- This topic is empty.
April 27, 2014 at 8:40 pm #22857
I discovered that my new OpenSprinkler v2.1u is a little eager to get on with its day to get to the start of the programmed schedules. 😉
It’s not really as bad as that sounds…
The clock on my OS runs a little fast. It gains about 10 to 11 minutes in a day. I noticed it the day after I had setup my watering schedules when I thought the incorrect station was running but then realised the first station had already finished because the clock was about 10 minutes fast.
The OS does have an option for NTP sync which synchronises it’s time with time servers. If I’m reading the source code correctly then the time sync looks like it’s only done once per day at which point the time should be corrected.
Some will argue that 10 minutes is not a big deal and to many it would not be, but unfortunately I work in IT (networks more specifically) and I’m quite a stickler for making sure all the kit I work on always have the correct time (we log quite a bit and incorrect times render the log entries pretty much useless for debugging or auditing).
That said, I’m not annoyed or upset that my OS clock runs a little fast as the fact that it even supports NTP is already a significant improvement over my last irrigation controller.
What I’ve found is that if I disable the NTP sync option, save the options and then re-enable it and save again, then OS immediately does a NTP time sync and the clock is correct again.
I’m curious to know if anyone else has noticed if their OS clock is running fast or slow?
As a work around I’ve written a little script that runs on my LINUX server here at home that runs via crontab to disable and then re-enable the NTP sync option thereby updating the time.April 28, 2014 at 6:14 pm #26599
10 to 11 minutes a day is indeed out of the normal range, it’s likely a problem with the RTC or crystal. First, can you see if the RTC on your board is DS1337 (2×10 pins, relatively big), or is it DS1307 (2×4 pins, small)? We’ve received two reports of clock issues related to DS1337, and I suspect it’s that particular batch of chips that had problems. If you have DS1337, it’s likely to be affected too. If it’s DS1307, we haven’t received reports of clock issues, but it still might be the crystal that has a problem.
The NTP sync interval can be changed in software. Right now it’s set to sync every 24 hours, but can be changed to every hour or something. This does require re-compiling the firmware and flash the controller.April 29, 2014 at 2:39 pm #26600
The RTC is a DS1307 with the small crystal (as in the picture in the build instructions http://rayshobby.net/wordpress/wp-content/uploads/2013/12/IMG_0003.jpg).
I read in another post (regarding the faulty chips) about shortening the connection from the crystal to the chip. I have mine soldered as shown in the build instructions but I could try solder the crystal directly to the chip if you think it’s worth a try?
My script seems to be working well so far so I’m not overly concerned at this point.
- You must be logged in to reply to this topic.