I have a din rail kWh meter with 1000 pulses a kWh. I measured that the kWh on the display of the din rail kWh meter, is perfect in line with the measuring of the solar inverter.
But: In dash there is around 14% more kWh than on the meters...
I looked at syslog and this is what I saw:
The counting is not good!
For example, this is the counting of flukso: 1,2,3,4,5,6,7,9,0,1,2,4,5,6,8
I saw my gas sensor the counting is perfect, but my watersensor has some some sometimes a slippery too! But not as much as my dinrail kWh meter.
14% is quite a lot!
What could this be???
Any suggestions?
| Attachment | Size |
|---|---|
| port2 flukso.jpg | 97.77 KB |
| Port3 flukso.jpg | 92.64 KB |
Hans, have you tried resetting your Flukso? My guess would be a defective Flukso meter. But I'm not sure off course.
1/ Make sure the sensors are properly connected to the screw terminal. Loose wires can trigger faulty pulses.
2/ Try shorting the + and - of the pulse input (with a meter constant set to 1) and verify in the syslog that the count increases by 1 each time.
Hello Bart,
Thanks for helping. I'm sure the cables are connected properly, I always pull the cables after screwing to see if they're good locked.
Yesterday I switched the cables from port 2 to port 3, and I'll look if it is still the same. I have set the constant to 1, after setting it to 0.877, because I got no grafics. But the reason why I've set it to 0.877 is that I have a 14% more kwh than the meter. But now I know in syslog, the counting is not good.
If I have the same trouble with port 3, I think my din rail kWh meter s0 is not good. Maybe too low voltage?
I will also check your advise to count with a single wire and shortcut the loop a few times. The + to the -
I'll come with new information in the evening...
Hans, about not getting any graphics, well, I noticed that when I use a Constant with a decimal, I don't get any readings either. Changed it to "1" and my solar and waterprobe started to work with graphics.
I don't know if it has anything to do with using a point or a comma to enter the decimal number. Haven 't tested that yet. But strangely, eventhough my solar meter uses a pulse of 0.01, i still get a correct reading when using a Constant of 1.
Compuprof, my kwh meter give 1000 pulses a kwh. The constant then should be 1. So every watthour there is a pulse. If your have a good graphic/results, your solar inverter gives also 1 pulse/watthour.
Otherwords 1 kwh 1000 pulses or
1 pulse 0,001 kwh
If im true
It says 100 imp/kWh. And yes, I was wrong when entering the constant of 1. I made it 10 and now the readings are correct. But most important: working with decimals, mine fails. Working with integers, succeeds. :-)
That still doesn't explain your counting problem.
I believe there is NO 14% wrong displaying done by the graph on the website. I believe it is the consequense of the missing counts (which will be approx. 14% of all countings). If you get what I mean (my English isn't 100%...) :-)
Once you sorted out the problem of wrong counting, you will get perfect results with Constant = 1 and have no need to use Constant values like 0.877.
Perhaps the new upcomming firmware might resolve your problem?
Today the flukso count was perfect... On my din rail 1,51 kWh(it has only 2 decimals), on dash 1,507.
So this is what I'm talking about!! :)
But what is the reason on port 2 it doesn't count good?
I don't know...
I have tried with a wire the function, but I did it fast. Maybe too fast, so somethimes it goes like this: 1,2,5,6,7,0,1,2,4,
I wondered how many pulses the flukso totally can handle, it seems 1 pulse/second.
But I have 4 sensors, and maybe this week the 5th arrived.
The one which will count my woodburning stove with warmed water, to see the outcoming flow.
I wondered if this is not too much for flukso?
Please see the picture above in my first title, I attached 1...
You are right compuprof. The missing counts are the 14% which flukso count above the normal count. So thats why I get 14% higher.
But the question now is, what was it? Is it a fault in port2 of flukso?
Didn't I connect the wires not good? (I didn't believe that)
Or is the connection of the connector (the pins) not good?
Port 2 and 3 are software and hardware technically identical on a FLMv2b; they use a switch circuit to either route a clamp via a voltage-divider to an analog-digital-conversion port of the micro-controller or use this port's pin-change-interrupt feature to count pulses without the voltage-divider necessary for clamp interfacing. If the count (PCInt mode) is higher on port 2 than on port 3 assuming the same (!) setup, that is cable and S0-meter used, this indeed sounds like a bouncing effect. Do you use twisted pair cable? Have you tried a ferrit core around the cable close to the FLM, in case there are crosstalks between other parallel wires? It is more than 1 pulse/sec possible as S0-specifies a pulse length of around 30 to 90 milliseconds (30-10 Hz) and 30 ms are quite long for the uC's code...
I used twisted cable... Should this be the problem?
http://www.bespaarbazaar.nl/din-rail-bruto-productiemeter-p-3260.html
90 ms. So I better can take internet wire or something? Ftp utp Cable? I now uses audio Cable... 1,5 mm2
FYI: I used a telephonecable to extend. It was a 4 x 0,6mm² cable and works fine for my water probe & solar S0 DIN meter. (Also cheaper than audio cable.) But my gas probe isn't working yet. Maybe the same problem of interferrence or might also be a wrong positioning of the sensor on the gasmeter. Still have to find out.
I also considered STP (SHIELDED Twisted Pair) cable, but haven't had the time yet. :-)
I use old Cat 5 network cable.
One run I have is ~30metres
Bart,
There is really something wrong with port2. See the pictures. I have the same sensor set to port 2 and set to port 3.
Port 2 count pulses there were there aren't...
Same settings, nothing changed.... only change the wires.
So do you know what the problem is?
@Hanskraayeveld
Please read your the personal mail that i sent you.
Or https://www.flukso.net/content/accuracy-flukso-solar-measurement.
There is definitely something wrong with port 2 on the flukso
Regards
I have installed the watermetersensor to my woodstove watermeter. First I thought I installed it fault. I change the place more than 10 times. Definately I changed/installed only the wires from port 2 to port 3 and now he is running fine! Not even 1 litre difference over more than 200 litres...
So Port2 is not good. Not on mine flukso.
So I need all the pulse ports... 1 for water consumption, 1 for gas, 1 for solarpanels, 1 for woodstove water
How can this resolved?
sorry port 2 to port 5
Bart, is it possible in my case to use port 6a and b? I have 1 current clamp for main electricity, to port 1. My solar din rail to port 3, gas to port 4, and now my wood stove water to port 5. My main water is now not installed. Is it possible to connect 1 of the listed to port 6? So I don't have to use port 2...?
Unable to send e-mail. Please contact the site administrator if the problem persists.
Your message has been sent.
Is there anyone from the business of flukso? Who can help me?
How long is the guarantee of flukso?
And where is the help? How long did it take, before being helped?
Come on Bart, lend a member a hand :)
It's a little bit frustrating. I did almost everything to contact. Did anyone know how to get attention? Is there vacation? I tried to call, emailed, contact forms, forum... Does anyone know an address? Maybe I can write a letter, or visit the store...
@hanskraayeveld I think I was successful in triggering the problem you're experiencing on a test setup, although the frequency of the phantom pulses was much lower than the 1/7 you're experiencing. I've added a 5ms mask after each pulse. With the filter I was not able to trigger a single phantom pulse out several 100's being generated. I would like to try this new firmware on your FLM to verify if this solves your port 2 problem. Let me know if you're ok with this, connect the DIN-rail meter to port 2 and then leave the FLM running as is.
Okay Bart!
Thanks! I did not know you're busy with it...
I will connect port 2 to the din rail kwh meter.
I hope you'll install the new firmware soon.
Please let me know when you have done it.
Thanks. Hans
Bart, I changed my solar panels din rail meter to port 2 and my woodstove water to port 3.
The 1/7 or 14% was when connected to my solar panels din rail meter. But I put it to port 3.
How much is was connected to my woodstove, I didn't calculate...
I have now the grafics from my wood stove on my solar panels, and the grafics from my solar panels to my woodstove...
I've just updated your FLM sensor board's firmware. I guess we'll have to wait for the next sunrise to know if the debouncing code fixes your issue.
My din rail shows 27,70 kwh. Yesterday 26,34.
So thats 1,36 kwh.
My dash is 1,351 kwh.
So it can be without problems. If it yesterday was 26,349 it could be 27,700 today.
Tomorrow I will update. It seems it is working good.
Oh I forgot he counts yesterday a few watthours. I changed the ports around 17.00, when the sun was allmost down. So I think, you have done a great job, Bart!
Thanks for helping me!
Today from 27.70 to 29.56. Thats 1.86 kwh. In dash 1.868. So it is perfect working...
Thanks for the feedback. The debouncing code will be rolled into the next official release.