Forum Replies Created
-
Posts
-
What does this mean:
Ű
Ű
is this what your browser returns, or what’s displayed on the LCD screen, or something else? Can you go to OpenSprinkler, click a button, like B1, and see if the LCD responds with the IP address?@kms: that’s a good point, I never thought about it. Will add the sync() as you suggested. Thanks for reporting.
To be sure: you need to unzip the whole package to a local folder, like your Desktop. DO NOT double click on the executable while it’s still in the zip file — Windows makes it look like it’s a folder but it’s still in the zip file. You need to explicitly ‘Extract’ or ‘Unzip’ the zip file first.
On the webpage do you see a switch named ‘Manual’. Just click on it to turn the controller back to program mode.
If you can attach a screenshot it would be easier to explain.
Make sure that you don’t press the button B2 for too long after inserting the USB cable. Release it within a second or two after inserting the USB cable. The LCD screen should remain off — if not, that means the device is not in bootloading mode.
Sorry about the confusion. Because very few users take advantage of the relay, and it’s kind of expensive, we’ve made the decision to remove the relay from all OpenSprinkler hardware (including OS, OSPi, and OSBo).
Also, due to a recent supply chain problem of ATmega644, it turns out that buying ATmega644 ends up costing more than ATmega1284, that’s why we decided to upgrade to ATmega1284 ahead of schedule. I will try to check in the schematic this weekend.
You are right that you can still add your own relay to use the feature. Unfortunately I realized that when removing the relay, I didn’t add the corresponding digital pin to the pin out area. So that microcontroller pin is not mapped out. At this point, your best option is to modify the pin definition (in defines.h) to use any other available pin, and recompile the source code.
The original reason it’s constrained to stations 1-8 is that I thought stations 1-8 always exist, so when you define a master station it’s always valid. However, now I think about it, there is no reason why it can’t be just any station. It’s pretty easy to modify the code to support that.
If I understand it correctly, what you are saying is that after running manual program, the controller does not return back to program mode, is that right? I am not sure which firmware version you are running, but it sounds like your version requires turning off manual mode (i.e. turning it back to program mode).
It looks like both requests are about having more than 1 master valve. The firmware currently only supports one master station. However, you can make use of the ‘parallel station’ feature. Specifically, you can define a station as a ‘parallel’ (i.e. non-sequential) station. When creating a program, include that station as well, and set its time to span the whole program. This way, that station will turn on when the program is on.
This is assuming you are running the Unified OpenSprinkler Firmware, which supports parallel stations.
Sure, I’d be happy to receive a flow sensor for testing. Please contact me directly at [email protected] Thanks.
Yes, I would be happy to integrate your code to future firmwares. You can attach the code here, or contact me directly at [email protected]. Thanks.
In Linux, you will need to run the Updater program in ‘sudo’ (or add the USB VID/PID pair to /etc/udev/rules.d/). This is a USB-permission thing in Linux.
You should always use the MAC address reported on the LCD display of OpenSprinkler. The WiFi adapter (assume you’ve set it in client mode) is a simply pass-through, so don’t use the WiFi adapter’s MAC.
For additional details, we have a tutorial video on how to set up fixed IP, and port forwarding:
https://opensprinkler.freshdesk.com/solution/categories/5000022938/folders/5000099521/articles/5000569763-video-how-to-set-up-port-forwarding-for-remote-accessTo second what Samer said: the rain sensor typically has a mechanical delay (which is also adjustable by opening or closing the vent size), so you don’t need to add additional delay in software. I have written a blog post about rain sensor. Check it out here:
http://rayshobby.net/using-rainfreeze-sensor-with-opensprinkler/@aruzsi: for OpenSprinkler 2.1, the instructions are:
1) Unplug 24VAC, unplug USB (if it’s already connected)
2) Plug one end of the USB cable to your computer
3) Press and hold button B2, plug the other end of USB to OpenSprinkler
4) Release B2 two seconds after USB is plugged inThe LCD screen should remain off. If it lights up, that means you have to hold B2 a bit longer. The device should show up as a USBasp device, and if you are in Windows, you will be promoted to install driver if driver is not already installed.
For additional details, please check the firmware update tutorial:
https://opensprinkler.freshdesk.com/solution/categories/5000022938/folders/5000099521/articles/5000381694-update-opensprinkler-firmware-with-downloads-Here is my suggestion: when it loses response, check the LCD screen and see if there is an arrow icon displayed at the right end on the second line. If you see that arrow, it means the controller is trying to ping the router but cannot reach the router.
If there is no arrow, try to click button B1 or B2. If the LCD does not respond to button clicks, it’s likely the controller is busy waiting for NTP sync to return, or waiting for weather / time zone query to return. These are likely due to issues with the Internet connection (or in case you have installed a firewall, perhaps the firewall is blocking incoming packets to OpenSprinkler for some reason).
If the LCD does respond to button clicks, then it’s really unclear to me why it has intermittent connection issues. We may have to ask you to send it back to us for further checking.
When you try to upload the firmware, is OpenSprinkler in bootloading mode? As I said earlier in the thread, if you already know you have OpenSprinkler 2.1, there is no need to click on ‘Auto Detect’. In any case, before uploading firmware, make sure OpenSprinkler is in bootloading mode.
Once it’s in bootloading mode, Windows should detect it as a device under libusb-win32 section, and the name of the device is USBasp.
Also, make sure that when you enter bootloading mode, do NOT press the button for too long. If the button is not release within 1-2 second, Windows thinks that the device has lost response.
@Rob, the blog post here explains the story behind the OSPi enclosure:
http://rayshobby.net/work-in-progress-opensprinkler-pi-ospi-a-will-finally-close-the-gap/
Long story short, we use the same enclosure for OS, OSPi and OSBo. In fact, figuring out that I can fit OSPi inside the OS enclosure is pretty much the primary reason I started designing OSPi. As you may or may not know, making a new injection mold is very expensive — including the design fee, it’s easily reaching 10K. At the volume we are selling, it doesn’t justify making a new enclosure for OSPi. I was actually glad that I didn’t go down that path initially — because RPi has so many different versions, it would be difficult to design one enclosure that fits all of them. For example, if I had paid 10K to make an enclosure for the old model A/B, it would be useless for the current A+/B+/RPi 2. So while $77 is a not cheap, as you can imagine, if we had designed a new enclosure, it would be quite a bit more expensive.Since people don’t look at a sprinkler controller on a day to day basis (unlike a thermostat), I figured that aesthetics is perhaps not that important. Especially since it’s a web-based sprinkler controller, hopefully you just need to access it remotely and don’t have to access it physically.
While $77 is not cheap, as a business we have to make sufficient profit to keep going. Also, I think you would agree there is no other competing product on the market that’s close to this price range. Most web-based sprinkler controllers are well above $200, closed-source, and often don’t provide as much flexibility / programming features. Again, we are not mass producing OSPi, so it’s not easy to get the cost down. I also don’t think it will become a high volume product — it’s mainly designed for makers, tinkers, RPi enthusiasts, and it’s not oriented to the mass market. Sorry about being wordy, but I am just explaining the rationale behind the price tag.
The microcontroller-based OpenSprinkler is the one that’s targeted to the mass market — it’s all fully assembled, with LCD and buttons, and the enclosure is specially designed for it. That being said, as described in the post above, I am working on a new version of OSPi specially tailored to RPi A+, which is likely the first OSPi that has LCD and buttons and fits perfectly inside the OS enclosure.
Ah, sorry, just came to this thread after replying to you in a different thread. Will definitely consider adding the support for flow measurement in future firmware code. If you want, you can post your suggestions/proposals here, or email me directly at [email protected] to discuss how to proceed further.
Well, I guess you are right — sometimes you’ve just got to trust the algorithm 🙂 To be fair, the algorithm is a pretty simple, heuristics-based one. For example — it uses the average temperature, which in your case is pretty low. But your highest temperature during the day may be quite high, and you would be wondering why it doesn’t decide to water during the day. Well, all I can say is it’s after all not a real gardner, so it can’t make very intelligent decisions. But hopefully the ups and downs will be averaged out, so that it does the right thing over time.
@Pierre, glad to hear it worked. Well, 4.5V is not unusual, but I thought 2.8V is a bit on the low side. Because that voltage comes from the 3.3V linear regulator, unless if the circuit is drawing a lot of current, it should be pretty close to 3.3V. If you want to test further, I suggest you unplug 24VAC, but instead power the circuit with a USB cable, and retest the VCC-GND voltage. See if it’s close to 3.3V. It’s just my precaution — I am pretty sensitive to these numbers, and I feel 2.8V is a bit unusual.
@Robert, glad to see that you are making use of the current system to log flow rate. Here is how the log data is formatted: each entry records the [program_index, station_index, duration, time_stamp]. When program_index is non-zero, this is a normal log record storing a standard sprinkler event. When program_index is 0, this is a special log record, and the station_index becomes the name of the event. The currently supported events are defined here:
https://github.com/OpenSprinkler/OpenSprinklerGen2/blob/master/main.cpp#L736
there are rain sensor, rain delay, and water level events. In the future we can add an additional event such as “fl” (flow), and duration would become the numerical value of the flow measurement.A blank ATmega644 does not come with bootloader, you will need to use an ISP programmer to program a bootloader to it. On the other hand, if you already have an ISP programmer, you can directly program the firmware to the microcontroller without going through OpenSprinkler’s USB port.
Sorry about the trouble you encountered. I still haven’t received my RPi 2 so I can’t test. But I am pretty sure the OSPi image (which was made last year) is NOT compatible with RPi 2. We will update the image soon to use the most recent Raspbian. Do keep in mind that you can install Dan’s Python program, or our new Unified OpenSprinkler Firmware directly, without using the pre-configured image. For both of them, the source code is released on Github, and you can follow their specific instructions to clone, install, and run the code.
Given that your temperature is quite low these days, the watering percentage is probably coming out to be 0 anyways. For testing, I suggest 1) change the weather adjustment to use Zimmerman method; 2) change your location to Sydney,Australia. See if this will result in a non-zero watering percentage.
