Anybody retrofitted a Roland PNC CAMM2 machine?

I've got a Roland PNC-2100 sat in the workshop, and it looks like a prime candidate for machining one off PCBs.

However I'm struggling to find much information about the steppers or spindle. Going by the wiring, they're definitely not using much current. I do have the original control box, but I was hoping I could find details without having to reverse engineer the original control to work out voltages/currents.


There is a model number on the spindle motor, but google is bringing up very few details, other than it appears to be a Canon Precision FN38 series, but it would appear to be a special version, as the full number isn't bringing up any useful info (it's FN38-H216N1E, with CN38-13101, and 6FKZ in separate boxes if anybody wants to try finding info on it).
And I can't see any model numbers on the steppers with them installed.

Isn't the spindle motor on that model DC? If so a variable voltage bench PSU and a tacho should make short work of determining the voltage/RPM slope.
I might be tempted to connect an Arduino with a RAMPS to run a stepper and see how well that works ?

Yes, DC motor, but I've not got a tacho. Could be an excuse to buy one though...

I looked at it again last night, and powered the control box up. It has one 36VDC supply, but I can't get into the other smoothing capacitor (it's rated at 25VDC) to see what the other supply is without some major dismantling. The feed for the motor appears to go to the speed control board on the machine, then down to the motor, and the motor has 3 wires (I'm going to gues possibly some kind of thermal cutout…).

It does seem to power up OK, but the motor switch does nothing, so I'm guessing it's also controlled by the controller.
I'm going to have a rummage and see if I can find an old XP computer, and see if it will run.


I'm still undecided what to retrofit it with. Arduino/GRBL is the cheapest option, but I am tempted by a DDSCV, as it seems a bit less fiddly to use and comes with an interface.

Just done a little bit more probing while waiting for a job to finish.

There are 6 TEA3718S chips that obviously power the stepper motors, which receive 36VDC.

The other big smoothing capacitor measures at 12VDC.

One half of the motor switch controls a 5V feed from the control board, and the other is switching 24VDC. I never checked but one side seems to come/go via a TO package bolted to the main heatsink. It's possibly on the supply side, and is what is providing the 24VDC from the 36VDC.
There are a couple other chips bolted to the heatsink, which at first glance I thought were stepper driver chips given the number of pins, but given the row of TEA3718S's I'll need to have another look.

Just been given one of these, a bit tatty. Had exactly the same thoughts about pcb milling and for that pcbgcode and linuxcnc would be an an obvious choice but I am also up for selling as is if someone need a donor machine, It moves and homes but the spindle does not come on from the panel buttons and the spindle drive belt is broken. I think GRBL would probably work well but it also seems like a very fiddly job to overlay a retrofit onto Rolands design