Just can't kill it no matter what I
>> do;
>> very fine Friday nite indeed.
>>
>> Before you ask #2: Yes, it will traslate into all step motors of
> all
>> sizes at all supply voltages at all speeds.
>>
>> Mariss
>>
>
>
>
>
>
>
> Addresses:
> FAQ: http://www.ktmarketing.com/faq.html
> FILES: http://groups.yahoo.com/group/CAD_CAM_EDM_DRO/files/
> Post Messages: CAD_CAM_EDM_DRO@yahoogroups.com
>
> Subscribe: CAD_CAM_EDM_DRO-subscribe@yahoogroups.com
> Unsubscribe: CAD_CAM_EDM_DRO-unsubscribe@yahoogroups.com
> List owner: CAD_CAM_EDM_DRO-owner@yahoogroups.com, wanliker@...,
> timg@...
> Moderator: pentam@... indigo_red@... davemucha@...
> [Moderators]
> URL to this group: http://groups.yahoo.com/group/CAD_CAM_EDM_DRO
>
> OFF Topic POSTS: General Machining
> If you wish to post on unlimited OT subjects goto:
> aol://5863:126/rec.crafts.metalworking or go thru Google.com to reach it
> if you have trouble.
> http://www.metalworking.com/news_servers.html
>
> http://groups.yahoo.com/group/jobshophomeshop I consider this to be a
> sister site to the CCED group, as many of the same members are there, for
> OT subjects, that are not allowed on the CCED list.
>
> NOTICE: ALL POSTINGS TO THIS GROUP BECOME PUBLIC DOMAIN BY POSTING THEM.
> DON'T POST IF YOU CAN NOT ACCEPT THIS.....NO EXCEPTIONS........
> bill
> List Mom
> List Owner
>
>
> Yahoo! Groups Links
>
>
>
>
>
>
Discussion Thread
Mariss Freimanis
2006-04-28 22:54:20 UTC
Geckodrive
juan gelt
2006-04-29 04:19:37 UTC
Re: [CAD_CAM_EDM_DRO] Geckodrive
lcdpublishing
2006-04-29 04:42:48 UTC
Re: Geckodrive
art
2006-04-29 08:15:23 UTC
Re: [CAD_CAM_EDM_DRO] Re: Geckodrive
Dan Mauch
2006-04-29 08:31:13 UTC
Re: [CAD_CAM_EDM_DRO] Geckodrive
Mariss Freimanis
2006-04-29 10:23:13 UTC
Re: Geckodrive
Alan Marconett
2006-04-29 10:30:26 UTC
Re: [CAD_CAM_EDM_DRO] Geckodrive
art
2006-04-29 10:38:39 UTC
Re: [CAD_CAM_EDM_DRO] Re: Geckodrive
Jack Hudler
2006-04-29 11:20:56 UTC
RE: [CAD_CAM_EDM_DRO] Geckodrive
Mariss Freimanis
2006-04-29 12:44:24 UTC
Re: Geckodrive
cncnovice
2006-04-29 13:02:00 UTC
Re: Geckodrive
lcdpublishing
2006-04-29 13:23:32 UTC
Re: Geckodrive
Jon Elson
2006-04-29 13:30:16 UTC
Re: [CAD_CAM_EDM_DRO] Geckodrive
Mariss Freimanis
2006-04-29 14:12:05 UTC
Re: Geckodrive
Larry Olson
2006-04-29 14:14:40 UTC
RE: [CAD_CAM_EDM_DRO] Re: Geckodrive
Graham Stabler
2006-04-29 14:21:59 UTC
Re: Geckodrive
Mariss Freimanis
2006-04-29 15:06:15 UTC
Re: Geckodrive
Jack Hudler
2006-04-29 16:09:33 UTC
RE: [CAD_CAM_EDM_DRO] Re: Geckodrive
Jack Hudler
2006-04-29 16:21:16 UTC
RE: [CAD_CAM_EDM_DRO] Re: Geckodrive
Polaraligned
2006-04-29 17:05:59 UTC
Re: Geckodrive
Mariss Freimanis
2006-04-29 17:37:37 UTC
Re: Geckodrive
Mariss Freimanis
2006-04-29 18:20:07 UTC
Re: Geckodrive
Jack Hudler
2006-04-29 23:47:44 UTC
RE: [CAD_CAM_EDM_DRO] Re: Geckodrive
lcdpublishing
2006-04-30 04:30:48 UTC
Re: Geckodrive
vrsculptor
2006-04-30 04:34:45 UTC
Re: Geckodrive
Mariss Freimanis
2006-04-30 08:08:14 UTC
Re: Geckodrive
Mariss Freimanis
2006-04-30 08:14:46 UTC
Re: Geckodrive
vrsculptor
2006-04-30 12:38:39 UTC
Stall prevention/MachIV/Automated Setup
ballendo
2006-04-30 14:50:49 UTC
Re: Geckodrive
Mariss Freimanis
2006-04-30 18:22:58 UTC
Re: Stall prevention/MachIV/Automated Setup
Polaraligned
2006-05-01 03:28:30 UTC
Re: Geckodrive
lcdpublishing
2006-05-01 05:57:20 UTC
Re: Geckodrive
Mariss Freimanis
2006-05-01 08:56:04 UTC
Re: Geckodrive
lcdpublishing
2006-05-01 09:44:43 UTC
Re: Geckodrive
Jon Elson
2006-05-01 09:50:47 UTC
Re: [CAD_CAM_EDM_DRO] Re: Geckodrive
Mariss Freimanis
2006-05-01 12:40:11 UTC
Re: Geckodrive
lcdpublishing
2006-05-01 13:58:42 UTC
Re: Geckodrive