I posted this over on the VWS forum and thought someone here might also benefit from it.
For the advanced tinkerer...
My VP2 Plus has always been a little cranky with comms between the ISS and the console. The ISS is out by the driveway, and the console is in the living room next to my recliner. The wiring is over direct-bury Cat5 patched to regular Cat5 on a 66 block in the garage. From there, it's patched on another 66 block in the Brain Room to a Cat5 that runs to the console location. The distance has to be well over 100'.
In the beginning, I thought the problem was interference from all the radios, scanners, controllers and the laptop at the console location. The console would lose comms to the ISS seemingly randomly. Coiling up the cables and adding ferrites seemed to fix it.
Then I changed the weather computer a few months ago and it went to heck again. There is another Cat5 cable for the serial from the WeatherLink back to a 66 block in the Brain room, patched to another Cat5 to the weather computer in a different room. The console would work fine with the ISS as long as the computer was not connected.
I should note here that all of this in-house wiring worked flawlessly with my original Weather Monitor II.
I did a lot more troubleshooting, trying different serial cables snaked through the house, even a pair of RS 232/485 isolated converters, etc. Nothing worked. I was reduced to moving the console into the computer room, where it worked fine, and I could keep VWS and SLOweather going.
I wondered if there was perhaps a ground loop problem but Davis wouldn't part with the schematic for the comms on the VP2 so I could look for possible interactions in the console. I did agree with the tech that, since the ISS comms were an RS422 differential pair, there shouldn't be any problems, but he had no troubleshooting suggestions for me.
The big puzzle was why it would work until the computer was connected and then ISS communications would fail. I've worked with 232, 422 and 485 communications for 15 or 20 years, and this was really frustrating.
Finally, after a little more testing, I realized that the common failure element was the ISS/console cable from the Brain Room to the console. Today I took apart the connections at each end to make sure I didn't have a pair crossed and to re-terminate them. Sure enough, I found a single broken data wire on the jack for the ISS in the wall by my recliner. In the interest of completeness, I completely repunched down the wires on that jack and cleaned up the other end on the 66 block in the Brain Room.
So far, so good. For a few minutes longer than it's taken me to write this, the console is talking to the ISS and VirtualVP is talking to the console, all over the house cabling. I can probably also remove all of the ferrites now, as well.
Since the RS 422 line was not properly connected, I'm amazed that it worked at all. And, I have had problems with lightning in the area locking up the ISS. I hope this fixes that problem as well.
