Current Cost Support Forum

Summer's drawing to a close here in the northern hemisphere, and it raises an interesting question about the collected history data.

Currently, the history aggregates usage into bins for each 2 hour slot. When I set my clock on the CC device back an hour, does that fill a slot with 3 hours of data, or possibly skip a slot, and therefore potentially collects only 1 hour of data depending at the time I change my clock?

How about if I change my clock by a larger amount (say forward 8 hours)?

I think because of those issues and the fact the internal clock is not a good time keeper I would ignore the history from the unit and use a more accurate time source e.g. the connected computer.

Just read the real-time data and ignore the rest and let your application build the historic data from that (with of course more precision).

Automan.

I'm going down that route, but I don't really want to leave my PC on 24 hours a day.

I'm starting to think that what I'll do is use an mBed or an Arduino/Fez micro to capture and process the data myself, and then use it to post to the web somehow.

Of course, that rather limits who we can sell any software we write to, to recoup the investment.

I used to run my weather station using one of these gadgets

http://wiki.meteohub.de/Hardware_Overview

My one of course is now obsolete

Automan.

Re timekeeping (automan's email), classic is bad, mine loses few mins per week but the time keeping of my envi and those of my live_sol beta testers seems fine few secs/week - we aren't talking atomic clocks here, so within a minute or so is plenty accurate enough. live_sol generates a warning email when difference is > 2 minutes so it can be reset if necessary.

The best way to proceed imo - I use it in live_sol - is to keep pc on during the day when there'll be most solar generation and capture the data in real time. Many use their pcs during the days anyway. Then turn off at night if you want to and use the historic data (2 hourly resolution so less accurate) to 'fill in' the gaps. Good enough! It's a lot easier to use/program a pc than start messing with arduinos etc, and with a laptop the extra power used is not great (20-30w). John

I've just been searching around for netbooks off eBay, etc, because while I've done embedded programming, it may well be cheaper for people to just pick up a ready made device.

I think that might be the best compromise, as they're low power, and I could possibly just leave them on 24 hours a day. I might even find some other things to do with it.

I have a Samsung NC10 which uses about 14 watts.

However, my server with SSD and no display or fans uses under 20 watts.

Automan.

Just advanced my clock 1 hour for summer time which started at 2am this morning (2/10/11). At 9:47 Standard time I advanced to 10:48 DST (there was a little clock drift as well).

The following is based on data retained by Techtoniq Energy Station from my Envi (CC128-v1.18).

• Live Data: There was an entry for 9:47 and also one for 10:47(why?). There was nothing in between as I would expect.
• Bi-Hourly: Entry at 9:00, but none for 11am (Why?). I can't see any reason TES would drop the 11am record as it should have seen both the 9am start and 11am finish for the 2-hour block. drivel removed
• Strange entry: TES has retained an entry for 23:59 on 2/10 (has not happened yet). Probsbly due to the clock drift. TES must be combining the Date from the PC with the time from the Envi without looking at the DSB

As far as I can tell TES just displays the content of the History records so It looks like Envi drops the data.

I will try to capture the actual history data to remove TES from the equation. [Might be able to work out when summer time finished and when I adjusted the clock (then again might not!)]

Edit: Oops silly me! The history download hadn't finished when I grabbed the data.
The 2 hour block is there but missing about half the data as seen in .....

_________________
Seeker
"The Truth is out there!"

Thanks for the useful heads up, Seeker.

Do you actually have 3 phases on your system out there?

No I'm on single phase.

Wondering what prompted the Question? If it was the phases shown on the lower set of graphs -- this is a standard chart from TES which shows the last hour of data and allows selection of any of the phases. If it is the 3 vertical sets They are separate tramsmitters. On on the solar feed and one on the collection of active wires after the main switch. The third marked heating is my experimental rig using a dev board which records when the central heating is running (much like an AM) -- still faithfully running on a breadboard!

_________________
Seeker
"The Truth is out there!"