Re: [foxboro] RE �: Communication problem b/w FBM231 & AB PLC

  • From: Corey R Clingo <corey.clingo@xxxxxxxx>
  • To: foxboro@xxxxxxxxxxxxx
  • Date: Mon, 17 Sep 2007 14:08:31 -0500

>>The AB Gateway (or Integrator30) limiting resource is usually the RS232 
baud rate, likely 19, 200.

Not in our experience.  Our Int30As will not saturate even a 9600 baud 
serial link, a fact which I find puzzling to this day, given the hardware. 
 We do generally have more than one PLC per A-B highway (but not an 
excessive number), but the Honeywell PLCGs I used to use fared far better 
in a multiple-PLC scenario than Int30s do, and Rockwell RSLinx is in 
another universe entirely  performance-wise. 


Since our efforts to optimize the gateways led to even more problems 
(frozen data with no system errors, primarily), we are switching slowly to 
an AW70 integrator.  It's not a situation I like, but the AW70 has much 
better performance (it uses RSLinx to deal with the PLCs), and it's about 
our only other option unless we decide to go to V8 and CP270s.  In 
addition (because of RSLinx),  it can read _any_ PLC data from _any_ A-B 
PLC, regardless of the PLC type or the manner in which it's connected. 
Int30s can't deal with anything not directly connected to the highway they 
are attached to.


That said, you are correct about the desire to pack the data and read as 
many (contiguous) registers with as few ABSCANs as possible.  The effects 
of ABSCAN processing overhead can be profound -- we noticed it the few 
times we were able to make the changes without any other adverse 
side-effects.


Now, back to Imran's question :)  The addressing format will be different 
in the FBM231s than on the DI I imagine.  I've never set up a FBM231, but 
to continue my previous example, a CIN that referenced a bit in B3 on the 
Int30 might have a PNT_NO of 3:0.1, where in the AW70 it would be an FCIN 
block with a PNT_NO of [PLC_NAME]B3:0.1.  For the AW70 I found this out by 
trial and error with an OPC test tool (the AW70 reads in data via OPC), 
but I found this documentation for my FBM224s in the Modbus FBM User's 
Guide.  So your block types will be different ("standard I/A" in the DI 
vs. DCI in the FBM231), and the IOM_ID and PNT_NO will likely be 
different.  Unfortunately I did not find a document on the Invensys site 
that described the addressing format for serial-connected A-B PLCs; they 
only seem to have one for OPC connections.


Corey Clingo
BASF Corporation






"Doucet, Terrence" <tdoucet@xxxxxxxxxxxxxxxxxx> 
Sent by: foxboro-bounce@xxxxxxxxxxxxx
09/17/2007 10:14 AM
Please respond to
foxboro@xxxxxxxxxxxxx


To
<foxboro@xxxxxxxxxxxxx>
cc

Subject
[foxboro] RE : Communication problem b/w FBM231 & AB PLC





Imran,

The AB Gateway (or Integrator30) limiting resource is usually the RS232 
baud rate, likely 19, 200.  Each serial communication   consists of 
overhead (addresses, security check, etc) that do not provide information 
about your process and data (your process good stuff) that all must be 
part of the 19,200 baud.   You need to try and minimize the overhead by 
reading (or writing) contiguous bytes of PLC data.

Most users get the PLC logic to move the various pieces of information in 
the PLC into contiguous data areas in the PLC and then the Foxboro 
Integrators read that contiguous data with the largest read that is 
possible.

If you just read your input data directly, most of your 19,200 baud would 
be overhead. It takes a little work to map the data in and then read it 
out on the Foxboro side but it is well worth the effort. At 19,200 baud 
you can usually read about 350 bytes per second of your data but there are 
a lot of other factors (like the number of PLC's on your DH+ network) that 
could lower this 350 bytes per second.  Also you need to keep writes from 
the Foxboro to the PLC to low numbers per second.





 
 
_______________________________________________________________________
This mailing list is neither sponsored nor endorsed by Invensys Process
Systems (formerly The Foxboro Company). Use the info you obtain here at
your own risks. Read http://www.thecassandraproject.org/disclaimer.html
 
foxboro mailing list:             //www.freelists.org/list/foxboro
to subscribe:         mailto:foxboro-request@xxxxxxxxxxxxx?subject=join
to unsubscribe:      mailto:foxboro-request@xxxxxxxxxxxxx?subject=leave
 

Other related posts: