Forum Replies Created
-
Posts
-
The LCD display is purely software (because the controller doesn’t have a feedback mechanism to tell if a solenoid is actually on or off). So if the LCD doesn’t display E1 correctly it’s gotta be a software issue. I will check to see if I can reproduce the issue.
Very interesting. I do have some technical questions because I think there are a couple of non-trivial aspects in your approach:
1) If I understand it correctly, the relay board you are using is ‘active low’ — which means a logic LOW activates the relay and logic HIGH releases the relay. Correct me if I am wrong.
2) You can connect an OpenSprinkler’s station port directly to the relay signal pin because the station ports are essentially open-collectors (albeit driven by triacs instead of transistors or MOSFETs). This matches well with the ‘active low’ relays because when a station is turned on, that station port is connected to GND, thereby generating an active low signal which in turn triggers the relay. It’s important that OpenSprinkler and the relay board must share a common ground, which is achieved by your USB cable that provide 5V power to the relay board and the wiring ensures that they share a common ground.
3) This is the interesting part: normally triacs are not meant to work with DC voltage / current because they require a zero-crossing to properly turn off. I am guessing that it works in your case because the relay board uses opto-coupler which consumes very little current. As a result, when you turn off the zone, the DC current running through the triac is small enough that ensures the triac will turn off. This is not the case if the current flowing through the triac is larger than a few milliamps — the triac will stay on and will not turn off. That’s why triacs are really meant for AC current and not DC current. I think it happens to work in your case because the DC current is very small.
Anyways, it’s a clever approach and I didn’t think that the triacs would work with these relay boards (therefore previously I have been telling people to replace triacs with transistors or MOSFETs).
A relay has two ‘coil wires’ or ‘coil pins’. Just connect the two coil wires like how you would wire a sprinkler valve. Regarding which two are the coil wires, you have to check the relay’s datasheet. For example, on the Omron relay:
http://www.amazon.com/gp/product/B005T742IA
the two gold colored pins at the bottom are the coin pins.
For the RT314524 relay, check the datasheet:
http://www.mouser.com/ds/2/418/RT314and424-205089.pdfI thought you said OSPi gives 5V if fuse 1 is bypassed — that means it’s working. What other issues are there?
If you have a multimeter, you can open the enclosure (there are four screws at the back of the device), and use your multimeter to measure the resistance between VIN to GND (these pins are available in the pin out area close to the top edge of the PCB. Let me know the resistance value.
I’ve responded to you on the support ticket. For the record, here is what I said (in case other users need to know):
If you do want to use your current RPi, note that the 2×20 pins are backward compatible with the original 2×13 pins — specifically, the first 2×13 pins on the new RPis exactly match the 2×13 pins on the old RPi.
It’s a bit tricky to explain the wiring without holding one at hand. So I will try my best:
First, open the OSPi user manual:
http://rayshobby.net/docs/ospi14_manual.pdf
On page 3, look at the ‘Plus’ version. That’s how the RPi is plugged into the pin headers. Your model B would be plugged in the same way, except that the yellow RCA connector would block it i the way. Anyways, this gives you an idea of which pin should go to which pin.In fact, OSPi really only uses the first 2×8 pins (go to page 10, look at the Pin Use graph, that shows which pins OSPi uses). So you just need to get either one 2×8 ribbon cable (male to femal) or two 1×8 ribbon cables to make the connections as described above. That’s all. Hope this makes sense.
The on-board fuse is a small SMD fuse. Part number is: CC12H2A-TR
Note that below that fuse are two pin holes, connected in parallel to the SMD fuse. These are meant for through-hole fuse which are probably easier to solder than SMD fuse. You can get a through-hole fuse like this one:
http://www.digikey.com/product-detail/en/3JQ%202-R/507-1643-ND/1010023
solder it to the two pin holes and that’s it.The green LED does look very bright. This is probably because the manufacturer used a green LED that’s different from our original spec. The current flowing through the LED is about (5-2)/1K = 3mA, which is pretty normal. But if the brightness bothers you, you can remove the green LED or the 1K resistor next to the LED.
@stewa9: if it’s Orbit brand, I am pretty sure it’s 24V AC.
I’ve never heard of the model B1+. I think the only available models are RPi 1 A/B, A+/B+ and RPi 2. From your description it looks like you have the RPi B. Note that OSPi 1.42+ does NOT work with the original A and B models — it only works with A+/B+ or RPi 2. Because the pins are backward compatible, you can still use ribbon cables to connect RPi B with the pin headers on OSPi. However, honestly I think the easiest solution is to just buy a A+, which is the cheapest RPi and it would work nicely with OSPi 1.42+.
“Today I finally made the 24V connection and updated the FW via USB to version 2.14.” — does this mean that both 24VAC and USB are plugged in at the same time when you did the firmware upgrade?
If the controller can power on over USB, that means the microconotrller is alive. I am still a bit confused by your description: has the controller ever worked with 24VAC, or has it never? If it has never, you should check the orange terminal block and make sure the wires have good contacts with the terminal block pins. If it worked with 24VAC before but not now, perhaps the fuse is blown or other components were damaged.
Note that it will only skip the station run if BOTH conditions are met. The first condition (watering percentage is less than 20%) is to make sure the user didn’t intentionally program a very short water time. For example, if someone programmed a water time of 10 seconds to begin with, this should probably not be skipped because it’s likely the user’s intention to water for a very short amount of time.
The relay seems fine and should work with OpenSprinkler. Could it be that you wired the relay incorrectly? Note that the correct way is to connect the two coil wires to OpenSprinkler just like a normal sprinkler station (i.e. one wires goes to COMMON, the other goes to a station port). The coil resistance should be around 30 to 40 ohm as the datasheet suggests.
It’s possible that the fuse on OSPi has burned. To check you can use a multimeter to measure the resistance across the fuse (it’s a small brownish component right above the green COM terminal block and it’s marked N). If the resistance is not close to 0 that means it has burned out.
The relay I used is the 22:22 — I am not sure what this means, is 22:22 a model number of the relay? Is it 24VAC relay or DC relay?
Hmm, this is not a known problem (especially since the LCD shows master station is on). Note that the latest firmware is 2.1.4, which has been available since a few weeks ago.
So I am able to reproduce this problem. The issue has to do with the fact that the the rain sensor checking is not bypassing master station when it should. The work around is as follows:
1) go to Edit Options and disable Master (i.e. select Master as None)
2) Now the master station becomes a normal station, you can edit its attribute and turn on the ‘Ignore Rain’ flag.
3) go back to Edit Options and enable Master (i.e. select a Master station)This should fix the issue for now. We will address this in the next firmware 2.1.5. Thanks.
@Randy: I don’t understand why your OS2.2 has 12MHz. Did you buy it from us? All OS2.2 units come with 16MHz crystals.
We don’t have a current part list for OSPi 1.4+ — first of all, you can open the schematic in EagleCAD and generate a BOM using the bom script; second, most components in 1.4+ are the same as previous versions, there hasn’t been significant change.
Sure, you can host web app on local server. Check this support article:
https://opensprinkler.freshdesk.com/solution/categories/5000022938/folders/5000099537/articles/5000164006-using-a-different-server-for-ui-assets@Rick: The package has been received and the unit has been checked. To me it seems something on the SPI bus is interfering with the microcontroller. I took the microcontroller out, put it in a standalone chip programmer, and re-flashed bootloader to it. After that, I popped it back in and was able to upload firmware from the USB port. It’s not clear to me what was the original cause. It could be a soldering related issue, since this is a DIY kit, it could be a flash memory corruption issue. In any case, you will receive an email to arrange for return shipping.
It’s possible but will be a lot of work. The OSPi schematic is all available in Github. But before you start, have you tried to use wired Ethernet with RPi, which is generally more reliable than WiFi?
@Augustin, so it looks like somehow the ‘sunrise to sunset time’ and ‘sunset to sunrise time’ are flipped. To make sure, in general, ‘sunrise to sunset time’ should always go with ‘sunrise time’ as the start time; and similarly ‘sunset to sunrise time’ should always go with ‘sunset time’ as the start time. Although the firmware doesn’t care if you mismatch them, it’s usually a mistake to mismatch them.
