DC Capacitive prox and a VFD

[olias]

I have a project I'm working on that has two Cutler Hammer VFDs (1/2 hp, and 2hp), a Micrologix 1000, a couple of 24vdc inductive proxs and one 24vdc capacitive prox. Everything is fine untill I turn on the 2 hp drive. Then the cap. prox goes on. I tried taking another prox and when I just wired the brown and blue leads to the power supply, left the load wire hanging, and turned the drive on, the same thing happened. The motor leads go out one side of the panel and the proxs go out the other. And I'm already using an isolated power supply. Any suggestions? Thanks.

Olias

 

[BTalbot]

how close is the capacitive prox to the motor leads? can you re-route them? is there a shield for the motor leads? this probably wont help but try changing the swithing freq.

 

[olias]

The leads are as far apart as I can get them. The prox is about 3 feet away from the motor. And I tried the freq. change with no luck. What else?

 

[BTalbot]

Not sure but different mfrs use different technology on capacitve proxes some use compensation electrodes internally (usually for temperature) some have no compensation. is this a low cost sensor? maybe a different mfr sensor would not be as sensitive to the motor noise.
Disclaimer:
I work for a Turck distributor, I will hook up a Turck cap sensor on my lab bench and put it next to a Control Tech VFD and post my results. Not trying to sell anything just trying to help and I'm very curious.

 

[olias]

I was thinking the same thing. C3 controls PN CPS-N18TCUZCP10.
Let me know what you find.

 

[BTalbot]

I used a 230 vac vfd certainly a 460 v would be a bigger culprit. The Turck cap sensor worked as expected 10mm range within an inch of the vfd, motor leads and motor. Have to hit he road now if need be I could hook up a 460 vfd maybe tommorrow. Good luck!

 

[olias]

Looks like it time to try Turck. I'll let you know how it works out. BTW, VFD voltage is 230v.

 

[OkiePC]

The culprit isn't necessarily the drive voltage, but the carrier frequency (4KHz or higher usually). I have measured induced voltages of 110vac with my fluke on a line I knew was opened at both ends. It was routed alongside VFD cables for thirty feet. When I grounded the wire to make this induced current go away, it produced a 1/2" spark!

In another case I had a (shielded and drained) 0-10vdc analog signal that had tiny little fifty volt spikes at 4Khz on it when you looked at it on a scope. My input card filtered out some of it, but my readings were off by a little bit only when a certain VFD motor was running.

My advice is to ALWAYS use shielded motor cable with ac variable speed drives. Stop the noise at it's source. You'll have fewer problems with analog I/O, communication calbes, and sensitive sensors in general.

Try to look at your isolated power supply input and output with a good scope. You may be getting noise onto your DC output and the capacitve prox is the only device that isn't tolerant of it.

 

[olias]

Screw it! I'll go with a light beam. I just need to sense the presence of a piece of cardboard. Thanks all!

 

[olias]

Got a call today from C3. They confirmed my problem and stated it is an EMI problem. They instucted me to wire the 24v-(blue) wire of the prox to ground. Worked! Now why didn't I think of that? Thanks all.