Sign in to follow this  
Followers 0
splicer480

871TM Inductive Prox Sensor Spike?

5 posts in this topic

Hello All, I have an automated cutting machine that has a turntable that starts at 0 degrees (Home) and rotates a full 360 degrees (Back to Home) to cut off the top of a plastic drum. I have an intermittent problem with just (1) of the (7) 871TM proximity sensors on the machine. The problem sensor is my (Home position) sensor. It tracks a bolt in the turntable carriage as a start and stop position. When it senses the bolt it is about a 1/2" from the sensor head and sends the 120 V AC back to I:0/2 of the micrologix 1000 plc. Any other time the turntable is turning, the sensor is at least 4" away from any metal surface. I'm getting a very fast intermittent flash or signal back to the plc from the sensor while it is looking for its home position causing the machine to think the cycle is done and shutting off too early. I changed the sensor and even reversed polarity and still the same thing happens. This machine is about 25 feet away from a 300 HP DC drive blow molding machine. The operators say when they shut down the blow molding machine for maintenance, the cutting machine doesn't miss a beat. When they turn the blow molder back on every once and a while the cutting machine will stop cutting too early. Do you think it is picking up "noise" from the DC drive (Flex pack 3000) or DC motor? Is there a diode of surge suppressor I could series in with the input I:0/2, or should I purchase a weld immune or transient noise protection version of the inductive prox? Any help is appreciated.

Share this post


Link to post
Share on other sites
If you can get ahold of an oscilloscope and record the behavior of your home input it definitely sounds like noise of some sort. I like the idea of a noise immune prox, but wonder if the problem isn't RFI into the wiring and not the prox.

Share this post


Link to post
Share on other sites
Those DC drives are extremely noisy. It's a 6-pulse arrangement even though it is DC. The input waveform looks horrendous. I know...I have 4 of them and crosstalk even among the drives is really bad. I'd highly suggest you upgrade to an AC system if you can justify it anyway. You didn't say if you are running 24VDC or 120VAC wiring. If it's 24VDC wiring, then your problem is probably that you are getting some sort of power problem such as a voltage droop or your negative is drifting in response to something else. Make sure that the negative (or neutral on 120VAC for that matter) and the PLC power are tied together and isolated electrically from all of your contactors, starters, etc. At least that way they are operating on a common reference (negative or neutral). If you have a choice, try to switch over to 24VDC anyways even if it means putting a small 24VDC/120VAC output module right in front of your PLC input. 24VDC signals have much faster switching times than AC ones, they are virtually immune to shorting them out, safer to work on (no shock hazard), and they are virtually immune to crosstalk type issues regardless of cable length since among other things, you need an AC signal to create crosstalk in the first place or very high speed DC ones simulating AC. If it's 120VAC (24VDC might have this issue but it is much more rare), then to start with, you may have either a ground loop or crosstalk through an electric field somewhere. If you are running your signal wiring with your power wiring in close proximity (in the same conduit), that's your most likely culprit. If you grounded your conduit or whatever at both ends, that's another likely culprit. If you can't easily eliminate the problem, run shielded multiconductor wiring and ground ONE END ONLY to the same ground as the PLC chassis (again...no tying drives, contactors, etc., to the same ground plane). If you tie both ends down, you create a loop antenna and couple noise from any magnetic fields into the system. Twisted wiring also helps eliminate electric fields because every other twist tends to cancel any induced fields on the wire. Schneider (Groupe Schneider) has a very good online free book called "Electromagnetic Compatibility" that does a good job of explaining EMC and how to deal with it. If this doesn't help, there are two more things you can do. For an AC signal line, you can buy an EMC filter. These are frequently just ferrite donuts that you loop your wires around for a couple wraps. For DC, the same concept consists of a simple capacitor. These require either experimentation or engineering on your part since you are designing a simple RC or LC circuit. Second, depending on your scan time and your timing requirements, you can program your way around it by reading the input with a high speed timer and waiting for the timer done bit instead of just reading the input directly. This is commonly done anyways just to avoid intermittent sensor firing at the "edge" of your bolt for instance.

Share this post


Link to post
Share on other sites
The 871TM is a two wire device isn't it ? I would suggest three wire devices for PLC inputs as the current drawn by the two wire proximity switch can be enough to cause the input to switch. You may be close to the switching curent with the two wire device, and slight noise on the wiring is enough to tip it over the edge.

Share this post


Link to post
Share on other sites
Maybe can add a delay timer about 10 to 20 mS in the plc program for the sensor input. This is to filter the spike.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now
Sign in to follow this  
Followers 0