Re: [foxboro] FBM 233 to Controllogix question

First thanks to all who replied.  Here's a summary of what was said, 
and what I found.

 From Terrence
If you do not pack the data and read contiguous bytes, the overhead 
load becomes very high, even with ethernet speeds.
...
I believe the controllogix uses a double word (32 bits) even for a 
boolean type.  The application we are interfacing with is a canned 
package, so we are not in a position to re-pack the information even 
though that would be better.  I am reading fewer than 275 tags total 
and only writing about 6 words back.  Loading would seem to be quite low.

 From John
I believed I have observed the issue you are having when we first 
brought our 233s online. If you haven't already done so, investigate 
adding the @ and TO settings to the DVOPTS parameter of the ECB201 
block representing the target PLC. The value entered after @ will 
influence the scan time. The TO value will influence the timeout value.
...
We have
   DVOPTS = EIP/CLX+TO=5+SN=1
I didn't have anything after the @, and will investigate that.

John also listed the pseudo-tags for FBM23[2,3] internal 
diagnostics.  (More on this below.)

 From Ron
What I found was that it appears that over time we would run out of 
memory. I believe it was due to invalid point that didn't exist 
(yet). I could only prove that the memory was reaching max(using John 
Dykes method) over a period of time.
...
That could well be the case, because the symptoms are consistent with 
running out of memory, although I tested each read after the block 
was built due to the known problem with bad tags.  I set up all the 
diagnostic counters for messages, but not for memory.  I noticed 
those in the first email that John sent.  I then started kicking 
myself and wishing I had built blocks for memory statistics 
initially.  I will see if I can get that done to help troubleshoot the problem.

Also from John
>With the earlier version of the driver, we found that misconfigured 
>points (which were deleted) still were actively polled by the FBM. 
>We could only clear it be rebooting the FBM and doing downloads of the driver.
...
Yeah, I knew about that.  In fact we were able to demonstrate the 
problem at the Southeastern users group meeting back in February 
before this was an acknowledged problem.  We have the version 1.22 
driver installed which I thought was the latest but....

TAC, says to upgrade to the latest drivers, and all will be 
well.  Apparently there is a still newer ethernetIP driver.  We shall 
see.  I hope that fixes it, because other than the brain damaged DCI 
blocks, the FBM23[2,3] is a pretty sweet interface module.  It beats 
a KF2 like a drum.

Thanks everybody.
David


 
 
_______________________________________________________________________
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:             http://www.freelists.org/list/foxboro
to subscribe:         mailto:foxboro-request@xxxxxxxxxxxxx?subject=join
to unsubscribe:      mailto:foxboro-request@xxxxxxxxxxxxx?subject=leave

Other related posts: