Forum Replies Created
-
Posts
-
There is no contradiction between the two documents:
– the schematic shows the index of the pin on the corresponding pin headers. Due to the layout of the mapped out pins, the entire P9 is split into three sections, therefore the pin number restarts from 1 on the SYS_5V pins.
– the online manual shows the actual logic name of the pin, and is the only thing that matters to you when writing software.The 33Kohm resistor is used to protect the rain sensor GPIO in case someone accidentally connected 24V AC to the rain sensor port (believe it or not, this happened in the past!). 33K, 30K, 27K all work fine, depending on what resistors we have in stock.
Sorry I misread your post. When you did the voltage test, did you power the circuit from 24V AC, or did you use a USB cable to power the RPi directly? If you’ve powered it with 24V AC, that means 24V AC is present on the board, it must be that a PCB trace from the terminal block to the COM wire is broken or something.
No, it doesn’t pulse at the zero crossings, it simply holds the gate high when you need the solenoid activated. It’s similar to how you would operate a transistor switch.
Technically the pulsing is more common and more power efficient too, but from the circuit design point of view, it’s simpler to just hold the gate high.
A jumper wire across the rain sensor terminals indeed simulates a closed sensor. However, the type ‘normally closed’, or ‘normally open’ is a software thing defined in the options. For example, if you choose the rain sensor type as ‘normally closed’, then the controller assumes:
– there is no rain when the two pins are shorted (i.e. jumper wire is inserted)
– rain is detected when the two pins are open (i.e. jumper wire is removed)If you choose the rain sensor type as ‘normally open’, then the logic is reversed.
Sigh, I have a feeling that letting users see the output messages complicates things… They are just the output from running the avrdude command. If you want to take a close-up look at the messages, you can bypass the GUI and run avrdude in command line directly. For exmaple:
avrdude -c usbasp -p m644p -U flash:w:firmware2.0.6.hex
is what it’s really running.Sounds like the connector on your AC in terminal is not reliably. Perhaps use a screw-driver to tighten the screws on the terminal block and try again. Also make sure you’ve inserted the plug all the way in.
I think it’s producing just a little too much A/C power.
Not sure what this means. I did hear many questions about whether the transformer will ‘pump in’ too much current to the circuit — so let me clarify this common confusion: the transformer does not ‘pump in’ current to a circuit, the circuit draws current from the transformer, and it can draw anywhere from nothing to the maximum rated current of the transformer. For example, a 24V AC / 750mA transformer is capable of delivering a maximum of 750mA current and maintain roughly 24V AC output voltage. But how much current it’s actually delivering to the circuit depends on how much the circuit draws.
I can’t measure any DC voltage across the LED or anywhere else for that matter,
There is a +5V pin in the RPi pinout area located at the upper-left corner of the circuit board. If you look at the user manual page 9:
http://rayshobby.net/docs/ospi14_manual.pdf
the third pin on that row is a ground pin. So you can put your multimeter probes across the first (+5V) and third (GND) pin to measure voltage. There are also some ground pins in the ADC pin out area close to the right edge of the circuit board.You should first make sure the LED lights up reliably between you plug in RPi.
There are two types of the mini relays: the earlier version used a black 110V AC / 2A relay, and the current version uses a yellow 250V AC / 3A relay. Both would work perfectly fine for driving a sprinkler solenoid, which only needs 24V AC / a few hundred milli-amps. I don’t see any big issue as the voltage / current both fall well below the relay ratings. The back EMF could be an issue, but the relay pins are relatively far away from the rest of the circuit, so it’s probably not a big concern in practice.
Dan’s interval program now provides a plug-in that shows how to control the relay. The OSPi user manual now also includes information about the pin that the relay uses:
http://rayshobby.net/docs/ospi14_manual.pdf
Basically outputting a high on that pin activates the relay, and low deactivates the relay.why there is a limit of 2GB on microSD card and FAT16? Standard Arduino SD library seems to handle well big cards (I’m using 32GB card) and FAT32 with no problems. Is it a hardware or software restriction?
Update: the statement below is outdated. See later discussions in this thread.
This is a limitation imposed by the tinyFAT library:
http://www.henningkarlsen.com/electronics/library.php?id=37
tinyFAT is a very compact SD card library, produces code that’s considerably smaller and consumes less RAM than the standard Arduino SD library. I’ve tried the Arduino SD library (you may have noticed that the earlier versions of the source code had a macro define to switch between tinyFAT and Arduino SD), but it consumes too much resources that I fear would eat into the other planned features.I’ve never used the HummingBoard, so I can’t tell for sure. As long as it’s pin compatible with RPi, I would say it should work.
Why pry open the enclosure? There are four big screws at the back of the enclosure. Just unscrew them. Honestly I don’t think this needs to be mentioned in the user manual…
@iDoug: if the webpage says firmware 2.0.6, it means you’ve upgraded it successfully. Regarding the error message, did you check the instructions, which explicitly mention that you can ignore those warning messages?
http://rayshobby.net/?page_id=732#uploadWhich firmware are you using? Firmware 2.0.5 and 2.0.6 have changed the way rain delay / sensing works, and they both bypass rain delay sensing for manual operation and run-once program.
Cool. Thanks for the update. Do you know if the latest Ubuntu for BeagleBone Black has fixed the driver issue? We will be making a new image soon using the latest Ubuntu version, and hopefully the driver issue is fixed in the latest version.
a) I need to open the box to install a 2GB microsd to activate the log ? Not sure if I’ve seen instruction on how to open the box and am afraid of ruining it 🙂 The bottom of the OS box is screwed to my larger waterproof box. Help please 🙂
There are four screws at the back of the controller. Just unscrew them and you can open the box. If you’ve mounted it to your waterproof box, you need to unmount it first, in order to access the back of the controller.
b) I’ve ordered those little blue blobs to protect from lightning. Instructions on how to use them ? EDIT -> Found the picture (but need reply to a 🙂 )
Assuming you have the fully assembled OpenSprinkler, you do not need these MOVs — the fully assembled OpenSprinkler has built-in TVS, which are similar to the MOVs.
c ter) How do I know which HW version I have ?
Open a browser, type in the controller’s IP address, and you should see the homepage. The first line of text on the home page specifies the firmware version.
d) I do not understand how to use the Weather button. It does not seem to do anything,
The weather button should open a new page that shows the WUnderground weather report. It doesn’t do anything yet, just to show weather for your information.
4.87 volt on the 24V AC pin is clearly not right. Just to be sure: the 24V AC pin should be measured using AC voltage level on your multimeter — if you measured it as DC, the reading would not be correct.
In any case, if the 5V pin is giving you correct voltage, I think at least the power supply section is working fine. Are you not getting 24V output on any valve, or just some valves? Do you see any visible damage on the board?
on version 1.8.3, if I try to add an interval program, the sprinkler basically locks up.
This is strange. Would doing an EEPROM / hardware reset help? (although when you upgrade 1.8.2 to 1.8.3 it should trigger an automatic EEPROM reset). How did you upgrade the firmware: using the firmware updater tool that we provide, or uploaded the program in Arduino IDE?
There is no difference between 1.8.2 and 1.8.3 regarding the rain sensor functionality. If it doesn’t work in 1.8.2, I think it may be a hardware issue.
Frankly I am not so fond of the microSD approach: think about having to open the case, take out the SD card, copy a file to it, put it back and close the case…
The current procedure is really only problematic for Windows 8 (and perhaps also the latest Windows 7 64-bit) users. The driver works fine on other Windows versions. The main issue with Windows 8 is that it enforces signed drivers, the only way to get these open-source drivers installed is to reboot the system into a mode that disables driver signature enforcement. It’s doable, just involves more mouse clicks.
I see, I forgot that NanodeRF uses a bootloader. The program you are trying to compile is version 1.8.3, which uses close to all the 32KB flash memory. As a result it can’t allow a bootloader (which takes 512B in the case of NanodeRF I think). Since OpenSprinkler 1.2, the board has included a built-in USBtinyISP (or USBasp) programmer, and hence the microcontroller does not use a bootloader anymore, freeing up some space.
There are two work-arounds you can use:
1. try to reduce the code size by commenting out features that you don’t need, such as the part corresponding to manual operation mode. You really only need to free up a few hundred bytes of flash memory space to make it fit.
2. alternatively, if you have an ISP programmer (such as USBtinyISP, USBasp, or use Arduino as ISP), you can use that to directly program the chip. Note that this will overwrite the bootloader, but if you need you can always put the bootloader back by using the ‘Burn Bootloader’ option in the Arduino IDE.I did think about allowing more than 1 master valve, but frankly I don’t know how common this is. For the moment, because the software allows concurrent (or parallel) running mode, you can manually set how you want valves to open/close in any pattern you want. Certainly not as convenient as having the algorithm automatically associate a valve with its corresponding master valve, but it can be done with the parallel running mode.
The IP Stack(or the monitoring code around it) has always been a little strange as the connectivity bars display in the lower right corner would eventually show disconnected(little X) even though it was fine. I was hoping 2.0.6 would fix that. Alas, it did not.
The monitoring code does the following: it pings the router (gateway) every minute to see if it can get a response. If it can’t get a response twice in a roll it will display a disconnection icon (even though the connection may still be valid). Also, if ‘Auto Connect’ is on, it will try to re-establish network connection. If DHCP is used, this re-connection can take a little while (the time out is 30 seconds). During the re-connection process the controller appears unresponsive, but it’s actually still alive. As soon as it re-establishes connection (or times out), it will come back to perform program scheduling etc.
Although uncommon, in some cases routers are set to reject ping requests, in which case the algorithm will clearly not work. In any case, I will figure out a more robust way to test connection and implement the auto connect feature. For now, turning Auto Connect off is probably the best work-around.
@jgrylls: when the device locks up, can you check the LCD and let me know what it displays? Does the LCD respond to button press?
@HFTobeason: the firmware updater tool now includes firmware 2.0.6. Please follow the download link and instructions here:
ttp://rayshobby.net/?page_id=732#uploadWhen you changed the status of the rain sensor, did you refresh the webpage? Note that the homepage is static and does not refresh automatically. The LCD does update in real-time though.
@iDoug: found the bug (for some reason the program worked fine on my Macbook Air but not on an older Macbook). Anyways, it’s fixed and the firmware updater has been streamlined, mainly consolidating all versions and pulling the executables out to make them more obvious. Please download the new package from here:
http://rayshobby.net/?page_id=732#upload
Let me know if this works. Thanks.Hi Mike,
Thanks for your suggestions. Sorry that it took me a while to get back to you since I needed some time to respond to each item on the list. See the relies below:
1) Give each program a “Name” as created
Yup, this can be added. It will take up more EEPROM space, but can be managed if we limit the length of the name. Sorting in the preview can be supported either (just a matter of changing Javascripts).
2) Change “Every” to “Run Every” (Add a New Program page)
Agreed.6) With the current method (All stations run the same duration) “Duration” should be changed to “Program Run Time”
3) Implement the “Run Every” option as either on or off (If turned off, #4 below is enabled)
4) To complement the “Run Every” option, allow the user to set Multiple Start Times for each program
5) Change label of “Time” field to “Program Start Time”
6) With the current method (All stations run the same duration) “Duration” should be changed to “Program Run Time”
7) And as previously noted in other threads, each station should be allowed an independent duration or “Run Time”. PLEASE!
9) Display Station Numbers in “Select Stations” section of Add a New Program pageYes, these are all being considered as a major revamp of how programs should be specified. See the related discussions here:
viewtopic.php?f=33&t=693&hilit=repeat+count10) Preview per station
This may be a bit tricky to implement, but something easy to implement is to add the support (in the preview page) such that when you click on a station it highlights all time blocks that this station is scheduled to run. Not exactly a per-station view, but is perhaps sufficient.
11) Run-Once Program, include a “Start In __ minutes” option
Sure, this can be done.
12) Program Preview, add “Previous Day” and “Next Day” buttons
This is already available in the mobile app’s preview feature.
13) Set Stations, add checkbox to disable a station
This can be easily implemented when the station attributes feature is introduced. See the relevant discussion here:
viewtopic.php?f=8&t=401&p=2659&hilit=attribute#p265914) Navigation
You are absolutely right! This should be done right now, and I will definitely get to it tonight or tomorrow.
