hello,

I am having trouble getting Modbus Tcp registers read in siemens tia v13.sp1.

i have an DB with Reals declared.

the values are written in Honeywell so it a different Endian , i get that.

ive made an block in the past that does the byteswap for me .

now if i enter 120000 in honeywell it comes out in siemens : something like 118000

so way off

this is the code in my conversion block:

 L     #R_In

      T     "md100"

      L     "mw102"

      T     "mw104"

      L     "mw100"

      T     "mw106"

      L     "md104"

      T      #Floatswap

Is there something wrong by recieving the data that is swapped as an real in Siemens tiaportal?

or is the instruction to swap wrong ?

im confused ...

thanks in advance for helping me .

Try this for 32 bit values (REAL, DINT):

L     #R_In
CAD
T      #Floatswap

and for 16 byte values (INT)

L     #R_In
CAW
T      #Floatswap

see https://support.industry.siemens.com/cs/ww/en/view/45523446 for more details

miami

Try this for 32 bit values (REAL, DINT):

L     #R_In
CAD
T      #Floatswap

and for 16 byte values (INT)

L     #R_In
CAW
T      #Floatswap

 

see https://support.industry.siemens.com/cs/ww/en/view/45523446 for more details

ive tried it  but didnt work for me as well.

the strange thing for me is if i look in my recieving DB (data written by honewell)

hex codes look different then when siemens writes them.

for instance":

typed in on the  honeywellsystem -

100000.0 

16#0000_47C3  ( recieved in my  DB)

so i swap it to this 16#47C3_0000

then i read it in an variable tabel . it reads 99840.0

when i write in TIA , 100000.0 to an MD  and i look at the hex code 

it reads :  16#47C3_5000 

im not sure if it can be looked at like this , but its strange . where comes the 5 come from...

small numbers have a little offset , but the bigger they get the more offset it gets.

92.0% gets 90.5%  so its a big problem for me .

Well 100000.0 does equate to the hex code 16#47C3_5000 and 16#47C3_0000 does equate to 99840.000.

So the TIA system is correct and the Honeywell system is not.

If the honeywell system is transmitting 100000.0 it should most definitely send the hex value 16#47C3_5000. You need to talk to the Honeywell engineer and ask him what is going on in his system as it is clearly not correct. Maybe some issue with rounding? But it seems a very long way out for a rounding error.

If he doesn't believe you, then show him the evidence using the links below.

Use this link to convert decimal FP to hex equivalent:-

http://babbage.cs.qc.cuny.edu/IEEE-754.old/Decimal.html

Use this link to convert hex equivalent to decimal FP:-

http://babbage.cs.qc.cuny.edu/IEEE-754.old/32bit.html

smiffy

Well 100000.0 does equate to the hex code 16#47C3_5000 and 16#47C3_0000 does equate to 99840.000.

So the TIA system is correct and the Honeywell system is not.

If the honeywell system is transmitting 100000.0 it should most definitely send the hex value 16#47C3_5000. You need to talk to the Honeywell engineer and ask him what is going on in his system as it is clearly not correct. Maybe some issue with rounding? But it seems a very long way out for a rounding error.

If he doesn't believe you, then show him the evidence using the links below.

Use this link to convert decimal FP to hex equivalent:-

http://babbage.cs.qc.cuny.edu/IEEE-754.old/Decimal.html

Use this link to convert hex equivalent to decimal FP:-

http://babbage.cs.qc.cuny.edu/IEEE-754.old/32bit.html

 thanks for the confirmation ,

i will pass it trough to the automation engineer from honeywell.

christian.vanputten

 

smiffy

Well 100000.0 does equate to the hex code 16#47C3_5000 and 16#47C3_0000 does equate to 99840.000.

So the TIA system is correct and the Honeywell system is not.

If the honeywell system is transmitting 100000.0 it should most definitely send the hex value 16#47C3_5000. You need to talk to the Honeywell engineer and ask him what is going on in his system as it is clearly not correct. Maybe some issue with rounding? But it seems a very long way out for a rounding error.

If he doesn't believe you, then show him the evidence using the links below.

Use this link to convert decimal FP to hex equivalent:-

http://babbage.cs.qc.cuny.edu/IEEE-754.old/Decimal.html

Use this link to convert hex equivalent to decimal FP:-

http://babbage.cs.qc.cuny.edu/IEEE-754.old/32bit.html

 thanks for the confirmation ,

i will pass it trough to the automation engineer from honeywell.

we found out what is was eventually , if honeywell sends out at holding register 40004 

i have to read it at 40002 . 

the other programmer was stating i had to begin at 4001 . 

so the 0000 in my second word was a blank spot .

the 5000 was used for the next real .

so thats why all the variables were off.

Off topic:

In TIA Portal Siemens finally invented SWAP instruction (disable IEC checks to use it with floating point) for all who couldn't live without it.