I noticed that Synergetic Micro Systems is moving from boards to chips with the introduction of their EC-1; an all in one fieldbus chip. http://www.controleng.com/archives/2001/ctl0201.01/0102tu.htm I guess they did not know who was going to win the fieldbus war so they decided to include CAN, Profibus and Ethernet all on one chip. This is a variant of the "any bus" concept and seems to be pushing the fieldbus technologies into a type of commodity. Are fieldbuses now regarded as a commodity and have they become cheap enough to compete with hard-wiring? George M. Thomas Tel: +1-630-963-7070 x100 Contemporary Controls Fax: +1-630-963-0109 2431 Curtiss Street [email protected] Downers Grove, IL 60515 http://www.ccontrols.com

 

It's a natural progression - expect a dozen such chips with various fieldbus combos out by end of 2001 (if you ask me). Most interesting option is a company making a product with a medium price tag (say $900 to $5000 range) like a bar-code printer or vision system can now offer a single product which can be a slave on a DeviceNet & Profibus & Ethernet as-is. No need for the "any bus" style add-in card anymore as these 3 options - along with an RS-232/485 port - will satisfy most users now (99%?). On lower cost products the added overhead of all the isolation transformers & connectors make this harder to justify. "cheap enough" is always a relative idea. I talked to the manager of one semiconductor OEM who makes large processing machines/cells for fabs. He says the switch from hardware wired to DeviceNet cut the wire-harness phase from 12 to 3 weeks on most cells. How do you put a real price on those 9 weeks? No doubt a dozen accountants could come up with a dozen different answers. Add to this the change in field-support & troubleshooting (some good - some bad). Add to this the "PR" value of "Oh, we have a WORLD STANDARD inside there" when selling - especially if much of the system is still $$$ proprietary. Add to this the ability to ADD options in the field - to add a series of optional sensors to a hardwired system either means you need to prewire the harness just in case a user wants this - or have this wiring adding on-site in the field. With a fieldbus, you can just pop in more sensors & worst case extend a bus run. That's far easier than running dozens of hard-wired runs thru an existing machine cell (especially one in a clean room!) Regards Lynn August Linse, Senior Product Application Engineer 15353 Barranca Parkway, Lantronix Inc, Irvine CA 92618 [email protected] www.lantronix.com Tel: (949)300-6337 Fax: (949)453-7132

 

Hey George and Lynn, This is an interesting subject for me, because as an employee of Synergetic (I also used to work with HMS Anybus) the most articulate customer was a guy who made motion controllers, who complained, "usually when you add a feature to a product (i.e. a new car) you can make more margin on it than you make on the car. With networks, you sell the added feature at cost with no markup, just so you can get the order." Putting a $200 daughter board inside a product to do COMM works OK, but everyone complains about how much $ it costs. Now it's hard to get around the connector issues, certainly, and we can't do much about that. But we felt the important factor for a lower priced controls product (just a few hundred $, not $1000+) was that the MFR could probably use the communication chip also as a processor, and save a lot of parts in the design. In that case, the only added cost may just be the transceivers and connectors, the communication and software stacks are already part of the package. I think the cost of networking has to drop to that point before fieldbuses are as accepted in factories as modems are in PC's. Hopefully we're starting to get there. Perry Sink