What's new
What's new

Motion Picture Fanuc m-16i

jermfab

Cast Iron
Joined
Jul 25, 2013
Location
atlanta, ga
Hello all... I’m Jeremy, I do special effects for the motion picture industry.
I got a doozy dropped off at the shop today. A pair of Fanuc industrial robots, possibly 2 of 4 in total.
53d330bcc1b3f59d544f0e7267e8b771.jpg

da49b6e0a20724810f48ca20b153bd5b.jpg

23b405ef1bc4c71717a6c6699d3fc5e7.jpg

The desire of the producers is to automate these machines, on camera... I’ve been clear that is likely far beyond my abilities. I don’t have the controller... I don’t have the cables and no one is even sure whether these machines even worked prior to the film procuring them.

I’m fairly certain I can disengage the stepper motors and “pose” the arms. I’m dead positive that won’t be enough for the people paying me.

This is way out of my wheelhouse.

Maybe there’s someone on this forum who can point me in a direction?

I’ve got a week.




Be safe and stay healthy





Jeremy
 
You can, at best, remove the red caps and turn the motors by hand.

The mechanical part of the machine is half, at most, of the robotic system.
 
I misspoke. They aren’t stepper motors, I’m not even sure if they’d be considered servo motors either. In any case I think I have a solution.

Elegant it is not, but I’m pretty sure I have a solution that meets my needs and can be accomplished in the allotted time.

The motors are 161V, 200Hz, three-phase motors. So my plan is to use VFDs and limit switches. Thankfully this approach also limits the amount of work I’ll have to do. The package per axis is about $250-300. It’s worth $1000 to the producers to see the arms move, so they get to pick any three axes.

The Drives and switches should be in tomorrow, I’ll follow up if my plan is a success.





Be safe and stay healthy



Jeremy
 
Jeremy . . . not to rain on your parade, but unless you have set up a brushless servo motor with a VFD before, you better have a good understanding of Ke, Kt, flux density, and brake management based on torque proving. At best you will be really frustrated, at worst, you are going to have a pile of smoking electronics.

I have done this kind of work with a company in Burbank called “Camera Support” . . . they have some sharp folks who can assist, but a grand may not even get them to hop in the car to come see you.

Camera Support
 
You can’t rain on my parade any more than my bosses ALWAYS do. I’m sure there’s a world I’m missing. The VFDs I have on order were recommended to me by a couple of guys at Fanuc.
They’re RS-232 units, which should be capable of managing the brakes, though I’ll likely have to use that output to trip a relay and provide the 24v DC for the brakes.
This is quick and dirty stuff. I need these things to move a handful of times and then they’ll head back to whatever scrap yard they came from.
I’ll take any advice, but after a couple hours on the phone yesterday with the two Fanuc techs I feel *fairly* confident my plan will work. At least for a while.
It ain’t perfect. It ain’t elegant. It most definitely ain’t the type of motion control these machines were designed to have... but all I need to do is get a slight amount of movement from three axes and NOT knock some actors head off.




Be safe and stay healthy




Jeremy
 
... all I need to do is get a slight amount of movement from three axes and NOT knock some actors head off.
Actually, that is a real-world danger. Industrial robots are often - usually ? - fenced off so that no one can get near them in operation. If you aren't used to industrial equipment, I'll just point out that this stuff kills. And it doesn't even say it's sorry afterwards.
 
Pay the shipping and I’ll put them in a box and send them to you... if the arms actually go to scrap and I’m able to. There’s a chance they go into studio holding.

You are correct about the brakes as well. From what I can tell, the brake takes a 24V DC signal to disengage. Easily done with the VFD RS-232 brake output and most likely a relay.

As to safety... yes. It’s a major concern. Another reason to use the VFDs. I will set an E-stop for all three drives. Safety is ALWAYS my first concern. Especially with people who are insured for more than I’ll make in my lifetime. I did most of the design work and my crew and I built the Gremlin Tom Cruise chases in “American Made”. Talk about rigamarole. They wanted the effect of a satchel-charge under the drivers seat... of a gremlin... with Tom Cruise running behind it. The whole gag was done pneumatically; hydraulic cylinders, powered by high-pressure nitrogen did all of the work. The only pyrotechnic device was 18” or so of cannon fuse to put some smoke in the car.

Regarding safety from the robots... yes. I was fully aware even before my new friend at Fanuc reminded me: “these robots are ALWAYS fully separated from their human overlords by cages...” Funny enough, my first industrial “big-boy job” was as a welder/fabricator making component machine guards in an injection molding plant. I see the safety issue as another reason to use the VFDs. I can vary both frequency and voltage to slow the motors down from their peak parameter. I expect to wind up supplying the motors with less than the 161V and 200Hz they expect.

As I’m bypassing the positioning elements and driving the motors directly I figure as long as I don’t deviate too far from the motor-plate parameters I shouldn’t let magic smoke out. At least not immediately. I already know the magic smoke doesn’t come out with suicide wiring into delta-wye, spike-leg, 208V, 60Hz shop power.

The biggest difference between my current world and that of most of the members of this forum is the impermanence of my world. My job is to make these things move, safely, a maximum of thirty times, plus whatever I do to gain confidence to present that much. Worst case for me is I do burn up my spare motors and have to punt. Punting will be reducing drag and dropping weight as best I can and using cylinders to make the move. Then I pose the rest of the arms to suit camera.

The goal is noticeable rotation plus a knuckle movement on one robot and the knuckle moving on the second. 2”-3” bore, 6”-12” stroke cylinders will accomplish that goal. The casting is plenty thick enough to just drill and tap for a mount for a rod-clevis.
I’ve got a good selection of “Bimba” low-pressure cylinders, likely more than enough for my purposes here. I’ve got fittings, solenoids, shuttle valves and the like. I should have the VFDs and switches tomorrow. Biggest issue I have is time. I’m gonna give myself tomorrow to play with the motors, first just on the bench and then I plan to start with the rotary drive at the machines base. I figure I’ll lock the rest of the arm off, sticking straight up, and prove my theory with that axis. You know, make sure the magic smoke stays in with a load and that the limit switches will work fast enough I don’t tear them off. Once I’m comfortable there I’ll move on to the knuckle pivoting.

If I burn up a VFD, I’ll punt. If I smoke a motor immediately, I’ll punt. If I can’t slow the motion down enough and I’m worried the unit will overwhelm the limit switches... I’ll punt.

As long as I have SOMETHING, I’m good. And the project gave me the excuse to add a couple of decent VFDs to my inventory.

Here’s a couple fun ones from a car gag early last year:
59d9533601fc92ac68293d147e1e4213.jpg

cce5c8c9c82b18b50261734e68145350.jpg

Pyrotechnic cannon, hence all the fire and smoke. If memory serves, we got the car on a Monday or Tuesday and I was cutting the cannon out LATE the next Thursday.





Be safe and stay healthy





Jeremy
 
As to safety... yes. It’s a major concern. Another reason to use the VFDs. I will set an E-stop for all three drives.

It might be a better fail-safe to use a deadman's switch (I guess they are called enabling devices now). It is quicker to open your hand/finger to release a momentary switch than to push an e-stop when something goes wrong. It can be as simple as putting a normally open momentary switch on the 24V brake release signal.
 








 
Back
Top