Howdy!
I am in the middle of a conversion of a Mid- 90's custom CNC router. This was designed for prototype PC Board routing and drilling, along with front panel engraving. The system originally DOS PC driving a PMC DC2-PC control card and basic PA10 hi-power opamp drivers for the X & Y axis DC Servos. It has a IMS IB462 Stepper driver for the Z-axis, which I may update as well. Suggestions for that? I am moving to a CSMIO/IP-M controller on DIN rail with a cleaner power setup. It does have a nice AC power control unit that I am working through to re-use for all MAINS control for the various supplies, etc.
It will also handle a lot of the E-Stop shutdown functions.

The Spindle drive is run by a LUST VF223 AC Drive (runs on 220V thru a step up transformer). I was able to track them down and get a manual for this, so I may be able to use it, but may also use a 1/2 size router for wood cutting/carving projects. This was running when unplugged many years ago and only have well less then 100 hours on it then, so inside the acoustic cover it is like new condition, just a few years old and older technology.

My first issue is in the CSMIO IP doc on Differential outputs, Section 2.1.2 (interfacing to optocoupler inputs) is close, but I would like a bit more protection from noise. My thought is to run Twisted-pair between the two units for STEP and DIR signals, using DIR+ and STEP+, leaving the low side outputs open as called for in the doc. I would use the hi side of the signal as shown, but the other twisted mate would wire to Pin 2 of the Analog I/O 8 connector, as covered in the doc. On the Gecko side, The STEP+ would go to Pin 9 and DIR+ to Pin 8. The two low side wires would be tied to the common Pin10. This would twist the low side with each signal but wire them as shown in the doc on page 5.  This should help to reduce noise in the signals. Correct?

Also, on the G320X, I plan interface to existing X & Y axis RSF Elektronik Linear Encoders ( MSA 6708 ML 670mm, 5V) which have differential outputs. The G320X is single-ended. I will most likely just put a differential receiver in between to convert the signal. This look to be too good just to toss away.

Just looking to conformation that I am headed in the right direction as I do not want to damage the controller before I even get started.

Regards,
Don
Katy, Tex.

Not sure if that would make any difference, I was under the impression that the differential signalling worked because there was an opposite of every pulse and as such if there was noise it would be cancelled out.
I think I would fit a receiver to just before the Gecko and that way be able to use the diff output.
Hood

Yes, Hood, that would be best. I did find on the CNC4PC site both a driver and receiver boards for RJ45 connection. I just ordered some of the Receivers to fix my encoder output conversion to single-ended on the input of the G320X. Should have ordered a few more. Just was not thinking about the CSMIO side for that issue, but would work.
The G320X's will be ~1-1.4M from the CSMIO controller, so trying to head off any issues with noise. The RJ45 cabling looks like a clean approach and will give me twisted pair all the way to the receivers, then a short jump to the G320X and all should be good. Will need to wath the grounding but plan to have ONE common ground point for the whole system.

thanx

Don

I discovered after poking around on the CNC4PC site, some hardware made for the G320(X) drivers. A C34G320 mates directly with the G320X and provides both of the rec'vr and transmitter pairs needed for the Driver input from the controller and from the Encoder's diffirential outputs. PCA's are small, using surface mount parts. These all use standard Cat5e or Cat6 RJ45 cables, so you get 4 twisted pair lines, running off a small 5V supply. Really cleans up the connections. The driver it a CG6, and the Driver/reciever pair is a C45,. The C45 manual show all types in use, so good reference. Cost is cheaper then I can make a PCB and much cleaner. The boards run off of a 5V supply, so will need to check spec's on the CSMIO/IP-M side. I could see using a USB Hub to distrigut the power for these, if I need to, but less connectors is always a better solution in the long run. It seems to require 24V to power it's outputs, but it may be 5V - 24V range rather then 24V only. I hope. Will check with CSMIO directly on that. Too many warnings in the manual about blowing drivers. I did find a typo in the manual, page 21, Blue Dot on the bottom of the page references to connect power for the Digital Output drivers to pins 9 and 23 (V and Gnd), but the conector pin out show GND to be pin 22. Pin 23 is label Output 1 for the relay contacts.
The CNC4PC site is a bit tough to find the manuals, etc for the these, but scroll down and Save the target to a file. Clicking on the MANUL link got me blank pages. I will say CNC4PC was quick to ship and they are on their way in hours. This will save me time and resultin a very clear installation. All twisted pair diff signals for these control lines.

At any rate....Progress!

I made up line driver boards for my machines when I was using controllers that could only output single ended signals. First few were just routed copper clad then latterly I got a board produced and populated it. Now however I am starting to use the CSMIO controllers and they have the Diff out so I no longer need to do such thing

Hood

YEs, I like Diff lines much better, but my old hardware/electronics has a mix, including the G320X. IMy day job is a world of twisted -pair and sometime shield cables for low freq analoge, so am use to working with them. The CNC4PC boards are too small and low cost for me to not go that route, but there was a day when I built my own. Just wish the IP-M unit had a few more inputs.

Don