Modbus, read in registers as unsigned integers

[tactical]

Hello,

I'm using the MSG command with RSLogix 500. The values are being read in as signed integers. Is there a way I can read them in as an unsigned integer? Do I have to convert them?

 

[Ken Roach]

SLC-500 integers are always Signed Integers; there is no unsigned datatype.

Are the values intended to be unsigned integers with values greater than +32767 ?

 

[tactical]

SLC-500 integers are always Signed Integers; there is no unsigned datatype.

Are the values intended to be unsigned integers with values greater than +32767 ?

Yes, the values are unsigned integers per the manual. Is it possible to convert them to unsigned?

 

[Ken Roach]

That's the problem: there is no unsigned integer format in the SLC-500 controllers.

The MicroLogix 1100/1200/1400/1500 have a Long Integer format, which would display a 16-bit unsigned integer correctly.

Otherwise your principal option is to convert the value to Floating Point, with some possible loss of precision (though with a 16-bit value, not much).

You'll have to do a little bitwise manipulation to do this: use a Masked Move to remove the sign bit, then add +32768.0 to the resulting value if the sign bit (the highest bit, /15) is true.

 

[tactical]

That's the problem: there is no unsigned integer format in the SLC-500 controllers.

The MicroLogix 1100/1200/1400/1500 have a Long Integer format, which would display a 16-bit unsigned integer correctly.

Otherwise your principal option is to convert the value to Floating Point, with some possible loss of precision (though with a 16-bit value, not much).

You'll have to do a little bitwise manipulation to do this: use a Masked Move to remove the sign bit, then add +32768.0 to the resulting value if the sign bit (the highest bit, /15) is true.

Not sure if that would work? It's a 16 bit integer so the value that corresponds with -1 should be 65535. Doing what you stated would give you 32768. Instead would you do the following?

- IF the value is greater than or equal to 0, then do nothing and use that value. Else, add 65536.

- So if the value is -1, you'll add 65536, which gives you 65535

 

[Ken Roach]

Here's an example:

Source Decimal Value: 54321
Source Binary Value: 2#1101010000110001

"Strip" the highest bit out using a Masked Move with all but the highest bit set, and the result is 2#0101010000110001

The decimal result is 21553. Add +32768, and you get 54321.

 

[tactical]

Here's an example:

Source Decimal Value: 54321
Source Binary Value: 2#1101010000110001

"Strip" the highest bit out using a Masked Move with all but the highest bit set, and the result is 2#0101010000110001

The decimal result is 21553. Add +32768, and you get 54321.

awww a see what you're doing now. I'll have to give that a try when I am back in the office tomorrow.

 

[tactical]

Okay, I tried the masked move but couldn't get it to work out. I believe it's because the value I have is in a data file of type N. Do I have to convert it to a binary file before the masked move? See picture below for the rung.

 

[Ken Roach]

That ought to work.

Basic troubleshooting: are you sure that rung is being scanned, and nothing else is writing zeroes to B3:3 ?

 

[Ken Roach]

The "mask and add" method I suggested is just because I prefer to make it really explicit what I'm doing when I manipulate data.

"If VALUE < 0 Then Add VALUE + 65536" is functionally the same thing and will give the same result.

 

[diat150]

Just wanted to say thanks to Ken for posting a solution. I was trying to pull a unsigned register from a modbus device using a micrologix 1400. unfortunately the modbus message doesnt support writing the unsigned registers directly to a long datatype so I was able to workaround these shortcomings using his posts.