Log of weather station trials and tribulations

Back to orf5.com

[12/5/05] I recalibrated the humidity sensor (2.5 clockwise turns on the potentiometer) on 12/4 as I was consistently reporting dewpoints 4 or so degrees F too high. I will keep my eye on it for the next couple of weeks and decide whether any further changes are needed.
[1/27/06] My new computer which I recently hooked up to the instrumentation melted down and I am offline for at least a week while the manufacturer repairs it.
[2/14/06] Computer is finally is repaired, and we are reporting data again.
[4/2/06] Had a problem with wind direction due to corrupt register on unit, fixed today. Maybe that will also fix the Mach 1 winds that are sporadically reported. Also am now reporting in "rapid fire" mode to the Weather Underground
[4/16/06] The last couple of weeks have shown bad temperature and dewpoint data due to water getting in a female-female modular plug junction. The wire had been just lying on the lawn and I had kludged a little PVC pipe + plumber's putty shield for the junction which lasted a lot longer than it should have, actually. Yesterday I finally redid the entire run of wire with clamping connectors for splicing (cut off the modular plugs) and ran it all through conduit which will be buried soon. No more water problems! I am still seeing spurious rainfall rate spikes. These are happening in software. I am considering writing my own code. I am using code written in Perl for the Ultimeter 2000 which I have been tweaking. Final note, I am back to reporting data every 5 minutes instead of "rapid fire."
[4/19/06] Finally found the source of my "spikes" - occasionally the raw data stream from the unit itself is corrupted with an extra character. This messes everything up beyond where the extra character is found. I confirmed with Peet Bros that this is not normal and I will most likely get the unit repaired/replaced. In the meantime, I have found how to deal with it in software and will collect some bad data for him before deciding where to go from there.
[5/16/06] Regretfully I am sending back the control box for the 2100 and the anemometer/wind vane for repair. I have never been satisfied with the winds - the cup anemometer is sticky. After seeing what a much more professional system can do I simply can't live with this performance. And while I have dealt with the data stream corruption in software, I am concerned it will get worse with time. So, you won't see any data from me for the next couple of weeks.

On another note: I tore up all the phone line and wired the temp/humidity and rain gage all with one run of cat-5e. Being red/green colorblind made it difficult but I actually got all the wires right the first time! No more splices, and one cable contains all three instruments. Also, I put the wire in 1/4 inch PVC conduit, cemented the conduit together, and buried it. Simply elegant. I noticed after doing this my relative humidity values seem biased a bit low. Hence, I will probably end up doing another recalibration after I get the system repaired (see 12/5/05 entry)
[7/2/06] At the Dayton Hamvention I visited the Peet Bros booth and they were unveiling a new anemometer/wind vane which addresses the low-wind problems I'd been having! So I called Peet Bros and had them change my repair order to a replacement/upgrade order. It involved a firmware upgrade on the controller box. Anyway, I finally got the upgrade and control box yesterday. However, they neglected to send back the base/cable part of the setup that I had sent back as part of the repair!! ARGH! So tomorrow I will call them and request that they send me one ASAP. So as of today I'm back online (wunderground only not CWOP) without winds.
[7/16/06] New anemometer is finally up and running. Adjusted humidity pot 3/4 counterclockwise turn (to the left); dewpoints have been low - this is a correction against my 2.5 turns to the right I did last year. Will decide whether to recalibrate again in a couple weeks.
[2/28/09] Instrumentation has been running flawlessly. Only problem currently is that the computer directly hooked up to the instrumentation is unstable, rebooting every couple of days. This can lead to strang things happening, especially in the precipitation reports.
[3/7/10] Flawless operation still. I disabled the wireless on the routerboard (problems with the Linux driver) and instead have a WAP plugged in. The routerboard has 100 day plus uptime and only goes down when I upgrade.


Back to orf5.com