MPG knob inputs

[cwatson1982]

Hi,

I picked up one of the typical cheap 6 axis MPG's. I wired the A/B signals to my kanalog and wired the rest of it to JP6 per the various examples out there.

My issue is that the selectors bits are all randomly bouncing high causing multiple axis to attempt to move and jerky motion. With just the A/B signals and power connected I get smooth motion (hard coded the C program to just move the 0 axis channel).

The cable run is ~20 feet from the back of the machine to the operator console. I have tried the original twisted pair shielded cable that was run through the machine for the original VGA and touch screen connections. I have also tried using an SFTP cable. Same results. With a multimeter on the selector outputs I get .02-.04 volts when they are inactive. 3.3V when active. When disconnected from kflop they are 0v when inactive and 3.3v when active.

Kflop/kanalog/konnect are powered by the same PC power supply that runs the computer. I tried an isolated power supply and had the same result.

Any simple way to solve this? I thought about trying to drive the selector switches at 5v and attaching them to kanalog opto in instead of JP6 but would rather not if I can avoid it.

[cwatson1982]

I tentatively might have found the cause, if not the solution. My e-stop circuit is 24v, removing the MPG e-stop switch from the circuit seems to have stopped the selector switch bits from going crazy. I have another switch mounted to the console so if I can't find the underlying issue I may just not use the additional e-stop button on the MPG.

[TomKerekes]

Hi Watson,

idk. You would have to include more specifics on the wiring and grounding for us to help.

You might try adding filter capacitors (0.1uf Ceramic capacitors to KFLOP GND, close to KFLOP) to suppress noise glitches.

[cwatson1982]

The machines existing wiring is still all in place as it was before the retrofit (other than the operator console and pendant wiring). Everything is star grounded. The kmotion products and the PC are on their own 120v circuit that is completely separate from the machine circuit (other than the mounting plate-board-cabinet mounting screws and standoffs)

I was incorrect yesterday. It is not the e-stop circuit but noise from the VFD/Spindle. As soon as I enable it, it goes haywire.. Running a wire from the Kflop ground to the cable shielding helps but the bits for the rotary switch are still activating constantly, too fast for the I/O screen to show. Checking via a C program and outputting to the console shows constant switching.

[TomKerekes]

Hi Watson,

Running 3V signals 20 feet across a noisy machine is likely to pick up noise. Twisted pairs don't really help if not differential signals.

You might try adding the Filter Capacitors as suggested.

Or yes higher voltage signals and optos will be less noise sensitive.

[cwatson1982]

I switched the MPG control switches to 5v and over to the kanalog opto inputs. Everything is stable and functioning properly now. Even was able to use the old pre-run shielded cable and a distribution block I had laying around for grounds. Next will be pots for speed over rides and mounting it all...if my aluminum for the panel ever shows up!

The spindle drive or motor must put out an awful lot of noise. I'm not sure what the voltage for the old MPG was but guessing it probably wasn't 3.3v.

If I end up needing the kanalog opto ins in the future I'll pick up a couple of 4 channel 5v to 3.3v optocouplers and go back to JP6.

Thanks as always, don't know how you do it every day!