Modbus Value scalling Problem

Hi All

I am doing a Tank Gauging project. In there all the tank parameter like level , temperature , pressure are coming into a master instrument(Tank side monitor made E+H) through multiloop HART protocol. the master instrument send data MODBUS rs485. Now i am able to get data from the Tank side monitor through Modbus to a Micrologix 1400. But the problem is i am getting value of when tank level are for
0 meter - 0
1 meter - 16256
2 meter - 16384
3 meter - 16448
4 meter - 16512
5 meter - 16576
At 20 meter - 16880
My tank high level is 20 meter.
Now in HMI i want to display the tank level in meter. Now how can scale this value into meter. i use SCP instruction for scalling but the values cnages with every 1 meter is not uniform.

Tnanks
Suvra

Does the PLC have a linear approximation function? they are generally pretty rare but I use Omron and use it all the time for this sort of stuff.

I dont think so that the AB PLC have this function.

Hello,

F(X) is needed.

It is also called a characterizer or FGEN or look-up table.

People call the function a function generator (FGEN), characterizer, or look-up table. The values used frequently called breakpoints.

Here's a screen shot of an explanation from a controllogix manual:

This sound like a badly chosen or miswired sensor. I wouldn't try to derive useful information from that signal.

I had to read the original 4 times over and ponder what he was asking, too. I eventually assumed that he's doing "tank gauging" and getting a count that isn't exactly 16384 for each 1 meter interval, which indicates some variance that needs to compensation. 'Tank gauging' (inventory of high value product, like petroleum or petrochem products) gets concerned about minor discrepancies, as opposed to 'simple' level.

If he's getting 16384 ±20 or whatever for each of those levels, then yes, something is drastically wrong. But I think he's trying to say that the 1 meter intervals are not exactly 16384 different than the prior level, and how does he compensate for such.

If the measurement between 1 and 2 meters is only the difference between 16256 - 16384 = 328 counts, then there's something critically wrong in the setup, because 98% of the resolution is between 0 and 1 meter, leaving only 2% resolution for the remaining 95% of the span.

original values:
0 meter - 0
1 meter - 16256
2 meter - 16384
3 meter - 16448
4 meter - 16512
5 meter - 16576

He had mentioned various reading sources. Could he be actually be reading a valid value but one not related to the depth of the product?

Hello,

Or byte swapped or swapped and rotated or wrong format or ...

Thanks all of you. I will look that if any thing is wrong in my setup or not. i will try your suggestions.