Omron Function Block

[bdmfab]

First time with structured text and something isn't working correctly. Simple block, copy of the SCP in Allen Bradley (scale with parameter). The math is off, I think its syntax. Driving me nuts. Its a CJ2M series. Thanks.

[hooey]

see attached file for my version for scaling int to real
hooey

[bdmfab]

Thanks Hooey,

The conversion to real was helpful.

Getting roughly the same result. I'm running the simulator, here are the values I am using.

In = 1000

InMin = 100

InMax = 20000

SclMin = 1

SclMax = 100

SclOut = -49597.80.

My trusty excel sheet yields 5.48

This is only my second Omron project so I am still green on this. I attached the .cxp with the FB. Am I using the correct registers? Cheers.

[BobB]

Hear is a link to an Omron scale with parameters FB.
http://forums.mrplc.com/index.php?ap...s&showfile=523

[hooey]

Originally Posted by bdmfab

Thanks Hooey,

The conversion to real was helpful.

Getting roughly the same result. I'm running the simulator, here are the values I am using.

In = 1000

InMin = 100

InMax = 20000

SclMin = 1

SclMax = 100

SclOut = -49597.80.

My trusty excel sheet yields 5.48

This is only my second Omron project so I am still green on this. I attached the .cxp with the FB. Am I using the correct registers? Cheers.

bdmfab
you use of d registers might need reviewing, d registers are 16 bit storage locations so used to store a real type number will use 2 16 bit locations d0 and d1 and as you have used d1 as a input to the function block as constant you will end up with corrupted results . define symbols for inputs to the function block as type real ie. InMin type real, InMax type real, SclMin type real, SclMax type real, SclOut type real and give that a go.
Hooey

[BobB]

I have basically given away using FBs for almost everything - too slow and painful. I used to use FBs for floating point maths. I often have to extract data from Caterpillar, and other brands, engines. It is extr5acted by Modbus RTU but is in J1939 format - absolute pain. Value, precision and offset. I have attached a routine I recently used on a job - fairly obvious what is going on. Admittedly not scale with parameters nut may help you get a handle on floating point in Omron.
It is part of a standard routine I developed - the engines all come with different information available. I stick in the standard routine and then delete what I do not get from the engine. The only other thing I do is use FIXL (fix long) to read the values in decimal. I then delete the FIXL function when I am happy with the readings I use the floating point values on Omron NS screens.
Hope this helps you get a handle on things.
Float all values first, including the constants you have to use for precision, multiples to make the values useful and offset values.