Topic: WittBoy = GW2001 (station) = WS90 + GW2000 on Kickstarter & Beta test

[Ra1nman]

For anyone who's interested, comparison of wind plots today, WS90 top, very close nearby station with standard anemometer bottom. Of course I can't vouch for the height of the other station but gusts seem fairly close, suspect the WS90 is just showing more sensitivity in between

Wind direction has been mostly the same all day, so this won't be causing any issues with results.

[Ra1nman]

Posted this in another thread https://www.wxforum.net/index.php?topic=44811.msg455005#msg455005 if it helps anyone then all good:

Initial 7 day review of Wittboy station...

Introduction

It has been 7 days since my Wittboy (WS90 + GW2000) has been running now, thought I would note down some initial feedback. I have owned a couple of basic fine offset weather stations over the years, the odd wireless rain gauge and cheap wireless temperature sensors!

History

Have admired the Davis VP2 Pro for many years, but, in the UK it would be around £600-800 for this setup, likely £1100+ for solar/UV bundle. For me that is a lot of money on a weather station. Now and again I have glanced at options on Amazon but nothing too much seemed to change. In December this year I stumbled on the Ecowitt brand, fast forward to the Wittboy mega thread and a WS90 combo was ordered.

Setup and build quality

Setup was a breeze, Energizer Lithium AA's inserted, some silicone grease on the rubber seal just in case. Upgraded firmware via USB and screwed onto the pole, job done. The GW2000 connection was painless also, done in about 2-3 minutes, all set and reading data from the WS90. The WS90 seems pretty solid to me, I have engineering background and am pretty critical! the GW2000 looks a bit 'cheap' in comparison but works as intended.

I spent some time looking over the ecowitt.net site, similar the iOS app, all seemed OK, main text on the app and website seemed a bit small, not so graphical but functional and with history.Days passed and I played with CumulusMX again, looked into Smartmixin app, was a little frustrated the Ecowitt app didn't update realtime (8.8 seconds) but hope that will come with time. Generally the data available is great,graphs are well laid out, no problem.

Rain and Wind readings

Sadly I don't have much to compare accuracy against, manual rain gauge and that is it, generally the rain seems pretty close, I have tuned gain a bit, not much, tracks so far pretty good with the manual gauge. I suspect light rain is going to be the area for concern, also heavy dew potentially, but, lets see how that goes. Wind seems fine, I suspect maybe under reading a bit compared to what my eye's see but close enough when compared to very local weather stations, better in the sense it picks up the low strength winds which they often miss.

Solar and UV

Now I like seeing the solar and UV, OK, maybe they aren't calibrated as such but it is nice to see, gives me an idea of what is going on, also good to link this with capacitor charge and voltage. Can easily see if it has been a real dull day or if the sun has been out for several hours, not essential but it was on my list if the Davis ever happened!

Voltage readings

Having access to voltage for AA's and capacitor, plus soil moisture or other sensors is also useful, I added a soil moisture because, well, why not, they cost less than a good meal out and around 10x less than a Davis equivalent. It works in line with rainfall, after a dry day the moisture drops from maybe 68% to 62%, so you assume it is working as intended. This should prove interesting in summer when ground really dries out.

Temperature, humidity and accuracy

Temperature accuracy I cannot say too much, I cross check various readings with local Fine Offset station and not quite so local Davis VP2's and they are close enough for me, within +/- 1 degree essentially, often less. Similar for humidity but this can be maybe a little more out. I have a WH32EP sensor awaiting install into a stevenson style shield, this should I hope cover that area pretty well for accuracy, also the height above ground will be addressed.

Conclusion

In conclusion I am pretty happy overall, I paid around £200 for the setup, soil moisture was £20 and WH32EP £55. My biggest concern is whether the setup will still be running in 2 years time, I hope so. In the meantime I like the fact it has no moving parts, a lot, accessing the data is easy, I've got an old iPad Mini acting as console for now, in general, it is nice to have a weather station running again. Hope that this might be useful for someone, this forum has been such a help for me while researching, contributors spend a lot of time testing and it is appreciated.

[Ra1nman]

Does anyone have gain adjustments for wind on their WS90? almost certain mine is reading low, have set gain to 1.3 for now, trying to benchmark against closest local stations.

[Gyvate]

I'm not sure if this is a good idea - you'd first have to check if the wind sensors of your "close" local stations are mounted at the same height above ground and have the same free area around as yours. Otherwise you might compare your microclimate with their microclimate which could easily be apples and pears.
But for an order of magintude comparison using the AWEKAS service for that purpose could imo be helpful.

[Ra1nman]

The only thing I can be sure of with the local PWS is they are in relatively built up areas and lower ground than myself. Sadly that is all I know, and of course brand of PWS, this is a shame, height of anemometer would be nice to know. With that in mind I figure they should if anything be reading lower than mine, but, maybe that will not necessarily be true. My location is essentially in the middle of a field quite high ground, no obstacles near by.

I will look into the AWEKAS, appreciate the information and help as always

[Gyvate]

The only thing I can be sure of with the local PWS is they are in relatively built up areas and lower ground than myself.

Depending on how that really looks some sort of wind channels might be created where higher speeds can be created compared to free exposure, not to speak of turbulences.

[Ra1nman]

The only thing I can be sure of with the local PWS is they are in relatively built up areas and lower ground than myself.

Depending on how that really looks some sort of wind channels might be created where higher speeds can be created compared to free exposure, not to speak of turbulences.

I did wonder if that might have been the case, my WS90 did register 37mph gust a couple of days ago, so it is picking up higher speeds. Now I see how people end up with three different weather stations and various sensors, testing and comparing can no doubt get out of hand. My aim is to try and get this current setup running as accurately as possible, the WH32EP should cover the temperature and humidity at least. I will dig through this thread and look for anyone who has compared wind readings with other stations on the same site, pretty sure I read some previously but can't be sure. Don't remember the WS90 being far off for wind accuracy otherwise I'm sure it would have stuck in my mind.

The fact that the wind speed is similar to other stations does lend more weight to your theory about wind channels, turbulence and such like with wind gusts. Do wish I had something like a Davis or similarly calibrated anemometer to compare with. Plans for the future are to raise the height of the WS90 once I have offloaded the temperature, humidity and rain, also moving it to an even more exposed location.

[Rover1822]

Wind speed is one of the hardest things to measure, getting consensus values from nearby stations is near impossible, ballpark values, maybe.

And honestly, if you look at wind direction from nearby stations, if you are like me, you will find them all over the place as well

A good place , if you want to pursue this on turbulence, height , etc., are the DIY / home site wind turbine forums, because it really matters there. (Along with the other obvious , aircraft, fishing, sailing vessels etc). My turbine is at 30', and I can't legally go higher , it needs to be at 60'.  I did it as a hobby, but , I have learned a lot.  There is a huge amount of turbulence near ground level , and the height you have mentioned will not escape it.

Also you mention the residential area that you reference against. They may be roof mounting, or something , so hard to compare. Also as Gyvate mentioned, the tunnel effect, wind hits a bunch of buildings and accelerates. Large cities with sky scrapers have this issue bad, but it exists even at low level communities.

Or you could just go out and buy a whole bunch of sensors (wait, some of us do that) , and try to figure out what is correct. 

[Gyvate]

The fact that the wind speed is similar to other stations does lend more weight to your theory about wind channels, turbulence and such like with wind gusts. Do wish I had something like a Davis or similarly calibrated anemometer to compare with.

A Davis or a Barani MeteoWind will meet the same topographical conditions and impact - unless you put one close to your WS90, it won't give you much.

[Ra1nman]

Wind speed is one of the hardest things to measure, getting consensus values from nearby stations is near impossible, ballpark values, maybe.

And honestly, if you look at wind direction from nearby stations, if you are like me, you will find them all over the place as well

Thank you for the comprehensive reply, using local stations as a gauge is always going to be tough. Really I need to get the WS90 higher and it will most likely happen, along with relocation to an even more exposed site, currently the North route is partially compromised with a building, once I relocate it, all surroundings will be clear.

You can most likely put me in the whole bunch of sensors camp, it wont be long I am sure

Thanks for the wisdom, it is greatly appreciated.

[Ra1nman]

The fact that the wind speed is similar to other stations does lend more weight to your theory about wind channels, turbulence and such like with wind gusts. Do wish I had something like a Davis or similarly calibrated anemometer to compare with.

A Davis or a Barani MeteoWind will meet the same topographical conditions and impact - unless you put one close to your WS90, it won't give you much.

If I ever end up with another measuring device then for consistency I will try to mount it alongside WS90

[giusCB]

 [ You are not allowed to view attachments ]
snowstorm, no problem so far , apart the rain gauge that was dead months ago

[BaseLine]

Not really designed for the winter, is it?  (but none of them are)
 [ You are not allowed to view attachments ]

External PV panel? Probably useless too because when it's needed the most, the sunlight is so weak a tiny PV panel cant really charge anything.

[Rover1822]

I'm going to guess, if you have the heater hooked up, it has no effect on top frozen material .

This would answer a question posted in another topic

[BaseLine]

No heater. It was not possible to remove the plug/cord and there is no power outlet anywhere close enough. There is one but that's 100 m.

[Adog33]

Had my GW90 inaccurately report rain events. Is this a common issue?

Could a bird sitting on the station cause this?

[Gyvate]

Had my GW90 inaccurately report rain events. Is this a common issue?

Could a bird sitting on the station cause this?

A GW90 doesn't exist  (it's a GW2001 station consisting of a GW2000 console and a WS90 outdoor array)
The haptic rain sensor (also called piezoelectrical rain sensor) of the WS90 has not yet reached high maturity when you want to have 0.1 mm accuracy at any moment.
It can be calibrated to some extent, but it's a dificult and long exercise, especially as most of the deviations happen in the low range (0.1 - 2.0 mm) and are not linear.
Also, calibration depends on the firmware version of the WS90. Plus, depending on rainfall type it can vary to quite some extent.
Ecowitt are working on it ...
But in principle, bird dropping on the "right" place may have an impact ...
Interestingly, on a larger scale, the deviations are not that bad. Let's say the corridor in which the WS90 moves with its readings compared to an Ecowitt WH40 rain gauge (with rim extension) and a WH65/WS69 rain gauge (the 7-in-1 outsoor array) is not so wide and even more interestingly irons out in a weekly or monthly view. I have them all three (plus two manual rain gauges) and it's imho actually not as bad as some accuracy purists here say.
Have a look at http://www.meshka.eu/weewx and draw your own conclusions. Have a look at the daily, weekly, monthly and yearly graphs and tables.
the WS90 charts start in June 2022 only ... that's why you will see a bigger gap in the yearly charts/tables.

The best way to "serve both masters" is to run the WS90 together with a WH40, which is possible in parallel.
A HP2560 console can only show/display either of the two at the same time, whereas a GW1100/GW2000 can show them both in the WSView Plus app.
The ecowitt.net cloud can show both (piezo and classical) rain gauges in parallel on one page

[Adog33]

Thanks for all the information. I’ll look at adding a WH40 into the mix. I have had good luck with the WS90 until this point. I have a feeling something made physical contact with the array.

Here’s my site: Timelapse’s update every hour & 24 hours.
https://precisionconditions.com/pwsWD/index.php

Had my GW90 inaccurately report rain events. Is this a common issue?

Could a bird sitting on the station cause this?

A GW90 doesn't exist  (it's a GW2001 station consisting of a GW2000 console and a WS90 outdoor array)
The haptic rain sensor (also called piezoelectrical rain sensor) of the WS90 has not yet reached high maturity when you want to have 0.1 mm accuracy at any moment.
It can be calibrated to some extent, but it's a dificult and long exercise, especially as most of the deviations happen in the low range (0.1 - 2.0 mm) and are not linear.
Also, calibration depends on the firmware version of the WS90. Plus, depending on rainfall type it can vary to quite some extent.
Ecowitt are working on it ...
But in principle, bird dropping on the "right" place may have an impact ...
Interestingly, on a larger scale, the deviations are not that bad. Let's say the corridor in which the WS90 moves with its readings compared to an Ecowitt WH40 rain gauge (with rim extension) and a WH65/WS69 rain gauge (the 7-in-1 outsoor array) is not so wide and even more interestingly irons out in a weekly or monthly view. I have them all three (plus two manual rain gauges) and it's imho actually not as bad as some accuracy purists here say.
Have a look at http://www.meshka.eu/weewx and draw your own conclusions. Have a look at the daily, weekly, monthly and yearly graphs and tables.
the WS90 charts start in June 2022 only ... that's why you will see a bigger gap in the yearly charts/tables.

The best way to "serve both masters" is to run the WS90 together with a WH40, which is possible in parallel.
A HP2560 console can only show/display either of the two at the same time, whereas a GW1100/GW2000 can show them both in the WSView Plus app.
The ecowitt.net cloud can show both (piezo and classical) rain gauges in parallel on one page

[giusCB]

DONE ! 1710m slm. at Roccaraso Aremogna Ski Resort
I don't honestly know what will happen to the sensors. 
The area has a extreme climate and the wind often exceeds the limit of the anemometer (144kmh). If it works here, i think will works in every place...
Live data here: https://www.turismometeo.it/stazioni-meteo/roccaraso-aremogna-ski-area

The station is completely dead after about 4 months.
The battery charge suddenly discharged along with that of the supercap. As in many other cases already reported here, the ws90 suffers from problems of water infiltration and mouisture inside the PCB. There is clearly an error in the design or assembly phase. 

I think next time i'll test the ws69

[ilc]

I've just had a 6 day outage with my WS90/GW2000. It is remote so I haven't been able to check it, but it has come up again of it's own accord.

Any ideas what might have caused it? Are they prone to intermittent outages?

Thanks, Ian.

 [ You are not allowed to view attachments ]

[Gyvate]

as it has picked up reporting to ecowitt.net again as it did before - and hadn't been touched, it rather looks like a network issue than a station issue.
If you configure your remote router accordingsly (e.g. port forwarding), you can access your GW2000 WebUI also remotely and look up its configuration etc.

If you can't connect in such a situation, you have obviously a network issue, either locally (at the remote location) or with the internet access. Difficult to say just from the picture.

[ilc]

as it has picked up reporting to ecowitt.net again as it did before - and hadn't been touched, it rather looks like a network issue than a station issue.
If you configure your remote router accordingsly (e.g. port forwarding), you can access your GW2000 WebUI also remotely and look up its configuration etc.

If you can't connect in such a situation, you have obviously a network issue, either locally (at the remote location) or with the internet access. Difficult to say just from the picture.

Thanks for your reply. I could access security system (which is on ethernet) so I know the network was good to the router, however the GW2000 is on WiFi. Unfortunately the router connects to the internet via 4G and therefore CGNAT prevents me from logging into the router directly. I can access some details via a monitoring website, and can see that the GW2000 connected to WiFi now, I'll check this first if it happens again.

I should also have mentioned that it went offline during a minor rain event after 2.2mm at 13.2 mm/hr, and I have read some people have had problems with water ingress. I guess time will tell.

[Gyvate]

I should also have mentioned that it went offline during a minor rain event after 2.2mm at 13.2 mm/hr, and I have read some people have had problems with water ingress. I guess time will tell.

if there was water ingress at all - my two WS90s don't show any with much stronger rainfalls over one year - it would have been with the WS90, (hopefully) not with the GW2000. So the GW2000 should at least have continued posting air pressure and indoor T/H if the WS90 (outdoor sensor array) had had an outage.

[ilc]

I should also have mentioned that it went offline during a minor rain event after 2.2mm at 13.2 mm/hr, and I have read some people have had problems with water ingress. I guess time will tell.

if there was water ingress at all - my two WS90s don't show any with much stronger rainfalls over one year - it would have been with the WS90, (hopefully) not with the GW2000. So the GW2000 should at least have continued posting air pressure and indoor T/H if the WS90 (outdoor sensor array) had had an outage.

Yes, good point, the GW2000 is inside and should continue to report. So obviously a WiFi issue between the router and GW2000. It's the only device on WiFi when we aren't there so I don't have any other devices to check, but I'll be as little more analytical if it happens again.

Thanks again fore your help.

[giusCB]

DONE ! 1710m slm. at Roccaraso Aremogna Ski Resort
I don't honestly know what will happen to the sensors. 
The area has a extreme climate and the wind often exceeds the limit of the anemometer (144kmh). If it works here, i think will works in every place...
Live data here: https://www.turismometeo.it/stazioni-meteo/roccaraso-aremogna-ski-area

The station is completely dead after about 4 months.
The battery charge suddenly discharged along with that of the supercap. As in many other cases already reported here, the ws90 suffers from problems of water infiltration and mouisture inside the PCB. There is clearly an error in the design or assembly phase. 

I think next time i'll test the ws69

I have been informed that the ws90 currently suffers from a bug which, from what I read here, has also been experienced by some of you. The chip that manages the RF part of the external sensors suddenly starts consuming a lot of energy from the sensors, knocking them out (AA batteries and supercapacitor drained). Currently, ecowitt looking for a solution (hopefully Henry will give us some more information). I will be sending my ws90 to the factory for analysis.