[wxqc] Re: Steve - Vantage Pro barometer fix

[John Webb]

Steve,

Have you finalized your fix for the Davis barometer incompatibility? I 
would be interested in testing this to help it along. It would be nice to 
have a fix for this before I add another station at nearly 3,000 feet 
elevation. Anyone know what the status of this problem is with Davis?

John Webb
CW3212

At 08:39 PM 12/16/2005, you wrote:
>It's been known for awhile that the Barometer data sent to APRS/CWOP is 
>problematic because the sea level reduced pressure is submitted instead of 
>the expected altimeter pressure. This is most noticable at stations above 
>a couple thousand feet, expecially during high pressure periods. If you 
>use the barometer calibration to "fix" the problem, your results will 
>drift off when the high pressure moves out and low pressure returns.
>
>I've been writing a program to see if this is easily solvable, and 
>happily, it is (if the weather software developers put in the fix).
>
>If the software is controlling the com port, it just needs to get the 
>BARDATA occasionally (each 1 to 5 minutes). The data provided in the 
>BARDATA is sufficient to convert the VantagePro sea level barometer 
>reading to a station pressure, which can then be converted to an altimeter 
>value. The R reduction ratio in the BARDATA is not useful because it does 
>not provide sufficient precision to correctly reverse the SLP reduction 
>calculation. However, the other data in the BARDATA can be used to do 
>this. The thing that was confusing is that the data item called 
>VirtualTemp in the Davis docs is misnamed. It is actually the mean 
>temperature (average of current temp and temp 12 hours in the past). In my 
>test code, I was able verify the results, and they exactly match the true 
>raw station pressure.
>
>If the software is in listen only mode, where it cannot get access to the 
>BARDATA, the station pressure can still be obtained through calculation. 
>The hard part here was replicating the humidity correction value (the 
>Davis docs omit the tables they use). After some digging, I found a 
>formula that produces a close approximation of this value. With this, and 
>as long as the software can provide the temp from 12 hours ago from its 
>log, the station pressure can be calculated. In my testing, the results 
>were accurate to within .005 inHg or less (typically not more than  .001 
>inHg off).
>
>To quickly try this out, I added code in my test program to write to a 
>data.csv file every two seconds (with the altimeter value for the 
>barometer), and configured WeatherDisplay to use this as its input and set 
>WeatherDisplay to send data to APRS. My station is now providing good 
>barometer data. Now I will work on code to have my program send directly 
>to APRS so that people who only use WeatherLink can use it  to provide 
>APRS data with altimeter barometer data.
>
>One interesting thing I noticed was that once I started using the 
>altimeter value, I set the barometer calibration back to 0 and found that 
>the altimeter value was very close to that reported by the nearest 
>airport. I think it might be the case that most of the calibration people 
>do on their VantagePro barometer is because of this SLP/Altimeter issue, 
>not because of the barometer actually needing calibration.
>
>If anyone is interested in source code for incorporating this technique in 
>their software, or are interested in getting a copy of my program for 
>testing, please contact me.
>
>While I was doing this I also found a trick for forcing the VantagePro to 
>update the barometer more frequently than 15 minutes. I can share that as well.
>
>Steve