sinking or sourcing remote enable kollmorgen drive

Hi Len, I wanted to let you know I ‘ve seen your post and when I get home this evening I’ll look over your attachment and make a write up my project so you know where I’m at on my integration.

So glad to know there’s another out there sharing the same unknowns. Servo Star CD drives are discontinued so I don’t think we will get much in the way of factory support, I did however find the Kollmorgen / Danaher repository for this product line. It looks to have the whole history of support papers and firmware. I’ll make sure to get that link to if you haven’t already found it.

-Doug

Doug,

Please send the link, but I think I may have already been there. Although this series is disconnected, when I emailed them two years ago, they were happy to share the download link to Motionlink and were quick to say give them a call if I had trouble getting things to work. I remember being pleasantly surprised at the support level for their older product line. Granted, they still sell a newer version that is very similar.

In the near future, I may want to pick your brain on your experience with Motionlink if possible.

Len

Hi Len, My project is a CNC retrofit for my bridgeport style knee mill using the Kollmorgen drives, motors and cables that were left over from a previous project from …well…nearly 20 years ago. On the linux cnc side of things I’m using a intel atom single board motherboard, mesa 5i25 with 7i77 (analog). My goal is to mimic a Trak mill from Southwest industries, those mills are the best in-between mills I’ve ever used mostly because they still have handles and dials on them so you can quickly use them for dirt simple things without any programing hassles but can also run programs and even 3-axis cnc machining. It took a lot of time to design and build a motor and timing belt interface so that I can keep the dials and handles, well different handles that is.

The Kollmorgen motors and drives I have are resolver based but the drives came equipped with an encoder equivalent output on C4 that is connected to the 7i77 encoder input. the resolver from the motor go to C2 on the drive. It seems to work fine near as I can tell so long as the parmeters in the drive are set right and the encoder scaling is correct in Pncconfig. The mill is in one location and the computer and panel with a single motor and drive are at home where I can work on it in relative peace and comfort. I verified the resolver to encoder equivalent counts and timing belt-leadscrew ratio on the x-axis prior to having linuxcnc fully configured using the Motion link software.

Man, I wish there was a Kollmorgen forum similar to the wonderful forum here at Linux CNC. Although I was deeply involved with the these motors and drives before, that was twenty years ago! Fortunately I still have the laptop from back then with a functioning serial port. I’ve not been able to get the USB to RS-232 cable & driver to work with my current laptop, I can see it transmitting but the drive never replies back. The old laptop running windows 2000 and Motionlink from that time period works just fine. I have found almost nothing to help with using motion link, takes a lot of time to muddle through it. I hit the estop/disable button a lot. To enable the drive I wired my estop using the normally closed side in-between pin 8 on C3 and -24V on the PS. +24V to pin 7 on C3. This weekend I’m working on getting the drives and PC next to the mill and start routing the cables and limit switch wires.

Here’s the Link to the Servo Star CD repository.

www.kollmorgen.com/en-us/developer-netwo...D=4176&title=&page=1

Good to hear Kollmorgen gave you the time of day when you called in. I expected it to be more of the same “we don’t support that version anymore”.

Your machine is using linear motors. One of the axes I delt with back in the day was a linear type. Man when those things go unstable its BIG deal. The one I had hit the hard stops in both directions a couple of times before I could hit e-stop. To be fair the travel wasn’t all that long.

Doug

Doug wrote:

Your machine is using linear motors. One of the axes I delt with back in the day was a linear type. Man when those things go unstable its BIG deal. The one I had hit the hard stops in both directions a couple of times before I could hit e-stop. To be fair the travel wasn’t all that long.

Doug

Doug,

This has been one of my concerns during startup/testing. My X travel is approx 14 feet! That's a 30lb (guess) carriage moving 14 feet with 142ft lbs of torque. Could get lively. In the Kollmorgen docs, they suggest using 2x4s to block both sides of the carriage to limit the overall travel to 6" or so until you verify you have all wired/configured properly and have good control. Can't get much momentum in 6".

My end stops have beefy mechanical stops of course, but also Allen Bradley switches that are wired in series with the Estop circuit. The Estop circuit kills the 3 phases to the servo drive. Note: These switches are not the limit inputs into LCNC, those are proximity sensors that will be inside in working envelope of the ABs. The ABs are the backup. Perhaps a bit of overkill, but will make me happy. I will probably wire up a temporary switch to the drive enable to use while initial testing like you were doing, as hitting Estop will kill all power to the drive and Motionlink will have to reconnect, ect.

Thanks for the link and the writeup on how your system is configured. It sounds like you will have an nifty mill setup when finished. I like that you are keeping it semi-manual. I have always envisioned adding cnc to one of my old SouthBend lathes (1906 vintage) in such a way that a machinist from 1906 wouldn't know by looking at the lathe that it was CNC. Only when grabbing the handles, would they then realize that it was drive by wire. I even was going to have the screen hidden behind an appropriate vintage styled cast iron case so it wouldn't be obvious....and no power switch. Once the overhead line shaft starts the belt turning, it would trigger the power to come alive. Kinda silly to spend the time on them, but I love the old SBs. I have 6 or 7 of them. Andy on here has done a nice lathe conversion, I believe.

Len

superlen wrote:

My end stops have beefy mechanical stops of course, but also Allen Bradley switches that are wired in series with the Estop circuit. The Estop circuit kills the 3 phases to the servo drive. Note: These switches are not the limit inputs into LCNC, those are proximity sensors that will be inside in working envelope of the ABs. The ABs are the backup. Perhaps a bit of overkill, but will make me happy. I will probably wire up a temporary switch to the drive enable to use while initial testing like you were doing, as hitting Estop will kill all power to the drive and Motionlink will have to reconnect, ect.



Len

Hi Len, There's nothing wrong with having redundant E-stops when using Linear motors. Its a really good Idea to limit the travel with blocks. On my previous use of LMotors the instability was really strange. We overcame it by swapping the motor phase connections on the drive and it suddenly worked. We did this as a last resort only after double and triple checking all the encoder feedback connections. You will want to wire the estop switch on the drive enable especially when tuning with motion link. If you lose power before saving to the eeprom you will lose all your hard earned settings and have to try and recall from your own memory all the parameters you adjusted through the many tries.

-Doug

" If you lose power before saving to the eeprom you will lose all your hard earned settings and have to try and recall from your own memory all the parameters you adjusted through the many tries."

I hadn't thought of that. Yes, that would suck. Hopefully this weekend I can get one of the axis hooked up. I still have a lot of wiring and box brackets/holes/cable routing to finish up so I may not get it done, but I'll at least get farther along.

Len