Recently my smart meter has been installed. It took some time to get the meter working properly as the energy supplier hadnt deployed it correctly and all usage was registered on the night tariff (including the pv output). For a few days now it's actually showing the correct usage for day and night tariffs so I decided to hook up my Flukso v2b and start reading our energy usage. I have had the Flukso for a while and only used it for a short time to read out the previous gas meter with a probe. This all worked perfectly.
What I have done so fas is plug in the phone jack of the data cable in the P1 port of the smart meter (Iskra ME382) and disable sensor 1 to 5 as the P1 port should be recognized automagically. This didnt result in any pulses being registered. It just showed the heartbeat every 15 min. in the log. Strange thing I did noticed after connecting the phone jack is that the Flukso would reboot at least every 3 hours or so. When this happens there is something in the log about a user nobody that logged in. With the data cable unplugged again from the smart meter it does not do this. So for some reason connecting to the P1 port results in reboots of the Flukso. In another topic I read that the data cable supplied with the Flukso was faulty so I also created another one from a (working) phone data cable. This resulted in the same behaviour, no pulses and random reboots of the Flukso with a maximum uptime of a couple of hours.
Oh, I don't think it's relevant but just to be sure. I edited the firewall so I can access the Flukso through Wifi.
I really have no clue where to continue from here so it would be great if someone could help me out so I can finally get this up and running. Below is a clean syslog from right after a manual reboot. I will add a log where with a heartbeat and a random reboot later in the next post.
[log cleaned as it's no relevant anymore]
Did you check that the RJ11-connector is connected correctly?
Quote from the description in the shop-ad:"Connect the Rj11 pins as follows: pin 2 > 6a, pin 3 > 5-, pin 4 > free, pin 5 > 6b. Pin numbering is from left to right when the RJ11 connector has its lip pointing upwards."
If reversed, there is no GND potential to the data link and data-lines are opposite direction what may be the cause for the reboot according to Murphy's law or Dogbert's tech support ;-)
Your log just shows the boot sequence without any sensor registering
Regards, Markus
Thank you for your reply Markus.
I did check the pins of the data cable after noticing it wasnt working. The description from the shop isnt quite clear to me though as the data cable that came with the Fluksometer only has four pins/wires and the description is talking about pin 2 to 5. So I cant really check if it is connected correctly but I just assumed that Flukso knew what they where doing when creating this cable and supplying it with my Fluksometer. I also copied that same setup (also a 4-pin cable) on a data cable I created myself (to make sure the data cable that came with the Fluksometer wasnt broken).
I did notice that the Fluksometer hasnt rebooted with the data cable I created myself. But that could very well be because no sensor was detected at all as you mention from the log above. This evening I will reconnect the data cable that came with the Flukso and post another log after a reboot occurs.
(Or should I get myself a 6 pin data cable?)
No - 6- and 4-pin cables are physically identical; the "4-pin" is just the inner 4 pins of the 6-pin (1@4=2@6 to 4@4=5@6; 1@6 and 6@6 are not connected in the 4-pin variant - see http://en.wikipedia.org/wiki/RJ11#RJ11) - see also https://www.flukso.net/content/cant-get-flukso-work-smart-meter-and-corr... - so the issue may still be that you connected the cable reverse to the srew terminal of the FLM...
Happy trying :-)
It is "to be precise" 6P4C - 6 pins with 4 connected... - the RJ-something is quite misleading
Thanx for clarifying that. The data cable was already connected when I unpackaged the Fluksometer. But of course it could be that they made a mistake during assembly. I'll triple check everything once again with these new insights.
Well, after fiddeling with it I now have different errors in my syslog. Still no pulses though. Below is a new log. One thing that isnt clear to me is that the following line in the instruction: "from left to right when the RJ11 connector has its lip pointing upwards". Does this mean when viewing it flat on the table and the lip pointing upwards or having the connector in your hands and turning it around so the lip (and the hole for the cable) is pointing upwards.
And when doing these commands through SSH I see correct output now ... but I still dont see pulses in the log. Just these errors as mentioned above (and the cable that came with the FLM was incorrect. Wire 3 was cut instead of wire 4 which was connected to the FLM and which is why I had no registered sensor :-| ) .
See my earlyer post, same issue. You have a liander smart meter? Some how you have to sent a start signal. Bart (icarus) can help you with it. For me now it works for rasberry pi the same problem.
Regards Richard
It's supplied by Enexis but it's an Iskra ME382 which is the same as yours. Bart seems to be on vacation till the 15th of august so I guess I have to be patient a little more.
I'm not sure if it's the same problem as yours though. The cat /dev/ttyS0 command through SSH gives me proper output where I can see pulses with proper values. So the FLM is able to receive the signal from the P1 port of the smart meter but somehow it's not getting submitted to the Flukso website presumably because of that error in the log every few seconds:
Not sure if it helps but this is the output from 'cat /dev/ttyS0':
I'm pulling the plug from the FLM now in the hope Bart can help me out when he gets back from vacation.
I am also on holiday whats your firmware version? 223?
You have the same meter so it should work, maybe just a reset
Yes, I'm on the 223 firmware. I'll have Bart look at it before doing a reset. Hopefully he knows what is going on and why these errors appear instead of the pulses that are clearly being send to the FLM.
Has anyone spotted Bart already after the holiday break, this issue still hasnt been addressed and I get no response to a support question through the contact form ... :-|
I just had a quick peek at your FLM. Everything seems to be up and running, your configuration is correct and has been synchronised with the Flukso server. However, I'm not seeing the same output on /dev/ttyS0 as what you posted in the comment on 10/08. This is the 'rubbish' coming out now:
I don't know whether your FLM is currently hooked up to the smart meter. But please check the connection again, and report back via the forum.
Cheers
/Bart
Thank you for the response. I currently do not have the FLM connected to the P1 port because of the syslog was being flooded with these error messages every second, which just didnt seem good to just leave it like that for days.
I will plug it back in up this evening (CET) and leave it like that for you to check.
Where you able to check it now it's plugged in Bart?
Seems you're generating so much solar power that your electricity counter is reporting negative values. I've got a code fix for that, although there's still an issue plotting negative values on the flukso.net charts. That one will be fixed at a later date. Could you reboot the FLM once more?
We received the solar panels right before summer and the smart meter a little later then that. So yes, our generation is currently higher then our consumption. This will change during the winter but it's good to see you have a code fix so I dont have to wait until then to start using my FLM.
I have executed a reboot command through SSH, I still see the same errors in the syslog though. Not sure if these should have been gone now.
I've just patched /usr/sbin/fluksod.lua with commit [1]. It's now accepting negative counter values. You can double-check via the syslog, or by subscribing to the mosquitto broker:
If you're talking about this error in the syslog:
Don't worry. It's just a mosquitto client subscription timing out. It's re-established once fluksod has new readings to post to the broker.
[1] https://github.com/flukso/flm02/commit/47c59e1841a9d891e0d28fc4075913034...
Awesome, I see pulses in the syslog and my dash is showing results. Thank you for your efforts!
Could you explain me something?
The p1 telegram contains values for current power consumption and current return delivery. Both can not be negative values. If the generation is higher then the consumption the current return delivery will kick in and the current power consumption is zero. The P1 telegram also contains the values of the total taken and returned peak/off-peak power. These are not negative as well as far as I can see.
I did subscribe to the Mosquitto broker and I do see negative values there but I dont seem to comprehend what is going on based on the values and data in the P1 telegram. So my questions are:
- What value is exactly currently negative and where does the mosquitto broker gets this data from?
- What value is the smart-main exactly showing on the dash, consumption or consumption - generation? I guess it's the latter because if it only was consumption it could never be negative. It would be great of both could be plotted on the dash.
- Once the dash will be fixed to show negative values, will this also effect the api for pvoutput.org.
What I'm trying to achieve with the FLM is that I can show a graph on pvoutput.org that plots the energy consumption on my pvoutput.org. Because of these negative values I'm not sure if the data seen now is what I'm looking for.
http://pvoutput.org/intraday.jsp?id=22252&sid=20144
The FLM subtracts the production reading from the consumption one. So a negative consumption means net injection into the grid. If you want to obtain gross production and consumption values, you could add a DIN rail meter in series with the inverter. You'll then have to do a little arithmetic to get gross consumption based on the aforementioned readings.