Cumulus Support

OK - I have another issue to go with the UV one in this thread https://cumulus.hosiene.co.uk/viewtopic.php?f=6&t=7022

What I am seeing is the Wind is flicking to exactly 359 Degrees momentarily (i.e. one read cycle / 2.5 secs) and then back to the predominant direction (which has bee W to E for the last few days) - on the console and Cumulus thus the trends also. Its something thats getting worse - not better.
This excursion time varies greatly. It can be frequesnt 5 secs to 60 Secs, or more infrequent - 30 mins to an hour etc.

So what I see is something that looks like this on the trends: Note the red + along the top of the wind dir trend - thats the issue.
You can see from the wind speed that its not something directly related to lack of wind speed. I have observed the vane and it is in the same general e-w direction (in fact mostly S when I looked), certainly no excursions North.
Plus its always going 359 Deg, not 358, 357 etc.

I have checked the wind sensor and it is a smooth 0 to 20K Ohm through its travel with readings of 0, 5K, 10K, 15k, 20K (almost) at the four compass points. It stays at these values steadily without fluctuation.

Has anyone else had this issue before?

Given my UV issue and this 359 Deg issue I am starting to think the Sensor Interface Module may be faulty (My station is the cabled version not the wireless version).

Anyway opinions appreciated - I am talking to the supplier who has been fantastic and is willing to help in anyway he can. So this is really to gather as much user experience for him as I can.

Cheers

Phil

An intermittent short in the cable somewhere, causing the max value to be registered?

Possibly Steve,

I have made up a test resistor assembly using 2 1% 10 K Ohm resistors on a RJ12 connector to simulate 180 Deg wind direction.

If its fine weather tomorrow I will plug it in instead of the Anemometer and observe over a few hours. This should isolate the problem to the Anemometer or the ISS SIM.

The only problem with this setup is whether the lack of wind speed signal may cause the VP console to interpret that as calm and force 0 degs, but if it does then thats not 359 so should be able to interpret accordingly.

Phil

PS your dual axis addition is fantastic for me analysing my UV issue as it allows me to compare UV and Solar for obvious reasons. Perfect.
Is there a way though to turn off the Axis lines, it gets a bit messy with both scales selected (maybe a check box?).

OK, I did my test.

I located two Dale 0.05% 10K precision resistors and created a simulated Wind Dir signal at 180 Deg (10K = 180 Deg) test harness.

I also added an RJ12 female connector so I could plug this test harness in series with the anemometer and the ISS SIM. This meant the Wind Speed was connected to the ISS but instead of the Wind Dir POT being connected my 10K signal was. If you don't do this the Avg wind direction comes up as 0 Deg as the wind speed is zero.

End result was Console displayed 178 / 179 degs (flicked between the two values every few seconds) and also displayed actual wind speed. The flicks to 359 no longer occurred. I left this test on for about 5 1/2 Hrs and then removed the test harness. Interestingly the flicks to 359 did not immeadiatly resurface, it took about 1/2 to 3/4 Hr, and have done ever since.

Here is a screen shot of my test. End conclusion is the Anemometer POT or cable must have an intermittant fault.

If anyone is interested in my test harness I can post the schematic.

Cheers

Phil

Issue still happening and there weems to be a link between time of day and severity - worse at night.

Anyway I have just stumbled across this post in another forum:

http://forums.accuweather.com/index.php ... =8701&st=0

It appears to be exactly the same issue I am having with the 359 Degree reading and it points to the ISS being faulty - but this does not line up very well with my testing on the weekend.

But if it is the ISS then it may also solve the UV issues as well.

Phil

Well I may have gotten somewhere with this issue too.

I received a replacement Anemometer (and ISS) last week and installed the Anemometer on the weekend. However to my disappointment at the time, it did nothing for my problem. I was still receiving frequent excursions to 359 degs. So my plan was to replace the ISS this weekend.

However I did also uncover a theory that by running cumulus on a laptop, the lack of suitable grounding may cause this issue (I think it was on the Weather Watch forum). Anyway I thought on Wednesday night that it would not hurt to try running Cumulus on a desktop PC as they are obviously is well grounded. So I built up an old machine to run till the weekend.

Anyway since I did this I have had no 359 excursions except ones that occur when the wind is tending to the North which can be expected. At all other times not a single one. Since using the PC, the longest I have gone between consecutive 359's is about 26 hrs - a huge record for me. See the screen shot for the last week attached.

It is a bit early to claim success (the other forum that suggested it - did not turn out to be the issue, it was the ISS) so its very much a waiting game. The thing that has me wondering is surely I am not the only person running Cumulus on a laptop to a cabled VP2+ Maybe something is wrong with the laptop PSU.

At least I have a spare ISS should my luck fail.

Cheers
Phil

Phil, how is the Davis powered? Presumably on it's own PSU?
I've hit similar problems a few times (outside weather stations) with connecting up kit that uses double isolated PSU's (no ground). Each powered item has its own floating ground reference, disconnect the data cable and put a meter between the grounds on your laptop and Davis - you may be surprised

One solution may be to eliminate one PSU and power the Davis from your laptop via a USB port - I think it takes a 5V feed and the peak current is well within USB ratings? Using batteries to cover times when the laptop is powered down.

mcrossley wrote:One solution may be to eliminate one PSU and power the Davis from your laptop via a USB port - I think it takes a 5V feed and the peak current is well within USB ratings? Using batteries to cover times when the laptop is powered down.

Mark, I do not think this will be possible.

A wireless Vantage Pro2 console has its own supply (batteries and external regulated DC 5v 200ma) - a wireless Vantage Vue console also has its own supply (batteries and external regulated DC 5v 200ma).

A cabled Vantage Pro2 console has its own supply (batteries and external regulated DC 5v 300ma) - unfortunately, I could not find much specific information about the cabled station.

Mark, you have to remember that the Davis Instruments weather stations are designed to be used on their own, without a computer. To extract the console data into a computer, a WeatherLink adapter (serial, USB, IP, APRS, Irrigation Control, etc) is needed.

My Vantage Pro2 console is connected to a computer through a WeatherLink serial adapter - my Vantage Vue is connected to a computer through a WeatherLink USB adapter. Neither consoles use power from the computer (as far as I know). I might be wrong but, it is my belief that the WeatherLink adapter get its power from the console.

Hi Ray, I meant to make a custom cable to take power from a USB port and plug it into the console DC input.

mcrossley wrote:Hi Ray, I meant to make a custom cable to take power from a USB port and plug it into the console DC input.

Ok! Do you think a USB port supply enough current to power the console (around 300ma)?

Just remembered the ISS (integrated sensor suite) has its own power (battery, superCAP and solar panel).

USB is rated at 500mA, the Davis PSU is 300mA, but the console draw peaks around 100mA I believe?

mcrossley wrote: but the console draw peaks around 100mA I believe?

Yes, and most of that (80mA) is for the light. If you don't turn the light on, peak draw for a wireless VP2 console is around 30mA and for a wired around 15mA.

gemini06720 wrote:Do you think a USB port supply enough current to power the console (around 300ma)?

300 mA seems way too high for the console. I run mine on the three C cells instead of the wall wart, and although I recharge them on a two-month interval, they're not very far down the discharge curve at that point.

Doubtless it is far far higher than otherwise when the backlight is turned on, and as I don't know what kind of backlight Davis chose to use, I won't hazard a guess.

Hi Guys,

I run mine on the standard Wall Wart that came with the WS. Only issue there was the USA style pins are straight, where as here in Aussie we like em bent
So I got out the pliers and bent em or I guess more correctly twisted them (after checking that it can handle 240 Vac of course). Worked fine, the pins did not show signs of cracking nor did the case.
Point to note is the official Davis Wall Wart is two pin (so double insulated) - at least for the USA version. I am curious to know if the official Aust / NZ version is also two pin.

I had previously tried running the WS on console batteries only and it made no difference. So if certainly looks like I need grounding on the serial interface for it to settle down. Remember the only way the unit is getting a ground is since I started using the desktop PC which is grounded and the only thing from the WS plugged into the PC is the serial cable.

I do not run any cable near AC power, I have made sure that there is at least 300mm (or 1 foot ish) segregation and where it does come anywhere near AC power it is at right angles (i.e. standard practice).

Cheers
Phil

PS No 359s over night.

archae86 wrote:...300 mA seems way too high for the console...

I was giving the specifications as found in the Davis Instruments documentation.

As specified in the Davis Instruments documentation:

• Power Adapter: 5 VDC, 300 mA
  Console Current Draw:
  • Wireless: 0.9 mA average, 30 mA peak, (add 120 mA for the illuminated display, add 0.125 mA for each optional transmitter station received by the console) at 4 to 6 VDC;
  • Cabled: 10 mA average, 15 mA peak (add 80 mA for the illuminated display) at 4 to 6 VDC;

• Battery Backup: 3 C-cells
  Battery Life (no AC power):
  • Wireless: up to 9 months;
  • Cabled: approximately 1 month;