4-20mA Trouble

Here it is plain and simple, except its not that simple.
I have a 4-20mA signal being sourced from device to a signal isolator. The output of the isolator is connected to a Allen Bradley control logix analog input module. When the device is put in to standby (not running) the analog signal sits at 4mA's. After any time interval the analog signal drops to 0mA's. I have tried to remove the isolator and this still happens. I have jsut connected a VOM to the device and the signal still drops. Any idea!!!

jwhitesock said:

After any time interval the analog signal drops to 0mA's.

Click to expand...

What do you mean by this?

Is the time interval random? That would indicate a break in the wire to me. Does it always fail at about the same amount of time? Is the device supposed to do this (I would guess not, but I'd still check the documentation)?

How good is the power supply running to the device? Have you metered it (especially while you see this problem)?

If the device is behaving like this (and you've ruled out the power supply), I would probably have it replaced.

Yes the time interval is ramdom. No the device is not susposed to due this. Today the Manufacture from the UK was here and replace all the circuit boards and power supplies in the unit. The problem still occured. Could this be a groud loop issue?

I guess carrying what Jimbo says a little further. How do you recover from the issue? DO you power down and back up? DO you lose the 24V when you lose the 4mA? If so see if your power supply current rating is to low and it is shutting itself down. Is the device run from 110VAC or 24V separate supply?George

To recover, the loop has to be opened and then reconnected. The power supply is still @ 24VDC. The unit is supplied with 120VAC and a transformer is integrated to the circuit board. I actually have two units side by side doing the same thing. All parts were replace today in both units.

You mention it dies even just being hooked up to your mA meter. If it is doing it on the test bench, not likely to be a ground issue. Can you give a model number of the unit?


George

I would perfer not to disclose the unit information. They are also trying to figure out the problem. I am just trying to rule any of the things that I have already done to address the problem. I just thought I would see if anyone else had any other ideas that I didn't already try.

have you checked the input channel on the plc?

have you checked power supply and wiring? so far you don't mention getting any signal over 4mA. are you positive wiring is correct?

Is this only happenning when the device is put into 'Standby' mode?
Maybe you should (as a quick fix) avoid using 'Standby'; it sounds like the manufacturer of the transmitter has some problems.

Well today being 6/6/06 I really thought the world was going to end so I was just going to give up on this and let the manufacture figure it out. But as noon rolled around the day was still progressing. Darn.

Back to the drawing board I thought. What haven't I tried yet? Humm. Going with the ground loop theory I thought what the hell lets just disconnect the gounds and see what happens. That was at 1pm and when I left at 5pm all was well. But since this problem sometimes doesn't happen for 8, 10, 20, etc. hours I still have to go back to see if it worked.

If this solves the problem, I think I need a vacation.

Well, Well, Well..... Just the one thing that I didn't try happened to be the problem. GROUND LOOP!!!! Why is that the hardest problem to detect? Does anyone know of a good way to toubleshoot ground loops? If so I would like to know.. Thanks to all who had input on this problem!!

just check the drawings and then wiring. analog signals (like any feedback signals) should be shielded properly and this should be one of first things on checklist. feedback signals are to be grounded in one place only, it is good idea to make them continuous from sensor to plc (no splices and chances for branching - intentional or not). if there is any doubt (or no doubt but darn thing doesn't work) run second cable. doesn't need to be pretty or high flex even if this is moving part (almost any stranded cable will survive long enough to determine if it was cable fault or not). check power supply for sudden drops or if in doubt try non-resetable one like Omron S8VS (they have very fast protection and if there is a fault, they switch off and stay off until power is cycled). this will not tell where the problem is but it will identify problem with power or wiring (intermittent shorts).