Re: Servo controllers for table top machines. long
Posted by
turbulatordude
on 2006-12-18 08:55:03 UTC
--- In CAD_CAM_EDM_DRO@yahoogroups.com, Leslie Newell <lesnewell@...>
wrote:
either has a really large table or he too missed the size the OP is
talking about.
If you are talking about a plasma machine with 600 ipm rapids, servos
are the way to go, you just cannot get a stepper to move the mass of a
heavy machine.
If you are talking a chineese desk top mini-milll (not a HAAS) a
stepper is the better choice.
I as I see it, there are 3 points. #1 servo's have encoders. that
provides feedback. Steppers do not have feedback. advantage servo's.
Yes, you can add an encoder. rarely done, and I don't think anyone
tries to use it as feedback for the control. I am sure some have,
but, it is not mainstream.
feedback allows one to know where you are. there is no way to weigh
the individuals comfot zone on hoping steppers are in the proper place
and have not missed a step.
On the table top machines, a properly designed and build machine can
use stepers. In fact, as Mariss pointed out, steppers can offer
better performance at some points. Mariss has offered that somewhere
around 200 watts of power is where steppers starts to look less
attractive and servos start to be a serious choice.
So, #2) is the power needed to drive the axis. Under 200 watts,
steppers usually have the advantage, over 300 watts, servos usually
have the advantage. I said usually, as there are advantages of
steppers, more power at low speeds. and of servo's, more power at rapids.
There are some FAST small machines that cry out for servo's, and some
slow large machines that work well with steppers.
One the speed issue. if you take said mini-mill and need to rapid the
long axis, you are moving all of about 12 inchs. at 100 inches a
minute, it takes about 10 seconds. Now, if you have a 12 foot long
axis, it takes about 80 seconds. That can seem forever. Of course
you would have gearing to make the table move as the speeds that fit
for accuracy and motor.
And, that brings up drive systems. if you have a rack and pinon, you
probably have a 1-1/2" drive gear, that is about 5 inches of travel
per revolution. since a stepper is usuallly 200 steps per rotation
(SPR) you have no resolution. you need to gear the motor to get your
1 mil resolution, or 1/4 mill resolution or 1 tenth resolution. you
need to add gearing. But, your table at 1 tenth resolution per step
is the same speed if it is a rack or screw. Since you are gearing,
the advantage of direct drive is lost. the complexity of gearing and
mounts and such come into play but is now equal for motors.
Considder a leadscrew. to move the table at 600 ipm with a 5 tpi
screw, you need to spin the screw at 3,000 rpm to get 600 ipm table
movement. advantage servo's
Considder that you can use a 800 count (steps per revolution) encoder,
or you can use a 400 or a 2,000. so you can change the motor's steps.
That screw with 5 rev per inch uses 1,000 steps on a 200 step motor,
or 10,000 if a 10x microstepper. if you considder that only 1/4 steps
are solid holding positions, then you have a great 1/4 mil resolution.
but you are now limited to speed as Mariss said. even if you can run
at 45khz, the limit for MACH, the motor power is not there, so a 45khz
is about the ceiling for current software. and YES, there are
choices, some to 200khz, but this is well past steppers.
but at 45khz, you get a max of 270 ipm rapids, but you are past the
max power of the stepper.
A servo would have an easy time running that screw at those speeds and
beyond.
The thrid thing is that of steps. when a stepper gets a pulse, it
moves a step (hopefully) when a servo gets a pulse, the driver looks
at where the position is and sends a pulse as needed, but...
When a servo is moving, it looks at where the motor is and how far off
position it is. It needs to be off position to get a pulse. as you
can imagine, at standstill, everythings is honkey-dory. as the force
of the cut starts to push it off, the driver sees the error and adds
power to hold it.
As a step comes in from the controlller, the position is off, becasue
the motor is where it was, and the driver now wants it to be somewhere
else. so, to some small or large degreee, the servo is always behind
some portion of a step or some steps, when it is in motion. This is
not all that big of a problem, and it is one small reason why one
takes a light finish cut.
Considder your high speed plasma/laser/router is cutting a large
square of thin material. racing towards a corner, the moving axis has
to come to a complete stop, and the stopped axis has to ramp up to
cutting speed at that corner. Now you have inertia and motor pulleys
and such to wreaak havoc on the machine.
Sorry for the long post. as with anything, there are compromises and
there are serious concerns. Also, with a properly designed unit, your
drive motors should fall into place as the one that fit best,
naturally. It is better to design the machine, and then see what is
the most logical choice.
Also, when the time comes to decide and you find yourself in that 200
to 3000 watt range and either could do the job, price come into play.
A high power stepper could cost $150.00, a driver $150 and a power
supply about $150, a 3 axis unit around $1,000 or so.
A low power stepper system could be $150 for a 3 axis board, $30 for
surplus and small stepers and $150 for the power supply. or $350 for
the machine.
For servos; it is still abut $150 for the power supply, maybe $20 for
a cheap treadmill motor or $250 for a high quality servo, and another
$50 for the encoders. so, a low of about $600 to a high of about
$1,000, maybe more if you add in some other feature cards.
When looking at a table top machine, I would be hard pressed to sell
you a $1,000 drive system when a $350 one will do everything the more
expensive one will.
My appologizes for the miss-spelling as I am riten from my wife's MAC
and I don't know how to do a speel cheak.
wrote:
>mind.
>
> >
> > Because they want to sell you a more expensive servo system.
> >
>
> Not true. If my machines would work quickly and reliably on steppers I
> would use them. Servos give me speed and more importantly peace of
> I know that even if something goes wrong the woest that will happen iswhen
> that the drives will fault out and let me know. Steppers would either
> miss steps or stall altogether without the controller knowing.
>
> > If the pitch of the screw, axis weight and choice of drive is correct,
> > there should be no need for a brake,
> >
>
> There is nothing wrong with a brake. Motors are fairly free-running
> power is removed. How else do you stop a heavy Z axis crashing intoit's
> endstops when power is removed?If you check the subject line, it is for table top machnes. Les
>
> Les
either has a really large table or he too missed the size the OP is
talking about.
If you are talking about a plasma machine with 600 ipm rapids, servos
are the way to go, you just cannot get a stepper to move the mass of a
heavy machine.
If you are talking a chineese desk top mini-milll (not a HAAS) a
stepper is the better choice.
I as I see it, there are 3 points. #1 servo's have encoders. that
provides feedback. Steppers do not have feedback. advantage servo's.
Yes, you can add an encoder. rarely done, and I don't think anyone
tries to use it as feedback for the control. I am sure some have,
but, it is not mainstream.
feedback allows one to know where you are. there is no way to weigh
the individuals comfot zone on hoping steppers are in the proper place
and have not missed a step.
On the table top machines, a properly designed and build machine can
use stepers. In fact, as Mariss pointed out, steppers can offer
better performance at some points. Mariss has offered that somewhere
around 200 watts of power is where steppers starts to look less
attractive and servos start to be a serious choice.
So, #2) is the power needed to drive the axis. Under 200 watts,
steppers usually have the advantage, over 300 watts, servos usually
have the advantage. I said usually, as there are advantages of
steppers, more power at low speeds. and of servo's, more power at rapids.
There are some FAST small machines that cry out for servo's, and some
slow large machines that work well with steppers.
One the speed issue. if you take said mini-mill and need to rapid the
long axis, you are moving all of about 12 inchs. at 100 inches a
minute, it takes about 10 seconds. Now, if you have a 12 foot long
axis, it takes about 80 seconds. That can seem forever. Of course
you would have gearing to make the table move as the speeds that fit
for accuracy and motor.
And, that brings up drive systems. if you have a rack and pinon, you
probably have a 1-1/2" drive gear, that is about 5 inches of travel
per revolution. since a stepper is usuallly 200 steps per rotation
(SPR) you have no resolution. you need to gear the motor to get your
1 mil resolution, or 1/4 mill resolution or 1 tenth resolution. you
need to add gearing. But, your table at 1 tenth resolution per step
is the same speed if it is a rack or screw. Since you are gearing,
the advantage of direct drive is lost. the complexity of gearing and
mounts and such come into play but is now equal for motors.
Considder a leadscrew. to move the table at 600 ipm with a 5 tpi
screw, you need to spin the screw at 3,000 rpm to get 600 ipm table
movement. advantage servo's
Considder that you can use a 800 count (steps per revolution) encoder,
or you can use a 400 or a 2,000. so you can change the motor's steps.
That screw with 5 rev per inch uses 1,000 steps on a 200 step motor,
or 10,000 if a 10x microstepper. if you considder that only 1/4 steps
are solid holding positions, then you have a great 1/4 mil resolution.
but you are now limited to speed as Mariss said. even if you can run
at 45khz, the limit for MACH, the motor power is not there, so a 45khz
is about the ceiling for current software. and YES, there are
choices, some to 200khz, but this is well past steppers.
but at 45khz, you get a max of 270 ipm rapids, but you are past the
max power of the stepper.
A servo would have an easy time running that screw at those speeds and
beyond.
The thrid thing is that of steps. when a stepper gets a pulse, it
moves a step (hopefully) when a servo gets a pulse, the driver looks
at where the position is and sends a pulse as needed, but...
When a servo is moving, it looks at where the motor is and how far off
position it is. It needs to be off position to get a pulse. as you
can imagine, at standstill, everythings is honkey-dory. as the force
of the cut starts to push it off, the driver sees the error and adds
power to hold it.
As a step comes in from the controlller, the position is off, becasue
the motor is where it was, and the driver now wants it to be somewhere
else. so, to some small or large degreee, the servo is always behind
some portion of a step or some steps, when it is in motion. This is
not all that big of a problem, and it is one small reason why one
takes a light finish cut.
Considder your high speed plasma/laser/router is cutting a large
square of thin material. racing towards a corner, the moving axis has
to come to a complete stop, and the stopped axis has to ramp up to
cutting speed at that corner. Now you have inertia and motor pulleys
and such to wreaak havoc on the machine.
Sorry for the long post. as with anything, there are compromises and
there are serious concerns. Also, with a properly designed unit, your
drive motors should fall into place as the one that fit best,
naturally. It is better to design the machine, and then see what is
the most logical choice.
Also, when the time comes to decide and you find yourself in that 200
to 3000 watt range and either could do the job, price come into play.
A high power stepper could cost $150.00, a driver $150 and a power
supply about $150, a 3 axis unit around $1,000 or so.
A low power stepper system could be $150 for a 3 axis board, $30 for
surplus and small stepers and $150 for the power supply. or $350 for
the machine.
For servos; it is still abut $150 for the power supply, maybe $20 for
a cheap treadmill motor or $250 for a high quality servo, and another
$50 for the encoders. so, a low of about $600 to a high of about
$1,000, maybe more if you add in some other feature cards.
When looking at a table top machine, I would be hard pressed to sell
you a $1,000 drive system when a $350 one will do everything the more
expensive one will.
My appologizes for the miss-spelling as I am riten from my wife's MAC
and I don't know how to do a speel cheak.
Discussion Thread
enytned1
2006-12-15 15:31:16 UTC
Servo controllers for table top machines.
Anders Wallin
2006-12-16 02:54:41 UTC
Re: [CAD_CAM_EDM_DRO] Servo controllers for table top machines.
caudlet
2006-12-16 08:09:16 UTC
Re: Servo controllers for table top machines.
enytned1
2006-12-16 15:50:12 UTC
Re: Servo controllers for table top machines.
Mariss Freimanis
2006-12-16 22:01:40 UTC
Re: Servo controllers for table top machines.
Leslie Newell
2006-12-17 02:18:06 UTC
Re: [CAD_CAM_EDM_DRO] Re: Servo controllers for table top machines.
turbulatordude
2006-12-17 08:35:24 UTC
Re: Servo controllers for table top machines.
Dan Mauch
2006-12-17 08:49:14 UTC
Re: [CAD_CAM_EDM_DRO] Re: Servo controllers for table top machines.
John Dammeyer
2006-12-17 15:42:35 UTC
RE: [CAD_CAM_EDM_DRO] Re: Servo controllers for table top machines.
enytned1
2006-12-17 16:58:27 UTC
Re: Servo controllers for table top machines.
Mariss Freimanis
2006-12-17 17:46:37 UTC
Re: Servo controllers for table top machines.
Steve Blackmore
2006-12-17 17:49:55 UTC
Re: [CAD_CAM_EDM_DRO] Re: Servo controllers for table top machines.
Leslie Newell
2006-12-18 01:57:44 UTC
Re: [CAD_CAM_EDM_DRO] Re: Servo controllers for table top machines.
Leslie Newell
2006-12-18 02:39:36 UTC
Re: [CAD_CAM_EDM_DRO] Re: Servo controllers for table top machines.
Polaraligned
2006-12-18 04:52:37 UTC
Re: Servo controllers for table top machines.
David Bloomfield
2006-12-18 05:28:56 UTC
Re: Servo controllers for table top machines.
Fred Smith
2006-12-18 05:56:36 UTC
Re: Servo controllers for table top machines.
enytned1
2006-12-18 08:07:40 UTC
Re: Servo controllers for table top machines.
turbulatordude
2006-12-18 08:55:03 UTC
Re: Servo controllers for table top machines. long
Fred Smith
2006-12-18 11:07:30 UTC
Re: Servo controllers for table top machines.
Leslie Newell
2006-12-18 11:51:32 UTC
Re: [CAD_CAM_EDM_DRO] Re: Servo controllers for table top machines.
Leslie Newell
2006-12-18 12:00:12 UTC
Re: [CAD_CAM_EDM_DRO] Re: Servo controllers for table top machines. long
Steve Blackmore
2006-12-18 15:26:29 UTC
Re: [CAD_CAM_EDM_DRO] Re: Servo controllers for table top machines.
Leslie Newell
2006-12-18 17:16:45 UTC
Re: [CAD_CAM_EDM_DRO] Re: Servo controllers for table top machines.
turbulatordude
2006-12-19 04:29:10 UTC
Re: Servo controllers for table top machines. long
William Carr
2006-12-26 07:05:51 UTC
Re: [CAD_CAM_EDM_DRO] Re: Servo controllers for table top machines.
lcdpublishing
2006-12-26 07:41:47 UTC
Re: Servo controllers for table top machines.
Jon Elson
2006-12-26 11:05:56 UTC
Re: [CAD_CAM_EDM_DRO] Re: Servo controllers for table top machines.