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

[BaseLine]

In this case, I wish they had some kind of PV panel with a battery I could use to power the WS90 heater.
There is no way to get power to where it is installed.

At what distance do you have a WS90 from your house/apartment?

Far enough. Hundreds of meters to closest AC outlet.

[Autofill]

because my front-end is Home Assistant, I created a two compensation sensors for the heating element. In my case, the temperature is upto 13% off when the heater is running, so for example, the reported temperature is -30oC, but the actual temperature is -34oC (as measured by my WS69 which is about 10 ft from the WS90).

The template sensor below only compensates when the temperature is below +5oC.

Now, with this new algorithm, both my WS90 and WS69 are at most 1oC apart....which is acceptable for me.

Code: [Select]

    ## Compensated Outdoor Temperature ##
    - name: compensated_outdoor_temperature
      unique_id: 0e1a1a97-2244-4ff9
      state_class: measurement
      unit_of_measurement: °C
      device_class: temperature
      state: >
        {% set temperature = states("sensor.temp") | float(0) | round(1) %}
        {% set negative = temperature < 0 | round(1) %}
        {% if temperature > 5 %}
          {{ (temperature) | round(1) }}
        {% elif temperature == 0 %}
          -0.13
        {% elif not negative %}
          {{ (temperature * 0.87) | round(1) }}
        {% elif negative %}
          {{ ((temperature | abs * 1.13) * -1) | round(1) }}
        {% endif %}
    ## Compensated Feelslike Temperature ##
    - name: compensated_feelslike_temperature
      unique_id: a9517d37-d3f0-4552
      state_class: measurement
      unit_of_measurement: °C
      device_class: temperature
      state: >
        {% set temperature = states("sensor.feelslike") | float(0) | round(1) %}
        {% set negative = temperature < 0 | round(1) %}
        {% if temperature > 5 %}
          {{ (temperature) | round(1) }}
        {% elif temperature == 0 %}
          -0.13
        {% elif not negative %}
          {{ (temperature * 0.87) | round(1) }}
        {% elif negative %}
          {{ ((temperature | abs * 1.13) * -1) | round(1) }}
        {% endif %}

If Ecowitt can add a heater temperature offset (not absolute temperature, but actual decimal integers) where it only applies when the temperature is below 5oC. The user would have to calculate how much of an offset should be applied, but I'm around 13% at -30oC or perhaps this number is general for all WS90s but others would share their experiences here.

[Autofill]

to give you an idea how long the heater stays on when the outdoor temperature is -20oC, here's the plot that shows power draw from my battery. After 5 pm, the power is completely coming from the batteries (not the solar panels), those 0.8 A current drops, is the heater turning on.

 [ You are not allowed to view attachments ]

The heater stays on for 10 to 11 minutes, then off for 4 to 5 minutes, then on again.

When its warmer, it stays on less, as you would expect.

[Arkadiusz_w]

I connected the heating at my second location yesterday to a WS90 Wittboy. I paired this additionally with a WIFI socket so that I could have full control over the under-roof heating and only turn it on when necessary, i.e. during snowstorms and freezing snowy weather when instrument icing is most likely and as I have experienced in the past.

I have noticed a strange anomaly in the data which relates to the behaviour of the lithium batteries. After running the heater, the voltage in the cells, which are only 4 months old, dropped after a few minutes. Previously, according to data from the Ecowitt app, they had between 3.24V and 3.28V. The capacitor at the time was 2.4V. Currently in Poland we have the shortest days of the year, the weather is cloudy, which does not allow supercapacitors to charge. Often the WS90 Wittboy has switched to battery power in the past weeks.

Within minutes the voltage dropped from 3.28V to 2.8V. In the Ecowitt and Wsviev app, a low energy message appeared. Data from the socket control app, indicated 240V and an intermittently switching on heater, which drew up to 12-12.5W, then switched off in a few minutes. Voltage was applied from the mains all the time.

Interestingly in the morning, the voltage dropped to 2.22V, fluctuating slightly. The heater was still running. After sunrise, when scattered radiation appeared, the cell voltage increased slightly to 2.4/2.5V. The capacitor began to charge slowly, despite the heating cable being plugged in. Remotely in the app from the Wifi socket, I had the power disconnected and switched to capacitor and battery power. After a few minutes the low level message from the 2 lithium batteries disappeared, the lithium batteries returned to 3.28V. I am wondering what is causing the current in the cells to fluctuate so much when the heater is switched on and disconnected. Should I be concerned about this or is it just a bug in the applications.

I also have a question as to whether the cable from the heater provides power to the station when there is no battery and the capacitor is discharged. I also present you with a screenshot of the cell behaviour in the WS90 and a link to the station where these unusual fluctuations occurred.

When I have the heating on, it records a slight increase in air temperature, which is measured with an SHT40 in the manufacturer's radiation shield. Turning the heating off remotely, results in less temperature fluctuation. The overestimation of the temperature in my elevated location, however, is no more than 0.2 deg.C/0.5.C higher.

I look forward to hearing from others whether they have encountered strange anomalies in the WS90 Wittboy.

Link to station where anomalies with cell voltage levels are observed: https://www.ecowitt.net/home/index?id=77800#

I use, together with the WS90 Wittboy and the heater power supply, the following WIFI socket, which allows me to remotely control power consumption, heater operation. Link below from Ebay for those potentially interested in remote control of the Ecowitt station.

https://tiny.pl/wflgx

[Gyvate]

I also have a question as to whether the cable from the heater provides power to the station when there is no battery and the capacitor is discharged.

As per Ecowitt the power supply cable when connected to the power grid provides permanent power to the whole sensor array including the heater.

[Arkadiusz_w]

I also have a question as to whether the cable from the heater provides power to the station when there is no battery and the capacitor is discharged.

As per Ecowitt the power supply cable when connected to the power grid provides permanent power to the whole sensor array including the heater.

It is how to explain the anomalies in voltage on two AA batteries. When I turn on the heater, after a few minutes the volts drop and it shows low battery. Is this a bug or is it true. Is there some kind of overvoltage?

Is there any contraindication from Ecowitt to using the lithium batteries and the cable from the heater supply in parallel?

[Arkadiusz_w]

I received information from Ecowitt regarding the heater in the WS90.

"If you have heater connected. The battery voltage will be disconnected, and thus it is not possible to measure battery correctly.
So this is normal, and it is not a failure of battery voltage measurement" - informs Ecowitt Support.

[BaseLine]

In this case, I wish they had some kind of PV panel with a battery I could use to power the WS90 heater.
There is no way to get power to where it is installed.

At what distance do you have a WS90 from your house/apartment?

Forget the PV panel idea. It would require absurdly large (few kW) of PV panels. There is not enough sunlight from Dec-Feb.
Far enough. Hundreds of meters to closest AC outlet.

[BaseLine]

Wiped off some snow from WS90

 [ You are not allowed to view attachments ]

[wsNoordbergum]

I also have a question as to whether the cable from the heater provides power to the station when there is no battery and the capacitor is discharged.

As per Ecowitt the power supply cable when connected to the power grid provides permanent power to the whole sensor array including the heater.

I also have the same issue with my WS90 after connecting the mains power the voltage of the batteries drops significantly
If it is just because the battery disconnects after mains power is supplied then there is no problem, guess the battery voltage will drop to the level of the capacitor, at least that is what I can see on the graphs right now

The power supply does indeed power the entire unit, as mentioned in the faq on their website

My station is https://www.ecowitt.net/home/index?id=82025

It is how to explain the anomalies in voltage on two AA batteries. When I turn on the heater, after a few minutes the volts drop and it shows low battery. Is this a bug or is it true. Is there some kind of overvoltage?

[giusCB]

strange problems today..
Yesterday the gw2000 lost the signal from the sensor, than it recovered again but without the data from the anemometer.
 [ You are not allowed to view attachments ]
I've re-register the sensor ID and immediately recover the data from anemometer. It's seems (the gw2000) it not decode correctly the wind after it wake up from signal loss....
beyond this, gusts at 150-160 km/h are expected next Tuesday 

I've had a lot of reception problems, and I've discovered they are caused by the gw2000 firmware version 2.1.9.
With version 2.2.0 they seem to have been fixed, but are still in testing... 
luckily i have remote access on the gw2000 , otherwise i would lose all the data...

the reception problems have been solved on the gw2000, i have no more data loss.

Big problems remain on the gw1100 version 2.2.0 with ws80 and wh40 sensors, i haven't tested with wittboy or other sensors. I've made a lot of test , and in the same location with same station, an old gw1000 works ok a gw1100 not 

[Ra1nman]

Well, I have read most of this thread so thank you to all  First time posting and a few 'cheap' stations over the last 10 or so years. Many times looking at the VP2 Pro, but in the UK the price of around £600-800 for me is just pushing it, especially for hobby use, their console which hasn't changed also annoys me somewhat! especially for the price. Anyway, every now and then I have a look around at options, last week I somehow stumbled across the Ecowitt brand on Amazon, few hours later I was all set to get the analogue style GW1103 station with display. Few more hours reading and the GW1102 with ultrasonic wind was on the radar!

That was when I stumbled on this thread, several hours later I was totally confused and back on the 1103 again, but, decided what the heck, lets go with the Wittboy with a view to adding WH32 and WH40 at some stage. My theory is that for the price I paid in the UK (£200) including the bird spikes, even if I do add WH32 and WH40, it will still be a reasonable buy for wind, UV and solar readings alone. I can see how the rain works out and suspect the temperature will be fine from the WS90 even mounted at around 4 metres.

Much as I like to mount the rain and temperature sensors at ideal height, initially I will put the WS90 at around 3-4 metres, maybe a touch higher, at least during setup and break in so to speak. I do have a radiation shield which was DIY, took quite a while to construct and was built to the letter! so would like to use this if possible. I understand that the WH32 is potentially less accurate than the WS90 for temperature so will also keep that in mind.

Anyway, the Wittboy will arrive in due course then let the games begin 

Managed to find some images of the shield, took me several hours to make, matt black under the sections that matter, aluminium mount, then installed a cheap sensor  technically it my not be correct but I got carried away at the time! moved the images to this thread to keep on topic here https://www.wxforum.net/index.php?topic=44756.0

[Gyvate]

I understand that the WH32 is potentially less accurate than the WS90 for temperature so will also keep that in mind.

Not necessarily - not when the WH32 is put into a proper radiation shield (e.g. Davis 7714 or some well done DIY).
A WH32 has a SHT31 T/H sensor.
A WH32-EP (with probe) has a SHT35 T/H sensor (EP = extra precision)
The WS90 comes by default with a SHT40 T/H sensor (bigger number does NOT mean better  - SHT3x = industrial, SHT4x = consumer market).
The SHT40 T/H sensor of the WS90 can be upgraded to a SHT35 T/H sensor.
The story is, even though the SHT35 in a WS90 works well, you always have the compromise between different temp/rain/wind measure locations.
Therefore the separate WH32(-EP), WH40 and wind (WS68, WS69, WS80, WS90) setup is optimal.

[Ra1nman]

I understand that the WH32 is potentially less accurate than the WS90 for temperature so will also keep that in mind.

Not necessarily - not when the WH32 is put into a proper radiation shield (e.g. Davis 7714 or some well done DIY).
A WH32 has a SHT31 T/H sensor.
A WH32-EP (with probe) has a SHT35 T/H sensor (EP = extra precision)
The WS90 comes by default with a SHT40 T/H sensor (bigger number does NOT mean better  - SHT3x = industrial, SHT4x = consumer market).
The SHT40 T/H sensor of the WS90 can be upgraded to a SHT35 T/H sensor.
The story is, even though the SHT35 in a WS90 works well, you always have the compromise between different temp/rain/wind measure locations.
Therefore the separate WH32(-EP), WH40 and wind (WS68, WS69, WS80, WS90) setup is optimal.

Fantastic information and knowledge, thank you kindly, so for maximum accuracy the WH32-EP is the way forward in a good shield, then pair with the WH40. I am all for better accuracy if it cane be done. Use the WS90 for wind, UV etc and the better placed sensors for temp and rainfall. I wonder why they don't fit the higher accuracy sensor to the WS90 in the first instance, cost I assume, seems a shame.

[Gyvate]

I wonder why they don't fit the higher accuracy sensor to the WS90 in the first instance, cost I assume, seems a shame.

Cost is always a driver.
I think it is so because it is sold as an "average user" customer model in the first place and not for the precision, accuracy and detail-loving enthusiasts  .
They have upgrading options.

[Ra1nman]

I wonder why they don't fit the higher accuracy sensor to the WS90 in the first instance, cost I assume, seems a shame.

Cost is always a driver.
I think it is so because it is sold as an "average user" customer model in the first place and not for the precision, accuracy and detail-loving enthusiasts  .
They have upgrading options.

That is fair I guess, price for the EP sensor is £55 UK but it does say accuracy for temperature of the WS90 is:

Temperature : -40°C to 60°C ( Accuracy : ±0.3°C (± 0.6°F); Resolution : 0.1°C (± 0.1°F) )
Humidity : 1% to 99% ( Accuracy : ±3.5%; Resolution : 0.01 )


The EP sensor is listed as:

Temperature accuracy: ± 0.2°C
Humidity accuracy: ± 1.8%

I suppose that is quite an improvement, also the ability to fit the EP sensor in a shield and at a suitable height 

Well I will add that to my shopping list since it will make use of my shield and I can be happier with accuracy. The display unit for the EP I guess goes in a sealed box or something so you can view it whilst outside.

The Wittboy even with additional EP sensor and rain gauge is still below £300 total for me in the UK, compared to a Davis which would cost me closer to £700-800 it still feels like fair value, not to mention the potential benefits from the wind sensor of the Ecowitt.

[Ra1nman]

A question which hopefully isn't too off topic, I plan to use the Wittboy with Ecowitt.net and also the iPhone app, I wonder what is the most visually pleasing way to use a spare iPad as permanent summary display? In the longer term I might end up with a 2560 console, but I have a spare iPad Mini which potentially could fill the void.

Any thoughts appreciated, I seem to remember seeing a tablet displaying Ecowitt data via Wunderground but could have dreamt that! it looked quite a good layout.

Thanks in advance.

[ivano]

Hi, my ws90 runs out of batteries from one day to the next, I replaced them last night at 20:00 and this morning at 12:00 they were already flat. what will happen? [ You are not allowed to view attachments ]  

[olicat]

Hi!

Do you use the external power supply?
Then the reported battery voltage cannot be trusted.
Have you measured the batteries? Are they really empty?

Oliver

[Ra1nman]

My Wittboy arrived and I will hopefully be setting it up today after firmware update, keeping fingers crossed all is good. Are there any current recommended calibrations for the haptic rain sensor as a starting point or just suck it and see? 

I will add some silicone grease to the rubber sealing rings before it goes outside, shall be sure to do some testing at ground level initially, make sure it looks to be working OK.

Wondering whether to leave the bird spikes off or not, any thoughts on whether they collect water, maybe leading to more chance of ingress, or is this not an issue. 

[ivano]

Hi!

Do you use the external power supply?
Then the reported battery voltage cannot be trusted.
Have you measured the batteries? Are they really empty?

Oliver

yes, the batteries are dead ,I don't use the external power supply [ You are not allowed to view attachments ]

[Arkadiusz_w]

strange problems today..
Yesterday the gw2000 lost the signal from the sensor, than it recovered again but without the data from the anemometer.
 [ You are not allowed to view attachments ]
I've re-register the sensor ID and immediately recover the data from anemometer. It's seems (the gw2000) it not decode correctly the wind after it wake up from signal loss....
beyond this, gusts at 150-160 km/h are expected next Tuesday 

I've had a lot of reception problems, and I've discovered they are caused by the gw2000 firmware version 2.1.9.
With version 2.2.0 they seem to have been fixed, but are still in testing... 
luckily i have remote access on the gw2000 , otherwise i would lose all the data...

the reception problems have been solved on the gw2000, i have no more data loss.

Big problems remain on the gw1100 version 2.2.0 with ws80 and wh40 sensors, i haven't tested with wittboy or other sensors. I've made a lot of test , and in the same location with same station, an old gw1000 works ok a gw1100 not 

There were many months of calm. My WS90 Wittboy started to lose frames in one location after 2 months. Resets it frequently to restore reception. I wrote to Ecowitt with this. I am using GW1100 gateway soft 2.2.2 and WS90 1.2.3.

[Rover1822]

for @ ivano

I would contact Ecowitt support directly via email. support@ecowitt.net

This appears to be a hardware issue (defect) . I had a similar but not the same issue. In your case it looks like the super capacitor is charging, but your batteries are draining. In my case the super capacitor was not charging.

I know from experience that the unit will run for 7-10 days on battery alone (as my cap was not charging) , the fact that yours can't even make a day of running , leads me to believe there is a short of some kind in the circuit.

You could try the hard reset of removing batteries , and placing the unit in the dark for 4 or 5, days, then putting the batteries back in. I doubt that will achieve anything though.

[Autofill]

The wittboy has some interesting power mechanisms. For example, since the cold weather came (or temps are constantly below 5oC), my supercapcitor voltage has been steady at 5.2/5.3 V (24/7) which suggests that its pulling all its power from DC. In contrast, in the summer or when temps are above 5oC, it seems to use supercap more as the voltage fluctuates often even though its plugged into DC power.

[Ra1nman]

Got the Wittboy running now, some observations as a newcomer!

Since the Wittboy updates every 8.8 seconds, why is there no option for anything less than 1 minute in the settings, is this to limit network traffic. When using the app on iOS 16 the landscape view doesn't appear to work, not a big deal, sure that will get resolved. Hoping Ecowitt will develop the app more, would be nice to get some larger tiles or maybe options for changing text size, maybe a more graphical option if it is being used as a pseudo console. Can't please everyone I know! 

I updated firmware on the WS90 and seems to be working OK at the moment, will road test it for a week and report back with any findings.