[FUN] Host requirements

[Stephen Mitschke]

In addition to the requirements identified in the HIST specification, whattypes of things should Erwin keep an eye open for folks?Ian> Does anyone have any generic guidelines for specifying FF hosts? We will> be> implementing about the equivalent of 12,000+ I/O fieldbus metwork. What> would be the issues for this size network?> > Please reply to this list. If the reply is very involved, please contact> me> directly.> > Erwin Icayan> eicayan@bechtel.com> 509-371-4735

[Stephen Mitschke]

> Hi Ian/Erwin,> > Here are some of the things you should keep in mind in selecting a Host> System:> > (1) First and foremost is, use Foundation approved Host (meets HIST> specification) as far as possible.> > (2) Almost all the FF Host system vendor (whether it meets HIST specs or> not) test their system with approved FF devices (transmitters, flow meters> etc.) available on the market, on their own. So the specifying engineer> should check the list of approved devices with the particular Host system> they intend to use before making decision (i.e. getting the FF check mark> is> NOT enough for interoperability!!!). Firmware revision of the devices are> also an important thing to check.> > (3) Host system should be able to detect and automatically assign a> "device no." (I might have used the wrong name but, this no is in addition> to the user assigned actual tag name) to all the devices on the network.> > (4) Compare the available tools with the system software, e.g.> configuration, scheduling, documentation, simulation etc.> > Other things to watch out for:> > (1) Document your segment design well and KEEP IT UP-TO-DATE AT ALL THE> TIME.> > (2) While selecting a field device consider it's current consumption and> the time a device will take to execute the function block it has. This is> the feature which will differentiate between men and boys.> > (3) For a size of this system it is very important NOT to assign devices> randomly to segments. Most of the time the temptation of adding a device> to> the closest segment (for a lower wiring cost, as it seems initially) gives> grief in a long run. As more "unrelated" devices are wired to a segment,> the> segment becomes hard to maintain (and document!!) and a impact of a small> change may affect more unrelated equipment.> > Bye for now.> > Kalpen> Paton Controls Limited

[Stephen Mitschke]

> Experience (installed base) is an important selection factor. The features> Fieldbus Foundation test is another one. But a very important one is how> the> system handles an increased amount of information. With 4-20 mA the data> base for 10,000 points has a certain size. For Fieldbus it changes by at> least one order of magnitude. The other aspect is how the DCS internal bus> handles more data.> > Marcos Emerson Process Solutions