I am curious as to the user community's experiences with HON Experion. Most of my experience is with TDC-3000 and HON TPS.
There is a lot to like about Experion: the HMI web builder is easy to use and has a decent suite of graphic objects and picture manipulation capability, control builder is easy to use compared to TDC. I also like the ASSET model for alarming.
I wish they had more choices for history collection: I am only aware that they can collect the same HM history rates: 5 sec, 60 sec and averages (6min, 1 hr, 8 hr, 24 hr).
I don't know much, however, about the system performance and functionality in an actual process environment and would appreciate the insights of those of you who have used it in a plant.
Good to see an ex-Honeywell employee on the forum. Actually they have 1 sec fast history option. About the performance I can tell you that is much beter than the TDC was.
It is more user friendly. Graphic performance is excellent when the points in the graphics reside on the C200.
I read your comment on the forum about the callup performance of GUS graphics that you were able "to get the GUS display call up time from around 5-6 seconds average to 2 seconds." Now the TPS is getting integrated into the Experion. TPS is seen as an OPC server from the Experion side.
What is your opinion about this type of OPC technology based integration? I assume that you know the TPNServer opc interface from TPS.
> Actually they have 1 sec fast history option. <
Interesting: how do I configure 1 second history? I did not see such an option.
> About the performance I can tell you that is much beter than the TDC was.
It is more user friendly. Graphic performance is excellent when the points in the graphics reside on the C200. <
I am seeing 2-3 seconds on a graphic with about 400 parameters residing on a flex stations. My empirical (but not very scientific) observation is that The performance seems to be a function of the graphics card and how loaded the server is.
I don't entirely understand the most efficient way to implement the scan groups, but the server seems to be very slow the first time a set of parameters is accessed and I am assuming that it sets up a scan group, because on subsequent accesses it is faster. If I transisition to a configuration display for a while, the graphics call up is again very slow, and I am assuming that the server is purging the scan group after a while.
I was reading the knowledge builder and apparently it is possible to set up scan groups: I need to dig into this more. The problem is that if you scan all the time a high ambient load is imposed on the system, but in the case of a plant upset or critical alarm the operators would want to transition quickly to the display that corresponds to the plant area they are concerned with.
> I read your comment on the forum about the callup performance of GUS graphics that you were able "to get the GUS display call up time from around 5-6 seconds average to 2 seconds."
Now the TPS is getting integrated into the Experion. TPS is seen as an OPC server from the Experion side.
What is your opinion about this type of OPC technology based integration? I assume that you know the TPNServer opc interface from TPS. <
That is how I would have approached it, though I would like to see the faceplate call up for TPS points a bit faster. Based on what I know of the internal LCN/UCN network timings, I would think 0.5 second TPS face plate call up should be easily archievable.
Do you know if the faceplate transactions through OPC are sent at a higher priority level than display parameter and history requests? I would not want my faceplate controller changes competing with that lower priority traffic.
"I would think 0.5 second TPS face plate call up should be easily archievable."
I have never seen a faceplate or a change zone which appears in 0.5 sec. Even the good-old GUS faceplate ocx appears in 1 sec (average). With the HMIWeb style faceplates the data shows up in more than 3 seconds if the faceplate has been already called up at least once for that point. The first time call-up takes about 6 seconds (it needs to resolve NAME to Internal ID conversion).
If the client will not accept the performance I will try to boost the OPC connection by setting the priority to a higher level.
I'm not sure if they already assigned a higher priority level to the faceplate.
As far as I know, you can have only one priority level with one OPC connection. So, I don't know how to set different priorities to custom displays and faceplates.
"That is how I would have approached it,"
With this approach, the displays gets "validated" at invocation. (With the GUS pictures you could validate once.) If the OPC checkpoint file gets corrupted, than you will experience at the next invocation an other validation of the display.
I am very experienced with Experion SCADA and am just installing a Experion over TPS/Experion Controller soon...
When you build the Experion server you specify the frequency rate for Fast (the old 5sec history but can be 1 sec) for up to 1000 points, Standard and Extended history.
The only thing to watch from my perspective is that the graphics side is a little limited and if you want to do anything in VB script you will be very frustrated.
Write to me direct if you would like some more info.