reducing waving readings on a Level transmitter.

Hi everybody,

I need any hints for trying to solve a little prob here. I'm feeding an input channel of a 1769-IF8 analog module with a current based signal of 4-20 mA linked from a measuring field device, an ultrasonic level transmitter,a SIEMENS Sitrans Probe LU and I'm obtaining "jittering" input readings. If, for
instance, the admitted measuring range by the LIT is 300 centimetres and the real level for the water levelline in the tank is 100 metres, my input readings are waving in a range of 10 centimetres as I can monitor them on the RSLogix( 99, 105, 103, 110...like that, randomly). For the particular case of
my application, an inaccuracy of about 10 cmts is fairly acceptable, but still I wonder how to improve the quality for my signal.This is because afterwards I will program a routine for keeping the levelline steady at a certain setpoint (by means of a speedrive controlling a pumping out process)and this setpoint shouldn't go out of a range of 10 centimetres, (otherwise I'd have to start up secondarily another pump supporting the main one).

The field device seems to be correctly installed and programmed (no error messages displaying)...the fluid whose level I want to monitor (sewage waters) have no foams which might be disturbing the ultrasonic echoes of the LIT. And this "jittering" is happening when no flow is being drive to the tank inlet, this is, when the level is steady.

Any clue, tip, hint, strategy...?

Thanks in advance.

First guess is noise, double check the field wiring, what type of cable are you using, shielded and twisted (I hope). Is it near any AC wiring, VFD's in the same wire-way?
Many higher end transmitters have an option to average the signal over a few seconds, maybe this would help "smooth" out the signal if the wiring is okay.

Hello buhnen;
I have seen "jitter" of this kind on an ultrasonic transmitter, when they were installed with too long a support tube.
If the transmission cone of the transmitter is not allowed to exit the support tube unhindered, you will get unwanted reflexions from the tube, added to the ones from the target, and that would mean "ghost" reading, or what you see as "jitter".
We even had to relocate the sensor at one point because it was positionned too close to the tank wall, which was giving unwanted reflexions.
Hope this helps,
Daniel Chartier

I agree with the others. Try to solve the problem right at the transmitter rather than trying to put a band aid on the PLC program. Most ultrasonic level transmitters have a simulator function where you can force the 4-20 mA output to any value. Try setting it at a fixed mA and see if the PLC input is still jittery. If so, then you have to look at shielding issues or grounding problems with the transmitter.

There are several possiblities. As discussed above, transmitter mounting and wiring are two of the usual suspects, and you should follow the advice given to determine if they are the cause.

Another problem if the transmitter is mounted outdoors could be wind. This can actually blow the ultrasoninc beam about and cause problems. In that case a generously sized "stilling well" to shut out the wind could help.

It is also possible that you actually have waves in the fluid surface. This could be the case if you have fluid entering the wet well causing turbulence, although you indicate this is probably not the case. Wind can also cause waves.

A last resort would be to have digital filtering on the analog input signal. This will damp the fluctuations in data. You should also have a time delay before starting the second pump, to insure that momentary fluctuations in level don't kick it on.

It is not unlikely, as Tom describes that there is turbulence within the vessel.

The Probe LU has several settings to dampen the output:



The fill and empty update rate is established as part of the initial setup in parameter 003, Measurement Response


The Measurement Response values can be setup independently as fill or empty rates:


or empty update rate:


The caveat is that damping the response to turbulence also damps the response to real change, ie filling or emptying.

Dan

Thanks mates! Your hints have been truly useful to me. I've been talking to our electrician and he is going to check tomorrow whether the wire shielding has been correctly grounded. He suspects it's not been like that, so we'll see it tomorrow.

On the other hand, I'll check the P709 parameter thatDanw mentions above, as the liquid surface is a splashing one and it might be suitable to increase the default amount of secs for this damping filter setting.

Once again, thanks to all of you!

Hi again to all of you,

I've finally solved this little prob I had with the waving readings. It was all about a wrong grounding.

But I wanted to come back here again with something new, just a matter which might be of general interest...

I'm using a bluetooth freehand device paired with my mobile for communicating with another fellow electrician who usually helps me by cheking on field some events during the programming. The thing is that when I was next to the pool where the ultrasonic is placed I started to hear a noise coming from my bluetooth freehand and I didn't give much importance to it at first. But later on we realized that something weird was going with the LIT. The readings were saying that the level line was increasing when in fact, what was happening for real was that the level was going down. I switched off and switched on the LIT and it all again worked fine. As far as I could find out by checking more thoroughly all this, it happened that when my mobile was away from my freehand and therefore both got unpaired, the freehand was automaticlly attempting to get paired back to the mobile and it seems that there were disturbances between the LIT and the freehand.

I just wanted to post here this happening just in case anyone else had ever experienced something similar because I'm still surprised about it.

Thanks for getting back to us with the solution - it's nice to close the loop!