Value Fluctuation

[viscoelastic]

Short Story: Reading 4-20mA raw number fluctuates as much as 60 points on 12 bit. When I put a milliamp clamp at the wire right into the V200-18-E4XB Snap in I have an absolute steady milliamp reading. Why does the raw value before linearization fluctuate so much with an absolute steady mA.

Long Story

Before I got a milliamp meter I kept changing wire, ultrasonic level sensors, laser level sensors and a milliamp generator to a set value. There have been posts in the past by me as I have been working through this issue of fluctuation. I always assumed it was something other than the PLC. Now I have trouble shooted everything back to the PLC and it is the unit that is creating the fluctuating counts.  I am officially at the Grrr state of trying to solve this. Please help unGrrr me.

[Flex727]

How many decimal points of precision on your clamp? Also, are you sure the clamp doesn't have some built-in filtering?

The 60 point fluctuation is only about 1.5% on a 12-bit input.

[viscoelastic]

Hey Flex, so is that normal then to have that much fluctuation? I have the filter set to high and it does not change it at all. The problem is that it makesthe linerarized number hard to read.

 

[Flex727]

It depends on what the sensor is. Some are worse than others. You can add additional filtration if you need to, or show one less decimal on your HMI.

[viscoelastic]

I also used a milliamp generator instead of a sensor. All results are the same.

The reason I cant show less digits it because I am linearizing it to liters in tank. The result in an integer in 1 liter increments.  

[Flex727]

I need one of those 4-20 mA generators! They're too expensive.

I have no experience with the mA generator, but I would assume it would hold a steady reading. If you're still getting fluctuation with that, then maybe there is some kind of hardware problem with the I/O module or a power supply issue. You probably need some input from Unitronics on this.

[viscoelastic]

Thanks Flex, I was thinking the same, but I figured I would bounce it off of people who are smarter than me before I dropped the dime.

[Flex727]

7 minutes ago, viscoelastic said:

but I figured I would bounce it off of people who are smarter than me before I dropped the dime.

In that case, you need to ask @Joe Tauser.

 

[Ausman]

Hi Visco,

before you do anything else, you need to do a direct reading on your line, not with a clamp meter.  Clamp is not good enough, even though plenty of people will tell you so.

Just FYI, I had much the same thing with some combo temp/hum/co2 sensors doing a regular fluctuation.  I initially thought it was an induction problem within the shielded line which was about 40m long.  It eventually turned out that each time the co2 system triggered it's read pulse, the entire unit's outputs changed.  The (bigname) manufacturer was most unhelpful even when I proved it to them with videos.  In the end I had to filter it out with a fairly complex algorithm.  Not happy.  To prove the error, I used my mA generator.  They don't have to be expensive, there are many good types available that are well under $50US.  Mine is a bare pcb setup powered by a 9 volt battery and I think it was about $30AU years ago.  It does mA, mV, 0-10, 4-20, blah blah blah everything under the sun.  It's a little cumbersome to use but works great and is rock steady.

Try a direct read and pls report back.

PS Joe, sorry...beat you to it!!

cheers,

Aus

[Ausman]

And further to this,

have you tried your generator in both locations?  ie the sensor end of the cable, and also direct into the PLC?  Both of these need to be done with the non-clamp direct reading.

cheers,

Aus

[viscoelastic]

Hello all, Ausman, I have a millivolt generator and tried that also. So to put this thread to rest. This morning I found a V350 laying around and hooked the sensor to it. The reading is steady as a rock. So this defiantly comes down to I think I have a bad snap-in.

[Ausman]

That's great Visco....sort of!  It will be good to know if the replacement does the same thing...hopefully not.

Sadly for me, my nature is to find the reason/solution when I encounter a large Grrr, even if it is relatively small in the scheme of things.  Curiousity only killed the cat.....so as long as the human is careful it is ok!

Very glad you are unGrrrrd for now.

cheers,

Aus

[Joe Tauser]

You did a good job troubleshooting by taking a lot of readings to verify the problem was not in the signal.  The only additional thing I would have done was put an oscilloscope on the input terminal to see if there was any noise.  Swapping in  a different PLC was then the next logical thing to try.

The A/D converter on your module has become flaky.  But you already know that

 

Joe T.

[viscoelastic]

Thanks Guys! Now we have to get Joe Tauser to put up a profile picture with a smile in it.  Right now he looks Grrr 

 

[Flex727]

I don't know, that looks like a smile to me. After all he appears to be enjoying a fine cigar!