Servo Punch Press

Nice Jim! Really .great work! Give us an update when you figure out the noise problem...curious.
 
Great ideas guys. One thing that has plagued Galil controllers is a weak Ethernet link. I've never had a problem with a serial or USB link. I think they have fixed Ethernet in some of the later hardware, but this is a used card and and is a few years old.

I had this problem on another installation where I would periodically lose the Ethernet link, but that system had no motor controls anywhere near the control system, and the motor panel was optically isolated from the computer panel. The motor controls were just contactors, no electronics at all. I finally solved that problem by using the serial link.

I have the serial link working just fine now, a little software magic to get the timing right, and shorten up the message strings and it works.
 
Glad you got working even it is "Old School" :)
Will this limit the parts per minute enough (or at all) to be an issue?
 
Glad you got working even it is "Old School" :)
Will this limit the parts per minute enough (or at all) to be an issue?

Not at all, the computer just sends required pieces and the GO command, and the Galil takes it from there. The screen may not update the parts count in real time at higher speeds, but the Galil is keeping track of the required pieces and parts count, the screen reads the count periodically. About the most the screen count could be off is one piece when running, and would be correct at the end of the run.
 
Seriously impressive work Jim...!!

Regarding the Servos / Drives, i am assuming (bad move on my part) these are 3 phase motors with vfd drives which like all such systems generate copious amounts of electrical noise..

I would recommend using a separate enclosure for the drives with an ac input noise filter, as well as drive / servo recommended motor cabling (usually these are twisted conductors and may also be shielded etc..)

The motion controller and other power supplies would be in their own enclosure with ac input filter/s, ferrite chokes etc.. In some cases using a small isolation transformer / filters may be necessary ... especially since you have no control or what their power quality is like

Note that you also want to ensure your customer does not have issues with any of their other equipment and hopefully the use of a quality input line filter for the drives and careful attention to grounds / grounding should prevent that.

A video of full speed operation would be awesome... as would some detail of the control cabinet / wiring if that is ok to share

Thanks

Rich
 
Thank you everyone for the kind words. I will post a video of it running later, along with some pictures of the electrical cabinet.

Regarding electrical noise, I do have a couple of small line reactors that may get installed, we see how it goes. Once the electrics are inside the enclosure it may change the overall characteristics of the noise. The only thing the noise seems to affect is the Ethernet link. Nothing else is bothered by it, not even the serial comm link. The motors are permanent magnet, 3 phase, AC servos, and are inverter controlled. I have these same motors & drives in my lathe and have had no noise problems at all. But in the case of the lathe, the drives are in a separate cabinet from the computer hardware, and the controllers are PCI bus rather than an Ethernet link.

It's just about done, will be out of here in the next day or so.

Got the guarding nearly done, 1/2 inch polycarbonate. No place to get fingers into anything :) All of the works are enclosed in a box, just a couple of 0.400 x 1.500 slits to allow the material in and out. Computer screen is mounted on the guard, as is the switch panel (the white blob) above.

1540521602224.png
 
The control panel is nearly complete. I did it to myself again and didn't allow enough room on the panel so everything is a bit tight.

I was running the machine with the panel sitting on the bench with the servo cables strung across the floor and only the breakers and computer for control. It runs that way but I think maybe the customer would like something a bit more normal :) The safety systems are now in place, so the panel gets stuffed into the enclosure in the morning.

The stray wiring will be neatly rerouted and neatly tucked away before it goes into the enclosure.
1540693217040.png

I really hated to run the wiring down the outside of the guard, but there is no other practical way of doing it. Doing it like this and mounting the monitor and controls on the guard has the advantage of insuring that the guards stay on the machine.
1540693386320.png

When you have cables with big fixed ends on them, you need a way to pass the cable into the enclosure, they won't fit through any normal flex conduit, and I wasn't about to hang 2 inch flex on the side of this thing. So build some pass throughs. These capture and insulate the cables. I made these out of UHMW because I didn't have any Delrin on the shelf.
1540693729839.png


There is a set of these for each side of the enclosure, the control and computer cabling enters on the left side, and the servo cabling enters on the right side, near the drives.
1540693860048.png

I'll be test running this tomorrow so I'll get some video of it running.
 
Didn't quite get to the test run, but I did get the panel stuffed into the enclosure and completed the wiring.

One little item could have been a bit of of PITA to do. I needed to mount the 48V Galil power supply. The only mounting holes are on the bottom, M4-0.7, so how to transfer the holes to the panel without a lot of measuring and layout?
1540790911329.png

So here's what I came up with, just make a quick template

1540790990132.png

Tape the template to the enclosure in the correct location, center punch and drill.
1540791086285.png

Worked like a charm :)
1540791138225.png

I shouldn't have had to use a 48V PS for the Galil at all. It is clearly marked 24V, but for some reason it has a 36-72V DC to DC converter in it and it refused to run on 24V. The good news is that I had a 48V PS on the shelf.

And the completed control cabinet.
1540791235802.png

And the master switch. I still don't like the cables running down the outside, but not enough time to do anything else.
1540791296638.png
 
OK, the final post for this project. The machine is going on the pallet tonight, and will be at the freight terminal in the morning. It would have left yesterday but I had a bit of an issue with the drives faulting out on start up. Turns out that powering up the drives in the enabled (default) condition can make them do some weird stuff, so took a bit of rewiring to resolve that issue.

The operator panel
1540941713259.png

The spool rack
1540941653352.png

The material spool-off weight, adds some spring to the material so the feed doesn't have to accelerate the mass of the spool. Prevents miss-feeds from too much instantaneous tension.
1540941913053.png

Punch depth adjustment. Allows running different product punch patterns simply by adjusting the shut height. The punches are staggered in length.
1540942069344.png

And finally a video of it running.

This project is completed :dancing banana:
 
Last edited:
Back
Top