Forum Replies Created
-
Posts
-
By “12 sections” did you mean 12 zones? The main controller can only control 8 zones. If you want to go beyond 8 zones, that’s what the expanders are for: each expander uses one additional PCA9555 to add 16 zones.
The reason the main controller only does 8 zones is that among the 16 IO pins of PCA9555, 8 are used for zone control, the other 8 are reserved for other purposes: including button 3, boost converter control on DC-powered OpenSprinkler etc.
In our support website, we’ve added a Troubleshooting and Technical Help page:
https://openthings.freshdesk.com/support/solutions/articles/5000859137
your question is under the paragraph of “I got a blank page when accessing the controller’s homepage”. Please check the suggestions there.Right, it looks like this has to do with Catalina dropping support for 32-bit programs. Yes we will update the instructions shortly. Thanks for reporting.
I have no idea: this doesn’t have to do with the firmware, it has to do with your operating system. I just googled “Bad CPU type in executable” and found mentions of this error in Catalina. You can google it yourself and check solutions.
If you follow OS 3.0 schematic, please note that there are two PCF8574 IO expanders, one on the top level (master controller), one on the bottom (AC-driver). The top level is set to I2C address 000, and bottom is 001. The bottom level controls solenoid drivers.
I am not sure exactly which schematic you are following. As I said in my first reply: only version 3.0 board design uses PCF8574 — in fact, it uses one on the top level board, and one on the driver board. Revisions 3.1 and 3.2 do NOT use PCF8574, instead, they use a single PCA9555 IO expander chip. The reason is that PCA9555 is a 16-channel IO expander chip, while PCF8574 is 8-channel. So the original 3.0 design had to use 2x PCF8574, and later with PCA9555 it only needs one single IO expander chip. Also, PCA9555 uses totem output so can source much higher current into the traics on the driver board. PCA8574, on the other hand, uses weak pullup so cannot source high current (and for that reason OS 3.0 driver design had to use additional PNP transistors to provide high current).
If you present the same data to 5 different gardners, chances are they will tell you 5 different things and they all think they are right. The weather algorithm is just an algorithm, it’s not human intelligence, it can’t read your mind. That’s why you are given parameters which you can tweak to fit your own specific need. You can’t expect it to work perfectly with the default settings. I think your expectation is too high.
“150% in the afternoon when no sane person would water” — I don’t think everyone agrees on that. If you don’t want it to water in the afternoon, then just not schedule a program in the afternoon. As I said, the algorithm can’t read your mind, but there is enough flexibility in the firmware that you can configure it to better fit your need.
That said, if you observe the watering percentage change during a day (sounds like you do, since you mentioned 0% at night and 150% in the afternoon) — then perhaps you are hosting your own weather server? If you use the official OpenSprinkler weather server, it uses caching so the watering percentage during a day does not vary.
About ETo: yes, of course you should tweak the ETo. That’s mentioned in the support article that explains the weather algorithm:
https://openthings.freshdesk.com/support/solutions/articles/5000823370-using-weather-adjustmentsAt the moment the beta test is for OpenGarage only. It’s because OpenGarage firmware is much simpler than OpenSprinkler, so it’s a great platform for testing and debugging. Once we verify that OTC works reliably and iron out the technical issues, we will integrate it to OpenSprinkler as well.
The OpenSprinkler weather server uses a Darksky API key that we have paid for. Users do not need to sign up for a key or pay for anything, except if you plan to run the weather service on your own server.
legacy OS hardware design files are in the rayshobby github:
https://github.com/rayshobby/opensprinkler/tree/master/OpenSprinkler%20Controller/hardware
you can install EagleCAD and open the .sch/.brd filesThat’s partly the reason we haven’t officially released the firmware — we need time to prepare the documentation. Right now it’s only clear from reading the code (for example, the code clearly publishes to topic named ‘opensprinkler’), but we can’t expect everyone to be programmers so we need to write down instructions for regular users.
Which firmware does your controller run?
It’s certainly not by design that sometimes it’s missing the stations tree. You can do a json parsing on the result and if it’s corrupted ignore that call result.
While the PR has been merged, we have not officially released the firmware or the corresponding App yet, because we are making a few minor improvements like adding mqtt user name / password support. Also we need to do final testing on all platforms (OS 2.3, OS 3.0, OSPi) before officially announcing it and releasing the compiled firmware files. Should be done in the next few days.
This version will be numbered 2.1.9(4), not 2.2.0, because it doesn’t involve major changes to configuration files or flash layout.
There are separate commands that you can use, such as /jo for just obtaining the option values, /jn for just station data. These are all documented in the OpenSprinkler API document. /ja is a collector of all the data combined.
Yes, that’s why the DC-powered OpenSprinkler is shipped with a 7.5VDC 2A power adapter (it’s included by default for US/Cadana orders, but not included for other countries). It provides a holding current that’s the closest to a typical sprinkler solenoid.
I think 12VDC is still pretty safe: I’ve seen discussions online where users have been using 12VDC to drive their sprinkler solenoids for years without any problem.
We had a batch of DC-powered controller that used 2.5mm plug (more specifically, the plug is 5.5mm OD x 2.5mm ID). The current batch should all be 2.1mm, which is more common. The power adapters that we ship can accommodate both 2.5mm and 2.1mm.
You can use a pump start relay, which is pretty standard for controlling a pump in sprinkler system. You can search on Amazon and find plenty of options, such as this:
https://www.amazon.com/bomba-aspersor-%C3%B3rbita-arranque-sistema/dp/B000I19I5E
it’s usually rated by the horse power that it can handle. The horse power should be sufficient to match your pump.Program 1 just means the first program in your list of programs. You can name it anyway you want, but it’s whatever appears the first in the program list. Program 2 is the second.
At the homepage, click the lower-right corner icon to bring up the menu, then ‘Show Disabled’. This is also available in Edit Options -> Station Handling.
While the firmware doesn’t limit the number of valves that you can turn on at the same time, I would say that simultaneous zone runs should be limited to 4 or less. Running a lot of zones at the same time can significantly increase the current load of the system. If you do want to run many zones at the same time, adding an external relay board would be the best since you can then use very thick wires to allow much more current to flow through it.
Honestly, I have no idea. It looks like you were trying to plugging and unplugging wires when the controller was powered on. I highly recommend against it — any wiring or rewiring should be done when the controller is powered off, because you never know when you may insert wires into the wrong port.
As I said, there is one thing I am puzzled about: you measured 28VAC between COM and zone 2, and you have the solenoid wires plugged into COM and zone 2 when you did the measurement, and you said it doesn’t activate the solenoid. This I really cannot explain — whether triacs inside are burned or not, or whether the controller is working or not, 28VAC was clearly present between COM and zone 2 when you measured it, why does it not activate the solenoid? I cannot explain. This is like saying: I measured 28VAC across a resistor, but there is no current detected in resistor. How is this possible? I don’t know. This defies physics.
You last photo, which shows the solenoid writing: what type of plugs are you using at the end of the gray cable? The image resolution is pretty low so I can’t tell. It seems I can only see two solenoid wires that go to the plug, I can’t see where the other solenoid wires went. Did you try to connect the two wire from a solenoid directly to the 24VAC transformer wires, and see if that activates the solenoid? As I said, if you measure 28VAC between COM and zone 2, and you have a solenoid connected between COM and zone 2, it doesn’t make sense to me that the solenoid doesn’t activate. That almost certainly suggests there is a wiring problem.
I really cannot explain why it doesn’t work. Two things that are strange:
1. You measured about 28VAC on slot 2 with the zone wire connected, so that means at the time of your measurement, 28VAC is present between COM and slot 2. I don’t understand how that can’t activate your solenoid — it’s like saying 28VAC is applied on the two wires of the solenoid, and it doesn’t open. This does not make sense.
2. The screen shows the current draw is 3532mA — this is also very wrong, the current draw is usually about 200 to 300mA when a solenoid is active. 3532mA is way beyond the limit.Something is apparently quite wrong but I don’t know what.
