I got an email a friend last week, with this simple question...
"What's your take on reliability of programming [Brand A's] instruments via Ethernet & GPIB?"
This friend is a Certified LabVIEW Architect, and designs and implements a variety of test and measurement applications for her customers.
She's also one of my best friends, and her livelihood depends on my honest advice.
Here's what I told her...
1) GPIB just works. It's proven itself reliable and robust over a few decades.
2) LAN, I think, can be reliable, too. It's proven technology, obviously.
3) You'd want to put some thought into the network architecture. For example, I'd probably want to find a good network hub, and configure static IP addresses on the computer and all the instruments.
4) Does the host computer have to also be on a corporate network? If so, I'd have two network cards in the computer--one for the corporate network, and one for the instrumentation. In this case, I would probably ask corporate IT if they want to advise on any network settings/requirements they would like to see, since the computer is acting as a router in this case.
5) Related to #4, are there any security requirements to be aware of? E.g., would you need to block access from the rest of the internet trying to access your instruments? It can be done, but you have to know to set it up.
6) Is performance a concern? What kind of traffic is going on between instruments and the computer? LAN is faster for bulk data transfer. GPIB is faster for small, simple readings.
7) LAN allows for some interesting possibilities for network synchronization, web access, long distances between devices, and other features. If you think you need those, we can talk further.
Summary: Both work. If GPIB is sufficient for your application, I'd be inclined to start there as a path of least resistance. But I wouldn't be afraid to use LAN; I'd just go in expecting it to be a little more complicated.
Read more of this article...
