JZ20-R16, level transducer and lcd problem

[Bart000]

Hello,

I have a JZ20-R16 plc and a Aplisens 4-20 mA level transducer.  I do linearization in ladder to get the correct level value in meters.  Depending on the level the plc starts and stops a pump, and it all works greate. Start and stop levels are input by the user from keypad, also the max value of the transducer.

Everything works ok, the problem is with displaying current level on lcd. Right after linearization i copy the value to MI3 that is set to be shown on the "main" screen. It is displayed but the readout is jumping around/changing very quickly (level is constant), numbers get blurry and they are unreadable. Tried this on 2 plcs and 2 transducers with the same result. I've seen this working but can't find any reason for this behavior. Maybe it's reading the value to often?

 

Please help me. Any advice will be appreciated.

[Ausman]

If you could post your program that would be helpful.  Otherwise, screenshots of the relevant areas.

cheers, Aus

[Bart000]

For now i can upload screenshots. I'll post the whole program in the afternoon.

[Isakovic]

You should filter the value a little bit before displaying it, analog input jumps a little bit up and down, it's normal, I guess if you display it immediately lcd will look blurry. Maybe take a couple of readings in short time and display the average, or update register that shows value every 2-3 seconds. I reckon value doesn't change drastically in time since it's level.

[Ausman]

Agree with Isak.  Have you set the filtering at the raw input in Hardware configuration?

Screenshot not really!  U90 loses the dropdown during screenshots, so taken with a camera!

 

Have you looked at the raw input when in online mode?  If the number is jumping around a lot there, then you need to find the reason.  A correctly located level sensor should be pretty stable.  There are numerous factors that can affect this, though.

Also, you might like to think about using the linearisation available to you in the Variable, rather than doing it through the ladder.  Look at the bottom right of your Variable 1 screenshot.

cheers, Aus

[Flex727]

Just put a positive transition of SB3 in front of your Store function in line 14.

[Bart000]

Didn't know about the filter setting, gonna try as soon as i can.

Tried delaying the screen update by counting the 100 ms pulses, and copying the value when the counter reaches a certain value, but it didn't change a thing.

In online mode the value also jumps.

I worked with many other plcs and this is the first time i encounter such a behavior. The same transducer connected to another plc, just thrown into water shows a stable readout. That's the reason why i am sure that i'm missing some probably basic thing.

Don't think i can use the Variable linearisation because the max value of the sensor is input from the keypad and screen and i don't think i can do that in the Variable linearisation.

 

Here is the whole program.

 

2 pompy kolko delfin analog 204 delay.U90

[Ausman]

Haven't looked at the program yet, will get to it later.

Also, don't forget that the an input can be 10 or 12 bit.  This might affect things based on the numbers involved.

cheers, Aus

[Flex727]

On 11/20/2018 at 4:29 PM, Bart000 said:

numbers get blurry and they are unreadable.

 

3 hours ago, Bart000 said:

Tried delaying the screen update by counting the 100 ms pulses, and copying the value when the counter reaches a certain value, but it didn't change a thing.

You're saying the numbers are blurry and unreadable when updating the HMI only once per second?

[Bart000]

I wanted to slow down the updates but probably did something wrong and it didn't change anything. Still updated very quickly.

[Isakovic]

Idea was something like this. Maybe SB80 needs to be in a separate net, I'm not sure, can't test it.

[Ausman]

What U90 version is this written in?  I can't open it.

cheers, Aus

[Ausman]

In response to Isak's post, I dug out an old U90 program that has linearisation that I know worked perfectly.  Paste attached.  It was a 4-20ma temp sensor, set for 12 bit and giving the temp from 0-50.00°C.

U90 is quirky in the way you have to do all the linearise in one hit.  I'd give it a go along my lines and see how it goes. It might look odd but it works ok. Don't do any delays except the hardware maximum filtering. 

Also, ensure that you have nothing that can accidentally overlap areas of the screen usage.  I've heard of minor overlaps creating this sort of issue.

13 hours ago, Bart000 said:

In online mode the value also jumps.

Is this the raw value, or the linz result? 

cheers, Aus

[Bart000]

program version is 6.6.36

 

I had it at first just like in you're example above, everything in one line. When the problem occured i started experimenting. Didn't know about the hardware filtering, and i think this could be the source of the problem (no filtering set). Should be able to try it on the plc within the next few days, gonna let you know about the results as soon as i can.

[Ausman]

Pls still advise on the "raw value" question.

cheers, Aus

[Bart000]

Don't remember right now which one was it, will check.

[Bart000]

In online mode both values are changing very quickly. The "raw value" and after linearization.

Tried to set the filter and it didn't change anything on any of the 3 settings.

At the end i followed Flex727 advice and put a positive transition of SB3 in front of the value copying and everything is working perfectly. So i think the problem is solved.

Thank you everybody form your help.

[Ausman]

How much was the raw value changing?  If it was a few percent of total span, something is wrong with the setup, be it mechanical or electrical interference. 

I'm only saying this because an intermittent read of a pulsing source does not necessarily give you an accurate reading of the true average.

cheers, Aus

[Bart000]

It was changing by +/-  5-10, and it's 10 bit resolution. But this are two completly different applications. Except for the plc model, transducer model and program nothing else is similar. Before the end of next week i'm gonna be doing another pumping station in a completly different place. So i'll check how it behaves and let you know. The problem is omitted, but it would be good to know what caused it.