Forum Replies Created
-
Posts
-
it’s a little complicated because the case and PCB screws are all packed together with the PCB and terminal blocks. To provide a ‘no enclosure’ option would require taking these screws out. The OSPis are all packaged in batches and sealed in anti-static bag, Since it’s difficult to predict how many ‘no-enclosure’ orders there will be, we’ve decided to just remove this option to reduce the overhead on our side. If you do not need the enclosure, you can contact us directly and we can make adjustment accordingly. We just don’t want to provide this as a public option. Thanks for understanding.
I’ve compiled a new firmware and it can be downloaded at:
http://rayshobby.net/software/firmware1.8.3_new.zip
This should be able to fix the buffer overflow problem, and allow you to preview a large number of programs. Plesae follow the instructions about “Update a Precompied Firmware” to flash the new firmware to your controller:
http://rayshobby.net/?page_id=732#upload
Basically you need to install USBtinyISP driver (only if you are using Windows), and avrdude program (or WinAVR if using Windows). Then run command:
avrdude -c usbtiny -p m328p -F -U flash:w:firmware1.8.3_new.hex
I didn’t change the firmware number, so your program data should be preserved.I’ve identified the source of the problem (Ethernet buffer overflow) and I will update the firmware shortly today. I know I’ve promised bedo2405 to fix this problem a few weeks ago but managed to forget …
The ‘without enclosure’ option has been discontinued a while back. We don’t offer this any more because it takes time and effort to prepare stock with different options.
You are right, we (including our distributor OpenLab in Australia) have started shipping OSPi 1.3. I didn’t have time to update the webpages yet, sorry about that. Will gradually put updated information online. The main changes in 1.3 are added rain sensor port, fuse, mini-relay, per-station TVS. These changes pretty much follow OSBo and OpenSprinkler 2.1.
The most recent software update is announced in Dan’s post here:
viewtopic.php?f=28&t=467This post is similar to what I just posted in the SquareWear thread, basically, it’s an updated and improved VirtualBox image 🙂
>>
A quick note that I’ve uploaded a pre-configured VirtualBox Linux Mint 13 image to the following url:
http://rayshobby.net/software/rayshobby-VB.7z
this is intended for users who are having trouble installing the USB drivers in Windows 8 and/or 64-bit, and the image has pre-installed Arduino, OpenSprinkler 1.x, 2.x source code, and SquareWear source code. It allows you to set up a virtual Linux (you need to install VirtualBox software) and it can pass through all USB devices to the virtual OS, so you can compile an OpenSprinkler or SquareWear program and upload it to the microcontroller directly in the virtual OS, without any driver installation.Please use 7-zip or WinRAR to decompress the above file (the reason it’s in 7z format is that it’s a more efficient compression scheme that results in 50% smaller file size). Once you decompress, there is a README.pdf in the package with further instructions.
I just downloaded a copy and it seems fine. Perhaps a network traffic problem? Try again, and consider using a downloader like Flashget which can resume the download if it gets disrupted.
You said ‘I’ll make changes to be programmed manually without going online’, which made me curious why you still need ENC28J60 (the Ethernet controller) — I thought you mean program the controller with buttons and not with a web interface.
In any case, OpenSprinkler is designed to work at 3.3V. For example, the microcontroller, 74HC595, ENC28J60, microSD card, all work at 3.3V. It does have a 5V line because DS1307 and 1602 LCD require 5V, but these can both interface with a 3.3V microcontroller.
OK, that’s good to know. I suggest doing the following:
1. Measure the resistance between VIN-GND again, and the value should be somewhere around 4.5Kohm to 9Kohm. If it falls below 4Kohm, it’s likely that either there is a short somewhere, or a component is burned (most likely the zener diode, or the LM2596 switching regulator).
2. If the resistance is correct, then something on the 24VAC input line is shorted. There are only a few components on the 24VAC line (fuse, diode D1, capacitor C1, LM2596, so it shouldn’t be too hard to narrow it down.Which GSM shield do you plan to use? Basically it boils down to figure out the interface (SPI? I2C? Serial?) and power requirement (5V? 3.3V? how much current)? If you can post a link to the GSM shield I can tell you better how it can interface with OpenSprinkler.
The OpenSprinkler user manual explains how to do this:
http://rayshobby.net/?page_id=730#program
check the section entitled “Use the same time settings with different duration times”. Basically, you can set up multiple programs, each with the same start / end / interval time, but with different stations and run times.1. In Arduino, you need to select menu Tools -> Boards -> OpenSprinklerHW 2.1 (12MHz, USBasp, orange terminal block). I suspect I haven’t updated the VirtualBox image with 2.1 profile, so in case you don’t see this option, you can copy the boards.txt in the Github repository:
https://github.com/rayshobby/opensprinkler/blob/master/OpenSprinkler%20Controller/software/arduino-code/hardware/aopensprinkler/boards.txt
to your Arduino installation folder:
arduino/hardware/aopensprinkler/boards.txt
and overwrite the existing boards.txt. This should fix the first issue (i.e. sketch too big)
2. You need to run Arduino either in sudo, or alternatively add VID 0x16c0 PID 0x5dc to /etc/udev/rules.d/xxxx.rules, where xxxx.rules is one of the existing rules files in that folder. I think (need to double check) the one in VirtualBox image is named 60-microchip.rules. Just follow the existing lines to make a new line with the VID/PID pair. This should solve the second issue.Overall the issues you encountered is because I haven’t updated the VirtualBox image yet 🙁 Will get this done soon 🙂
Hmm, if I understand correctly, you mean the fuse is blown when you power the controller by USB? Since the fuse is on the 24V AC line and not on the USB power (5V) line, I can’t think of any obvious reason why this would happen. Did you measure the resistance between VIN-GND, as well as VCC-GND? What are the resistance values you are getting?
Both OpenSprinkler and OSPi support 8 stations, among which one can be assigned as a pump / master station. Moisture sensor is not supported by default but you can modify the software to add support. Both OpenSprinker and OSPi have analog input pins.
For comparisons between OpenSprinkler and OSPi please check the information here:
http://rayshobby.net/?page_id=5816/#faqThe all-through-hole kit has been discontinued. We are only offering semi-assembled kits right now. Most of the passive components like resistors and capacitors are SMT and pre-soldered onto the boards, the microcontroller and Ethernet controller are left for the users to solder. So while it’s not all through-hole, it still provides some flexibility if you want to change / upgrade the microcontroller and so on. Additional details can be found on this blog post:
http://rayshobby.net/?p=8147Sorry about the lack of clarity: we’ve already started shipping OpenSprinkler v2.1 which has on-board per-channel TVS (transient voltage supressor) for protection against transient high voltages. So you do NOT need to purchase MOVs separately. All orders of OpenSprinkler 2.0 will receive version 2.1. I haven’t had time to change the webpage yet, which should be done within the next few days. Thanks.
1) The schematic is using BT1308W as solenoid drivers. Some schematics still have MAC97 in it. Why did you change to the new ones? Performance?
BT1308W is a surface mount part, and MAC97 is a through-hole part (maybe MAC97 also has SMD version, but not that I am aware of). Their electrical specs are almost the same. I apologize that the schematics are not all consistent. Basically if it’s surface mount you should use BT1308W, and if it’s through-hole, use MAC97.
2) Are the RN2 and RN3 resistor networks only resistors to limit the current or is the ‘isolated’ somehow important? Some talk about opto ccouplers for the triacs or is that just necessary if we would want to switch 220V like main power ?
These are current-limiting resistors. Like transistors, triacs operate under a certain gate current. You need the current limiting resistors to make sure the gate current does not exceed the operating range. These are not for isolation — for that you need to use SSR opto-couplers, such as MOC3063 in conjunction with BT1308W. Even though BT1308W can handle up to 600V, if you want to switch power line devices like 220V device, I would recommend using a relay.
3) I am sure it works fine (I ordered some parts, but could not start assembling anything yet) but isn’t the LM2596 ment to be DC/DC and not 24VAC to DC
You are right, LM2596 is DC/DC switching regulator. The AC to DC rectification is done using diode D1 (S1A) and input capacitor C1 (100uF). These will rectify AC to DC. For additional details, please check my blog post here about 24VAC to 5VDC conversion:
http://rayshobby.net/?p=6002Coming back to this thread: the new enclosure for OSBo is now in stock. It has enlarged cutout to fit an Ethernet cable. You can shoot an email to [email protected] to arrange for an exchange.
I suggest first following the instructions here to set the RTC time:
http://learn.adafruit.com/adding-a-real-time-clock-to-raspberry-pi/set-rtc-time
because RTC was not set initially on OSPi, I’ve seen a few cases where it may cause the RTC to be detected as UU. Setting a valid time may help solve the problem.OpenSprinkler is currently designed to work with 24V AC valves only. It’s probably quite difficult to power these valves with battery / solar panel. Technically you need an inverter that converts DC to AC, but I’ve seen discussions that these AC valves can run on DC voltage (al beit less efficient / draws more energy than running on AC).
We have an upcoming product, OpenSprinkler Bee (http://rayshobby.net/cart/osbee-10) designed to be battery powered and work with latching solenoids. The controller will have an RF transceiver which can remotely talk to a main controller that has Ethernet connection. It’s probably also possible to add a bluetooth module to allow phones to directly talk to it. I keep telling people that I am working on it but the truth is that I haven’t made much progress yet 🙁
Interesting. I was talking to a friend a while back, who designed her own chicken coop, and we were talking about how to construct an automated chicken coop. If you google ‘Arduino chicken coop’, or ‘raspberry pi chicken coop’ you should find some existing projects.
Cool, glad to hear that it finally worked. Indeed it’s surprising that it suddenly just worked. Perhaps it was a wire connection problem?
The USB power thing is perfectly normal: the switch only controls 24V AC line, and does not control the USB power. So the unit will power on whenever you plug in USB cable.
The soldering quality seems pretty good, and I can’t see any obvious problem. One thing to double check if the four pins on the USB connector: those pins are very close to each other and are hence prone to solder bridges.
The resistances you reported are indeed quite low and indicate a problem. The VIN-GND resistance should be about 4.2 to 4.8 kOhm, and VCC-GND resistance should be even higher. So chances are that either there is a solder bridge somewhere, or a component has been damaged after connecting to the sprinkler transformer.
I forgot to ask: does the LCD light up when you power the controller via USB?
Thanks. Those are great comments, will definitely take them into account when I update the files next time.
Some notes regarding your comments:
1. About flash eMMC: to be honest, I was quite confused in the beginning myself how it works. I had thought that it’s like RPi: you pop in the SD card and that’s it. Later I discovered that you need to press the boot button when starting, and wait for about 10 minutes for it to flash into the internal eMMC. So I updated the instructions but perhaps there were some earlier instructions, particularly what I said in the video were outdated.2. Hmm, I have always used 2GB uSD card for testing and it seems to work fine. Will check again.
3. The password is not documented in the README.txt, instead, it’s documented in the homepage after the Bone boots from the pre-configured image. I’ve planned it this way so users will only see the relevant instructions step by step. For example, README.txt only describes how to install the pre-configured SD card; after the installation, the homepage describes how to run the interval program, and use the mobile web app, and there are further links to the user manual of the interval program etc. There are lots of information and I hope not to overwhelm the user right in the beginning by putting them all in the README.txt. So I can definitely put the password in the README.txt, but I just want to explain the stories behind 🙂
