DNP3 Analog In problem on sending to scada real variable value

[RLMM]

Good morning, 

I have a M580 PLC and i am trying to configure DNP3 communication with a SCADA server. I can communicate, it works with no errors, for all kind of variables, except for Analog Inputs, which when i try to send, for example, 60.12 on a real variable, using group 30 and variation 1, or any other kind of combination between groups and variations, it returns the value of 4354170. 

As it seems, the last two hexadecimal digits from the value thar i am sending to scada (60.12 in the example above) are being cutted, because 60.12 in hexa is 42707ae1, and the 4354170 is 72707a. 

Does anyone knows how to send it correctly?

Thanks

[BevanWeiss]

I'm confused...

You say '... when i try to send, for example 60.12 on a real variable, using group 30 and variation 1..'

Group 30 Variation 1 is a 32bit INTEGER value with FLAGS

So you can't send a real value with such a data type.

You should be using Group 30 Variation 5 (short 32bit floating point value).  And this should be configured in the DNP3 Slave (the hardware).

The M580 PLC doesn't natively talk DNP3, so you must have some kind of adapter and configuration, or an RTU.

You should supply screenshots for the configuration of these.

Using an M580 and a NOR module does work (communicating up to Geo SCADA Expert, and across to Kingfisher RTUs, or SCADAPack RTUs), but you need to configure it correctly.

If you are using an RTU, then there must be ModbusTCP or similar to get the data from the M580 to the RTU also.. so you'd need to configure that correctly also.

[RLMM]

I have descovered what i was doing wrong...

I am using a NOR card.

I was linking the variables and the PLC by the name, and using group 30 variation 5 as you said, but it did not work, so i have tried by associating them using the PLC memory words (%MW), and i dont know why, but it works. 

But thanks for the help, it helped a lot.

[BevanWeiss]

I don't remember the NOR configuration supporting symbolic addresses.

I was pretty sure that for each group/variation it required to set aside a certain memory range (%MW range) which would be converted to the configured data format.

This wasn't always a straight forward mapping (i.e. for booleans it wasn't always just one bit = one boolean DNP3 point... since there are all the DNP3 flags needed to be allowed for also).

I don't think I've ever seen a UDT used for the data mapping... but it would be much nicer (if it would work with data alignment etc).

I personally dislike the NOR module, since it seems poorly integrated into the rest of the logic design environment.  It would be great if it was upgraded and done in a similar fashion to the x70 SCADAPack devices.  Where DNP3 points are DDT variables (like DTM DDT entries.. but more structured).

[RLMM]

I agree with you, it could be integrated with the logic design environment, it would be much better and simple to be programming.

But thanks for the help Bevan,

See you on another doubt.