Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: alsetn on November 29, 2012, 01:38:14 AM
-
I will be moving a small object fast back and forth on an 80 ft. rail using one large stepper motor/driver and cable. I need to program a smooth acceleration / deceleration profile evenly over the whole length.
Can Mach3 run this?
TIA !
-
It depends on about 100 different things.
(;-) TP
-
OK BR549, let's narrow it down. Step resolution is not too important, high speed is. If the motor has about 1200 oz in. torque maybe I can drive an 8 dia drive wheel and idler wheel to move the cable at a high linear speed.
Can I program exact number of steps for the whole length and plot them along their curves? Can Mach3 or another program compute the curve?
-
You have the Motion planner and the accelleration and velocity settings that are a trap curve and that is it. IF you plan to wind it up on a drum then you also have variable diameter as the cable winds onto the drum. Without positive engagement you can have cable slippage on the sheave.
Steppers are NOT known for their high speed. They are low speed motors. Low resolution to gain speed can drive you into the cogging problem and step loss when driving them beyond their torque curve.
Servos have a much wider operational rpm band.
I think you may want to take a look at a good VFD and motor drive setup.
How fast and what accelleration rates are you talking about ?
What is the accuracy needed for the stop points ?
(;-) TP
-
I'd use one of these to do what you want.
http://motion.schneider-electric.com/products/mdriveplus_overview.html?gclid=CL2Q7aCr9LMCFSemPAodJE8AlA
You can write a program and download it to the unit.
Self contained just need a Power Supply and a means to trigger input the program to start and stop. One trigger to send it home, trigger to start, trigger to stop and what ever you can think of. You can change/set the accel/decel and speed during the program run. You just write a text program and then send it to the processor. Different programs can be run depending on the trigger that is sensed. Great for repeated routines. Fun units to work with too. But Servos are so much faster and more powerful, Parker makes a GV Drive that does this also to control Servos that I have used.
-
The new Gecko G215 can do this on it's own, but it hasn't been released yet.
-
Thanks Gerry for the heads up
-
I think you are right about servos vs. steppers. I will plan for that.
0 to 80ft. in .5 to1 sec. moving object about 4oz. mass. Acceleration gentle and decel the same. But the veloc. profile needs to be sinusoidal so at midpoint (40ft) accel ceases and decel starts. Decel could be triggered by a switch at midpoint.
The minor amount of slip can be corrected for and zeroed each cycle by switches along the slide and a home switch.
You mentioned the Gecko G215 as a possible dedicated controller without need for a PC. Any other options off the shelf?
The big unknown for me is how to program the specific velocity profile.
Thanks all.
-
80 foot in less than a sec ? That is not going to happen with Mach3.
There is NOTHING gentle about that motion curve (;-)
(;-) TP
-
"gentle" meaning zero velocity at each end during reversal.
"...That is not going to happen with Mach3..."
Can Mach3 be tricked into thinking the move's more conventional by using normal speed in the instructions but multiplying the motor steps in the driver settings to increase actual distance traveled?
-
I think you would be better served with another solution.
Just a thought, (;-) TP
-
Is my math right? Off the top of my head, that seems to be somewhere upwards of 75,000 inches/minute? Accelerating at something like 600G? As 1G would get you about 125" in 1/2 second.
-
80feet per second = 960inches * 60 seconds = 57600 inches per minute
80 feet/sec = 4800 feet/minute * 12 inches per foot = 57600
54.54 mph
80 feet /sec = 2.49 G
Based on my Convert Program
-
Black Powder for the accel. ....... and a Bale of Hay for the decel. ;D
Not exactly sinusoidal, but it would be hard to see the difference.
Russ
:)
-
Assuming my math is correct:
80 f/sec = 960 in/sec
5000 rpm = 83.33333333 rev per sec
Assuming a 4 in dia pulley 12.5664 inch circumference
1047.2 in/sec
87.26666667 ft/sec
-------------------------------------------------------------------------------------
80 f/sec = 960 in/sec
7500 rpm = 125 rev per sec
Assuming a 2.5 in dia pulley 7.854 inch circumference
981.75 in/sec
81.8125 ft/sec
Good Luck
-
You have to accel AND decel to a stop in that one second, so you need to go much faster than the 57600ipm. I guessed at 75,000. Accel goes up as well.
-
No doubt, Send us a video clip on first test. LOL
-
Not sure why this wasn't more clear yesterday, but if you're going half the ditance, the velocity just doubles, if accel is linear. So, it's 115,200ipm, accelerating at 96G. He wants an s-curve accel, so it will actually be higher than 96G, I've seen a linear stage accelerating at 3-4G, and it's so fast you almost can't see it move.
-
Guess I am lost by how you are getting the 96G's
115200ipm = 160fps = 4.973g from my convert program.
Help out an ole toolmaker. LOL
-
The online calculator I was using was broken (sort of. ;) )
Your right, but you have to double it to get to 115200 ipm in 0.5 seconds. So it's a little less than 10G. Still pretty damn fast.
-
Hmm.........
0-60 in 3 sec, head yanked back and ask put into seat, cop in back of me, slam on brakes, 60 to 10 in about the same time,
head touches stearing wheel, pull into parking lot, cops says why, said I was sitting at light ......what the heck just need some excitement in my life, he says, I believe you because you would need to be the dumbest old fart in the county to do something like that with a cop in back of you, get warning and lecture about acting my age.
Your accel and velocity, better to just envison it ;)
RICH
-
Too funny, Rich