Forum Replies Created
-
Posts
-
As of this morning the weather numbers are flowing again. I touched nothing on my LAN or OSPi. It just started getting data again from Weather Underground.
Maybe this was caused externally and was/is outside my control.
Ibcoming weather data to feed Zimmerman is important since it not only saves money but gives the grass and plants what they need.
I hope someone can let me know what is causing this.
Thanks
Attachments:
Ray,
I am still having trouble with this. The watering level remains at 100% and the feed from Weather Underground is good but has no data. I have changed nothing on my LAN or on OSPi. This just appeared and no matter what I try, reboots of router and OSPi it won’t give me elements in the data feed. It’s driving me nuts. Do you have any idea what might be causing this please?
The problem seems to be a lack of content coming from Wunderground as this screenshot shows. This has been going on for a couple days now. The watering system defaults to 100%.
Attachments:
@ray,
Yes you did say it. earlier in this thread, I think. That was why I showed a picture of a v1.4 device. Since you shipped me a v1.4 it is the only version I have ever seen.
Oh, for sure it is better to integrate into opensprinkler. There is not much point in having an internet enabled irrigation system that requires you to stand in front og the thing with a screwdriver in your hand.
I only said that to illustrate that this thing seems to require hardware, software and integration. I have no intention whatsoever of offering a manual solution.
The difficulty is that like any sensor system there are two main parts, detecting a condition and then acting upon it. My belief is that we will end up connecting this to the I2C bust and doing the rest in software. Each sensor (either wired or wireless) needs to end up being connected something that can output a 0-255 variable. If this is passed via the I2C bus to a plugin that can start or stop any given channel (from watering) based upon a plugin set threshold from on the 0-255 variable.
The hardware is fairly simple using an Arduino chip since this has 6 analog inputs and 13 digital ones. Using the 6 analog and two of the digital ones would seem to be a reasonable compromise. Obviously one could use two Arduinos and provide 12 analog inputs but, that would put the cost up.
I am not sure how other sensors output their data so for now I will simply stick to a 0-3V input.
Does that sound a bit more sensible? I hope so…
@Dan,
I am happy to have a go at this. I will accept any and all inputs from anybody but it is my belief that Ray kinda owns the hardware design for this whole project. I PM’d him yesterday to ask if he had any requirements (I2C bus, USB, etc.) for connectivity and if he had any definite no-nos for hardware connection. When I have his answer, I’ll post it here.
Meanwhile my two sensors have turned up and they work well. You can give them any voltage between 3.3 – 20 and they give a 0-3 volt output depending on how wet they get.
It will be easy to make something read these and to give an output. Initially I will use an Arduino because I have several available and they have 6 analog inputs. This should give me an output to the I2C bus. If I get no further guidance, that is the way I shall go until a better idea comes along. The I2C bus works well because it is available on both the Arduino and the Pi. I2C is also on the Beagle Bone.
There is one very simple way to do this and it is to make a stand alone system that simply manipulates to outputs of opensprinkler. This would then not require any changes to any opensprinkle and would be connected in series with each output. When the desired soil moisture was reached, it would simply cut off that channel until the humidity dropped. This method has the advantage that there is no hardware or software needed. The moisture threshold would be set by potentiometers on each channel The down side is, of course that moisture sensor management would be entirely manual which is not that great.
You could be right for version 1.2. I only have version 1.4 and I used a pic from Ray’s site of an OSPi 1.4 device to show the only case where I knew the rain sensor input to be. For v1,4 I am certain as that is where I connected my rain sensor and, it is working. I believe version 1.4 to be the currently sold version so whoever buys an OSPi board today should get what is in my picture.
I will try hard to make something that will work with the existing hardware. That way existing users can retrofit moisture sensors to their systems. This may limit hardware choices. Like I said earlier, I am waiting for my sensors to be delivered. Only at that point can I seriously begin. Yes, I know it’s only 0-3Voults but it is nice to have the bits on the bench.
Ray needs to be involved from the very beginning. He “owns” the hardware design.
If anyone else wants to either pick up this little add-on project because they have either done it before or, are simply better at it than me, please say so as I am only helping out. There is no ego in this for me whatsoever.
I think the system software is where the main development will need to be as, from what I can see, there is no provision to stop a single zone from working based on outsode input and that will need to be changed. Again dialog with the developer from day one is needed. The mobile app will need to be modified too so that needs consideration as well.
Any help or design you can give is most welcome. More news when I have some bits on my bench.
Andrew, Great information. Thank you. I too had been think of using 1-wire. I am still waiting for my VH400s to be delivered so anything I say now is pure conjecture. My first priority when they arrive will be to simply make them work. After that, I will take any advice going.
My son is a member of the Charlotte NC Hackerspace and I told some of the members there that I was looking into this and I now have possibly more help than may be good for me. The software for this little project will be interesting since Once I make a bank of 8 of these things work, they will most likely provide a channel number (0-255) and a value (0-255) giving the moisture level. This will have to be made to turn any sprinkler zone on or off depending on a threshold set in software by the user. I do not think that OSPi currently has a per zone on/off capability.
My plan is to try and provide two types of sensor. Wired and wireless. I need wired so that VH400s can be buried under grass so that the grass may be mowed. Wireless won’t work underground and anyway you would need to find them and to dig them up to change batteries. Wireless sensors are needed for plants and beds where the VH400 can be simply poked into the ground leaving the battery and wireless parts above ground. Clearly the big advantage here is no cables need laying.
I originally liked using an Arduino because I know a little about it and it has six analog inputs so my plan was/is to serve up 6 x 0-255 channels and 2 x on/off. However if there is a better way then we can do that. However I also want to make this as cheap as possible. The VH400s are nearly $40 each when shipped so the 8 channel controller needs to be fairly low cost and, it also needs to be able to be daisy chained.
In the picture below, the rain sensor input can be seen. It is the green two wire connector on the right of the device. On the far left are two connectors for the common return wire and the 8 way connector in the center is the 8 zone output block. On the left side of the case you will see the orange power input connector and out of view on the rught side is an 8 pin expansion connector.
The rain sensor may be either normally open or normally closed and this setting is changeable via the software.
On the latest version v1.4 the opening panel in the center of the case is blocked off but is still visible. It severs no purpose on the OSPi.
For me the important part is that the system report what is currently installed. That way, months later I can still see what is on my system even if (highly likely) that I have forgotten.
Oh btw, the Monthly Adjust plugin is working on my system even though it still says I need to install “Note: In order for this plugin to actually work, The apscheduler Python module must be installed.”
I did this and waited for 1st August (+10%) to come around and, yay, it did, BUT it still says that I need to install it to make it work.
Thank you.
It would be useful if the rain delay did not affect manual operation. I know this has come up before quite recently but it really seems to stop any testing or demonstration once the rain delay has grabbed control. Rain delay works fine. It just seems to be in the wrong place in the system.
Rain delay also would benefit from working two ways.
1). the current way that permits a delay when no rain detector is present. This is as it works today.
2). to impose a delay at the end of a rain detector delay. By this I mean that the system has a rain detector but a further delay needs to happen after the rain detector has decided that it is no longer raining. Thus a rain detector that simply detects the current presence of water and turns off immediately the water has stopped can be catered for.Is that clear? Hope so… If not please come back and grumble at me to be more coherent.
Well done. From looking at your pic, it works. If you want to set up your WiFi and to use a fixed IP address (most useful ‘cuz that way it will always stay the same) try this.
You need to change two files on your system, /etc/network/interfaces and /etc/wpa_supplicant/wpa_supplicant.conf. Edit both files to match the samples from my working system. Obviously you will need to insert your IP address into interfaces and the name (ssid) of your wireless system and its encryption type (WPA, etc.) and your wireless key or password in wpa_supplicant.conf.
You can use a number of text editors to edit your files. There is vi, nano or pico. I use pico. To edit a file type sudo pico /path/filename.ext So to edit interfaces, type
sudo pico /etc/network/interfaces. You need to command sudo (Super User Do) to allow you to edit and save system files./etc/network/interfaces [/color
#File /etc/network/interfaces
# This file will setup your wireless network (wlan0) to have a fixed IP address
#
auto lo
iface lo inet loopback
# If you wish a fixd IP on your wired LAN (eth0) replace wlan0 with eth0 and vice versa in the line below
iface eth0 inet dhcp
allow-hotplug wlan0
auto wlan0
# Put your network (IP-address etc.) here
# If you wish a fixd IP on your wired LAN (eth0) replace wla0 with eth0 and vice versa in the line below
iface wlan0 inet static
address 192.168.1.50
netmask 255.255.255.0
gateway 192.168.1.1
# The file wpa_supplicant.conf defines your wireless network login parameters and will not be needed if you are using eth0
wpa-conf /etc/wpa_supplicant/wpa_supplicant.conf
iface default inet dhcp
#File: /etc/wpa_supplicant/wpa_supplicant.conf
ctrl_interface=DIR=/var/run/wpa_supplicant GROUP=netdev
update_config=1
network={
ssid="your_wireless_network_name"
scan_ssid=1
psk="your_wireless_network_key_or_password"
# key_mgnt is the type of wireless encryption mine is WPA_PSK
key_mgmt=WPA-PSK
id_str="home"
priority=15
}
I hope this helps.
Rain delay may not do what you think. I discovered this when I played with it. I thought it would give me a delay in watering after the rain sensor had activated thus giving the ground time to dry out. Not so. It seems to be a way of providing a delay in watering that is absolute. By this I mean that if you set it for 12 hours then it will not allow watering for the next 12 hours. It seems to have nothing to do with the rain sensor that I have on my system.
Maybe someone can tell me if this is correct.
Well, maybe not offer a list of available plugins but at least give a way to show that a particular plugin has actually err, plugged in when a neophyte user has installed it. My original was perhaps a little obscure in that the plugin was already installed and it told me that I needed to install another module before it would work. I installed the module (I think). The OSPi software gave me no indication as to whether:
– the needed module had already been installed before I started. This would have been sensible since the plugin needs the module to work.
– it was working (or not) when I installed it according to the published Linux instructions in the linked Wiki.
– I broke something by installing a later version of the module that the plugin could not use.You get my point?
Yes, an option to choose measurement units would be nice. Hardly anybody but the US uses gallons and inches anymore. Even US gallons are different from UK ones.
I was born in the UK and brought up with the “imperial system” as they then called it but also learned the metric system in science classes. Later I moved to Switzerland where knowing the metric system sure came in handy. After 35 years there I moved back to the UK for 4 years only to find that they too had switched to metric. Now I live permanently in the US, I am back to using feet, inches and miles again. Actually I quite like using 1/6 and 32nds of an inch. It is very satisfying. Even my old analog 0 – 1″ micrometer has been resurrected.
For me both systems are valid but the world seems to have decided that it is easier to count to ten rather than 3, 12, 20, 36 or whatever.
Using the Web page on the OSPi (http://ipaddress:8080/) login and then go to the Options page. Click the System tab and it will drop down to show several fields.
Enter your location much as I have done and the designated weather provider will be given this data. Then it should workOnce you have a microSD Card with the image on it as per my earlier post, you just stick the card in Ray’s adapter and then into the Pi and power it up. OSPi is designed to work headless i.e. without a screen but if you have your Pi connected to a screen and keyboard/mouse then you can login to the Pi (Username: pi, Password: raspberry) and play with the Pi.
To manage OSPi it will be better to use PuTTY http://www.chiark.greenend.org.uk/~sgtatham/putty/download.html to access the OSPI once it is installed. You will need to find the IP address of your OSPi and this can be done in several ways.
– If you have a screen and keyboard connected, just typeifconfigat the command line. This will display the IP addresses of the wired (eth0) and wireless (wlan0) interfaces.
– on an Android phone please download Fing from GooglePlay
– in Windows try Advanced IP Scanner here http://www.advanced-ip-scanner.com/Both the above utilities will reveal the current IP address of the Pi so that you can use PuTTY. You will also need the IP address when you install and configure salbahra’s OpenSprinkler program or to access OSPi’s internal web pages using :8080
Once you can access the Pi, and the web pages please come back here and we can help you along further. All this is a little bit daunting at first and if, like me you are not a Linux guru it seems even harder. However learning is about stumbling about an banging into things until you learn enough to ask the right questions. Even then the folk in here will guide you towards the right questions.
It is worth it because once it works, OSPi is brilliant.
@Dan,
I am retired so I have some time in between babysitting my two granddaughters.
As I said in my previous post I shall now go and buy some sensors and see what happens. I already have some 433Mhz stuff kicking around. Got a box or ATMega and AT Tiny chips too. I’ll see what I can do in the next few weeks as stuff gets delivered and I cobble it together.
UPDATE: I just ordered two Vegetronix VH400 sensors that you are using
@ray,
I like your idea of capacitance measurement using an ATTiny44. Neat.
I have no objections to using any components either off the shelf or home built. What I would aim for first is a solution that works. After that I’d look for cheap or home brew methods. However my main concern is making a device that can be used outside in all weathers. One variant would need to be lawn mower proof too and that would seem to mean burying it. Ain’t sure right now how to get 433Mhz through 6 inches of dirt. Cables are the obvious alternative. Even if I ignore the problems of in ground reliability of thin cables, there are two major worries with cabling. The first is making reliable connections under the ground. Obviously this can be solved with a decent case but that brings the cost up. Second is the chance of a nearby lightning strike inducing a power spike into the cabling spread out over a wide area – makes a great antenna. MOVs and optocoupler/isolators will minimize this but it will never eliminate it especially if the MOV gets damaged and not replaced.
Having done all the doom and gloom stuff above I will now go an buy a couple of sensors and have a look at how I think this may be done and then I’ll chuck it back to you in here and have you rip it apart. That way we might get a workable solution.
I guess that would put the Arduinos at the nodes (sensors). These would need to be in watertight containers and you certainly would need locally regulated power since you could not just send 5volts over a long distance down a bit of CAT5 (or whatever). An ATTiny 84 would be admirably suited to this task since it has sufficient I/O and is dirt cheap.
Can you design this? Do you need help? Just let me know. I have a number of Adruinos running various lights in my house.
I am now even more worried about lightning strikes, etc. with buried cables, arduinos, power regulation and the sensors all in a watertight enclosure, some buried and some not buried. This is an interesting challenge indeed.
The direct download link for the pre-configured SD card image is here http://rayshobby.net/?p=7606 as are the links to the instructions as to how to get the image onto a microSDcard.
If you get stuck, come back here.
Dan, I am very far from being a programmer. I used to use Dartmouth BASIC (line numbers) a very long time ago. I am more a hardware guy (even that I learned on tubes…). So from a user standpoint it would be nice to have a verifiable installation that is as easy as possible to accomplish. Software installs, upgrades and removals are never easy and they often affect other stuff that the user never thought about.
Firstly is the question of upgrading OSPi itself. Could that be made easier? Could you put a button in the options screen that said. “Check for Updates?” Even better would be one that said “Update available, do you want to install it?” See what I mean?
With regard to Plugins, there were some installed when I opened up 2.04. One was a thing called “Monthly Irrigatin Adjustments” (Copy/Paste from the screen 😀 ). when I ran it, it said “Note: In order for this plugin to actually work, The apscheduler Python module must be installed.
Visit the Wiki for instructions.”. I followed the Wiki and it did the things it said and I got no errors. I also got no changes displayed on-screen when in the Spinkler webpage. Did something install correctly or, did it fail? How would I know as a user?I believe that it would be nice if somehow a user could see just what has been installed and, if it is going to work. I have no idea if this is possible but, I sure would like to know that my system is working as I intend it.
Please understand that I am not complaining. I think that OSPi is about the best piece of hardware and software I have bought in a long time. I have had lots of fun building it making it work. It has amazed my neighbors and gotten me at least one free beer so, I am a very happy bunny indeed. I guess it really depends on where you see this product going.
Thanks for a great product and your support.
Thinking about this, it would seem to be a good idea to opto-couple/isolate the sensors before they reach the GPIO pin(s). I could not see any way that the sensors could be connected to a bus so each sensor would seem to need its own 4 wire cable as opposed to a connect to one, then the next and the next type of install. That is a pity as it means a lot more cabling.
Using wireless is OK for plants and beds but is useless on grass with a lawn mower.
Since these sensors provide a variable analog output voltage it would seem possible to be able set a “moisture watering point” for each sensor. This might require some extra hardware to accomplish.
The combination of moisture sensing, water flow control (one flow control, multiple zone measurement) and weather forecasting is a near unbeatable combination for efficient watering.
