bpwwer

A quick search found these https://www.amazon.com/Wal-front-Vehicle-Detector-Inductive/dp/B07K1BNL59/ref=psdc_7459508011_t1_B07H3N41H4 Not much info, but for ~$20 it might be worth playing with.

The node server does try to minimize the data sent to the ISY (and thus minimize the log entries) by only sending data when it has changed. But some weather data values do change frequently and the only way to change how often is to change the poll interval. This is one of the reasons I try to be somewhat selective in the weather data presented by the weather node servers. The more different values the node server tracks, the more data that has to be sent to the ISY.

Setting the log level to error should reduce the messages down to just error messages. Re-Discover causes the node server to review the nodes it's created. It will do this every time it start automatically and if you change the Forecast Days parameter as that's the only way to change the nodes it creates. Remove Notices will clear any notices the node server has displayed on the Polyglot dashboard-details screen. Update Profile will cause the node server to resend the profile files to the ISY. This also happens automatically when the node server is installed, but in some cases, the ISY is unable to accept the files (maybe it was too busy) so the option is there to resend them manually.

I just wanted to point out that it's not a requirement to have to Polisy's or RPi's to support two ISY's. The current ISY ecosystem wasn't really designed with the idea that there'd be multiple ISY's trying to co-exist with overlapping data sources. Some type of creative solution will likely be needed. For the Davis, that may mean feeding the data to something else first, that can support sending the data to multiple consumers.

You can run multiple instances of Polyglot on the same hardware. It does have environment variables that allow you to change the ports used by an instance of the program. I'm currently running both version 2 and an alpha of version 3 on my Polisy. If you wanted to install a second instance of Polyglot running on a Polisy, you'd have to set up the directory structure and port configuration so they don't conflict with each other. You can also run multiple copies of Polyglot on an RPi, though I've never tried that.

No not OpenHab, I just searched on github and npm for projects related to velux. The ones I found provide simple examples to create a program that uses the module to communicate with the KLF200. The first one I found on github looked like you could load it on an rPi via PIP and create a simple python program that used it. That's where I'd start just to get an idea on how to use the module. Maybe look at more than one and pick the one that works best. At this stage, not even trying to do any integration with ISY. We're still working to define the PG3 API's, but there's no expectation that existing PG2 node server will run as-is on PG3. How much effort will be needed to migrate a node server is still unknown.

it looks like there are both Python and node.js modules available that implement the API. I'd suggest looking over those and maybe experimenting with them first to understand better how to interact with the KLF200. Then you have to decide what information is important to both for display and commands. Use that to define what you want the nodes to look like on the admin console. From there you can create the profile files and start creating a node server from the template available. Based on what little I've looked at, the KLF200 is a controller that supports a number of different devices, that means creating nodes with status/commands for all the different supported devices so it sounds like a fairly complex node server. Another consideration at this point is that PG3 has just started alpha testing. It's not ready yet to start developing node servers for it, but it will be soon. It may not make sense to put in a lot of effort creating a new PG2 based node server and then more effort to convert it to PG3.

Someone else may be able to provide a better solution, but I believe you need to set a flag (variable) and use that to trigger the notification. Something like: if AQI > 100 then high_aqi = 1 else high_aqi = 0 and a second program if high_aqi = 1 then send notification The problem you're having is that every time the AQI changes, it triggers the if condition to re-evaluate. By using a variable, the variable only changes when the AQI goes over 100 or under 100.

If you weren't aware, the admin account used to ssh in and the admin account used for the web interface are different accounts. That's more of an FYI for anyone following this. Have you reloaded the web interface browser page? My system will get stuck where I can't log in, but as soon as I reload the page it lets me log in. Something seems to be cached on the browser that prevents it from properly authenticating. I'm now in the habit of always re-loading.

All of the valid values are in the documentation here: https://github.com/bpaauwe/udi-owm-poly/blob/master/README.md Current condition node sys.node.[address].ST (Node sever online) sys.node.[address].CLITEMP (current temperature) sys.node.[address].CLIHUM (current humidity) sys.node.[address].BARPRES (current barometric pressure) sys.node.[address].WINDDIR (current wind direction ) sys.node.[address].DISTANC (current visibility) sys.node.[address].DEWPT (current dew point temperature) sys.node.[address].UV (current UV index) sys.node.[address].GV2 (current feels like temperature) sys.node.[address].GV4 (current wind speed) sys.node.[address].GV6 (current rain rate) sys.node.[address].GV7 (current snow rate) sys.node.[address].GV13 (current conditions) sys.node.[address].GV14 (current percent cloud coverage) sys.node.[address].GV18 (current percent chance of precipitation) Forecast node sys.node.[address].CLIHUM (forecasted humidity) sys.node.[address].BARPRES (forecasted barometric pressure) sys.node.[address].DEWPT (forecasted dew point temperature) sys.node.[address].UV (forecasted max UV index) sys.node.[address].GV19 (day of week forecast is for) sys.node.[address].GV0 (forecasted high temperature) sys.node.[address].GV1 (forecasted low temperature) sys.node.[address].GV2 (forecasted daytime feels like temperature) sys.node.[address].GV13 (forecasted conditions) sys.node.[address].GV14 (forecasted percent cloud coverage) sys.node.[address].GV4 (forecasted wind speed) sys.node.[address].GV6 (forecasted rain) sys.node.[address].GV7 (forecasted snow) sys.node.[address].GV18 (forecasted percent chance of precipitation) sys.node.[address].GV20 (calculated ETo for the day) Your 'address' is n001_weather for current conditions so just substitute n001_weather for [address] above. For the forecast nodes, it's probably n001_forecast_0, n001_forecast_1, etc.

I don't have any real plans, just experimenting. Seems like this could be the basis of a Insteon node server. Currently, I'm able to see messages from my mini remotes and that's about it. I don't think I have too many dual band devices as a lot of my Insteon network is old.

@evilpeteThanks! This is pretty cool! I just installed librtlsdr and rtl_433 on my Polisy and it's working. I had been using rtl_433 to pull data from an Acurite weather station so I had the rtl_sdr dongle already. What should I try to do with it now? I might have to dust off some of my old Insteon message decoding code. There is a freebsd package for rtl-sdr that includes librtlsdr so maybe that can be added to the Polisy's package repository? @Michel Kohanim?

@PeterBarker11I'm assuming based on your post in the other thread that this is now working for you? In case anyone else sees this while looking for the solution. I believe this was resolved by restarting the admin console. When a new node server is installed, the admin console must be restarted before it can properly display the new nodes created by the node server.

My node server does pull data directly from the device, I'm using the V1 API and it can be configured to pull as often as you want. It currently does not handle multiple sensors, at least not correctly, so I'd have to make some modifications to handle additional outside temp/humidity sensors and I'd need to see what the data looks like to do that. From looking at the docs, it appears that the WLL device does support multiple sensors for the current conditions, but I'm not sure I'd bet my own $200 on that.

The node server doesn't support any extra sensors. It pulls data from the 'current' 'indoor', 'soil' sections of the data. It only pulls one temperature/humidity value from the 'current' section. I can add more values, but would need to see the actual data from WLL with the additional data as I don't have a Davis system.

Great post! I have something like this also. I look at window status via a DSC alarm and will turn off the AC if a window is open. I also won't turn on the fan if all windows are closed. I use a personal weather station to get the outside temperature and look at the diff between the inside and outside temp. When the outside is a few degrees below the inside it triggers a program that can turn the fan on if a window is open. It will also turn the fan off if the outside is close to or above the inside. When this was all working, I could open a window when the outside temperature was about equal to the inside and as soon as it cooled down outside enough, the fan would automatically come on. When it started warming up in the morning such that the outside temp started to reach the inside temp, it would shut the fan off. There have been times that it got so cold inside that the heat would come on when the windows were closed so I like your idea of shutting off the fan if it gets too cool inside. I just created a Purple Air node server that monitors air quality so I'd like to incorporate that too so that we don't run the fan if it's smokey outside. We've woken up at night to find the house smelling like smoke inside.

I had a whole house fan set up with a bunch of rules including indoor/outdoor temperature. At one point I got it stuck in a loop where it would turn on, blow some air over the HVAC thermostat which would cause an indoor temperature change which would turn off the fan. Once it stopped the temperature changed again and the fan would turn back on. Probably not good cycling the fan like that. It took some playing around with temperatures to get it working right. It was pretty cool when it did. I had it also tied to window state so when the indoor and outdoor temp was about the same I could open some windows and then when there was enough of a difference the fan would come on and start cooling things down. And if it ever got too cold inside, it would shut the fan off, or in the morning when it started getting too warm outside, it would shut the fan off so that we wouldn't pull warm air into the house. Basically all we had to do to control the fan was open/close windows.

All of the weather conditions are considered status. Control would be used for things that don't really have a current status. Something like a lightning strike event could be a control, but the distance to the strike is a status. But I haven't created any controls in the WeatherFlow node server. Any program that triggers on a status, will get executed when the status value changes. So yes, if you had something like: if temperature > 76 then turn on fan else turn off fan would cause the fan to turn on/off every time the changed going above/below that limit. To prevent that you either need to guard band the value I.E. have the on temperature > than the off temperature or add time references so that it has to be above or below the threshold for some amount of time before switching. The node server is getting updates from the WeatherFlow hub at 1 minute intervals so that does limit somewhat how often things can change.

File -> Set Userid/Password from the admin console.

That did it, both are showing up now, thanks!

After uninstalling and re-installing I'm still getting Model SE7600A-USS20NHB2 is not yet supported Interestingly that's the older of the two models I have.

Yes, I've played around with the charts on the portal. I was really looking for real-time (at least every minute) info on battery charge/discharge rates that I can feed automatically to my database. While the charts are interesting, they're not real-time. I think the data only updates every 15 minutes and you have to regenerate the chart to see any updated data. I believe the API provides data updates every 5 minutes. It seems like the data should be available over the modbus, and that's probably how the battery communicates it's status to the inverter, but I haven't found any documents that indicate it's possible to access that data over the local network port.

Yup, I suspect the portal and the node server are using the same API calls to get the information. While I have the node server running, I'm not currently using it for anything. I am collecting data locally from the inverters and meter via modtcp and via an external meter. I dump this into a database and generate my own charts. The one things I'm missing is info on the battery. When the battery is charging, it looks like I'm not producing as much, but when the battery is discharging I see it as if I have PV production happening. So it works out in the end, but I was looking into the API to see if I could get some additional data on what's happening with the battery. There is access to a lot of different data vial the API but my understanding is that it's not real-time,

That's basically the system I have too. Two inverters with a battery connected to one of them. What I'm getting with this node server is a Site node which has an aggregate of the two inverters that has a single child node for the battery. However, I installed the node server after everything was installed. As a last resort you could delete the node server and re-install it.

I just pushed a new version that should fix this. Thanks for the report.