Hello!
Same problem that quite some people have here with measuring solar panels with a clamp.
An overview... maybe someone sees the logic:
- at night, there is a constant measuring of 50 Watt on the Solar circuit.
- if I switch off the fuse of the solar installation this drops to 25 Watt?
- during day, the reading is 50-100 Watt higher than the numbers shown on the inverter.
So I'm thinking to buy a phototransistor to stick to the red LED on my DIN kWh meter installed. Any pointers to a model that should work are welcome...
As far as I understand it doesn't matter how many volts or mWatt it produces?
Greetings,
Pieter
Hi Pieter,
Your DIN-rail kWh meter will most likely have a S0 pulse output that you can connect directly to port 4 or 5 of the FLM. Have a look at the Fluksometer Manual [1] for further details on how to properly configure the port.
[1] https://www.flukso.net/files/flm02/manual.pdf
Cheers,
-Bart.
Nope, no pulse output sadly :-(...
Can you physically move the clamp so it is not next to other wires or circuit breakers? I found that the clamp was misreading in one location, and by moving it further away from the circuit breakers it started to read more accurately.
I tried placing the clamp in different locations... but it didn't change the reading at all.
The different places were all in the electricity board though, as there isn't any other place where the (XVB) cable is stripped so I can put a clamp over the fase wire. The "best" place was +- 10 centimetres from other cables and ciruit breakers.
There is one change though since my first post: regardless of where I put the clamp, if I put off the circuit braker of the solar panels, the measured power drops from 50 Watt to 0 watt instead of from 50 Watt to 25 Watt before...
I had the same issues. There is I believe cross induction to the clamp.
I fixed mine perfectly with advise from here by putting a Kw/H meter on the DIN rail.
In the pic you can see the current clamp still in place . I left it there to see the defference but its output is now off in sensor setting.
http://i8.photobucket.com/albums/a19/bazzslk230/IMAG0315.jpg
http://i8.photobucket.com/albums/a19/bazzslk230/IMAG0570a.jpg
http://www.schnap.com.au/kilowatt-hour-meter/single-phase-electric-kilow...
I ran the pulsed output to the Flukso 2 and adj the multiplier in 'sensor setting" until I was happy with it.
Shows '0' at night now and Solar output during the day.
The 'Flukso' app at PvOutput.org works real well at showing solar output and current power usage.
http://i8.photobucket.com/albums/a19/bazzslk230/4f68bbeb.jpg
http://i8.photobucket.com/albums/a19/bazzslk230/4f68bd62.jpg
http://i8.photobucket.com/albums/a19/bazzslk230/4f68bdb4.jpg
Bazzle
Yep... I'm afraid I'll have to do the same... but as I have a DIN kWh-meter already (but without a pulse output), I'll try first to put a phototransistor on it to create a pulse output, as it's a bit stupid to install 2 meters right next to each other to measure the same thing...
Just need to get to the electronics store :-(...
PvOutput.org looks great... I just coupled it with my Flukso and it's already up and running. Thanks for the great advice, this spares me a lot of feature requests to ask for on the flukso.net charts ;-).
For $38 I would add a new k/h meter ready to go on the solar input.. Use the other one for another cct?
Bazzle
I'll continue work on the Flukso charting soon. So please do list those feature requests. What are the features you like on the PV-output charts?
I think something very similar to the PV Output charting would be great and keep it all in house..
Bazzle
The current, present kWh meter is an "official" meter and the serial number is registered with the government to receive subsidies... so not possible to just swap it...
But if the 0,49 € phototransistor doesn't work I'll surely go for a 34 € kWh meter right next to the existing one ;-).
Yep... ofcourse even (way better) if it is in-house!
Things I think about immediately, in order of my personal priorities:
1/ Logical one: get same features for the new charts compared to the old ones
2/ On the month view an option to see the total kWh for each day (alongside the current options for kWh/year, Watt,...
3/ Cumulative view on kWh of both solar and main on the day view
4/ A "line" in month and year views with the 'expected' amount of electricity for your solar panels per month as can be calculated using this tool (for europe) http://re.jrc.ec.europa.eu/pvgis/apps3/pvest.php#:
-> so it should be possible to fill out the expected kWh per month for a solar sensor.
5/ Be able to access data from previous years, or even to overlay different years to compare the same month of previous year(s) compared to now?
6/ Extra metadata for a "solar sensor": kWPeak, brand/type of panels, brand/type of invertor, angle, orientation, location,... and be able to search/browse fluksonians based on that data.
7/ (yet) an extra unit to choose to show in the graphs (alongside Watt,...): kWh/kWPeak -> you also need the kWPeak for the solar installation for this.
8/ Search fuksonians with similar installations (same general orientation of solars, similar angle, within 20 km of one another,...)
-> this is fun to find "reference" systems you can compare with...
9/ Send an email with a warning if the solar panels produce significantly less for X time than your choosen "reference system(s)" (easy to be alerted of a technical problem,...)
10/ Mentioned before... but be able to set your solar sensor monitoring to "Public", so anyone can view it. For the "Main" sensor this doesn't seem like a good idea, paranoia people might fear that burglars can use this info to see you are on vacation.
11/ Possibility to enable showing a graph with the temperature that day/hour/... according to a weather server. Nice to see if there is a correlation between temperature and energy usage (electricity + gas) and/or production (solar).
12/ Would be fun to be able to attach a temperature sensor to the analog ports (instead/alongside weather server mentioned above)... eg. this one (just googled it, I don't know anything decent about electronics, even though I might have to know given my studies :-( ). http://www.facstaff.bucknell.edu/mastascu/elessonshtml/Sensors/TempLM35....
OK... my inspiration is getting low (and it's getting late here)... if you need any more "wild" ideas, shoot, than I'll give it another "think"...
Okay. That's literally a dozen suggestions. Thanks! I'll keep 'em in mind. I might poke you once in a while to clarify some points.
PS: I took the liberty of slightly modifying your comment. The suggestions are now enumerated. Makes it easier to refer to specific points later on.
Poke at will :-)!
BTW... I got myself a phototransistor... but I must have been sleeping, as I asked for an IR one... so obviously it didn't work.
So I just ordered an extra DIN kWh meter :-).
Maybe you could add one to your shop for when people want to monitor their solar panels as it seems a regular problem to get them monitored properly? I can add by the way that I have an inverter with a Trafo inside... possibly that's the common factor for people having misreadings...
Bart,
Do you think it would be a good idea to use an issue tracker to manage the feature requests? This might give a clearer overview for everyone...
Eg.:
http://code.google.com/p/picasastarter/issues/list
Sure. For server-side issues and feature requests we can use the Github issue tracker of the fls project [1]. It might be better to map each feature request to a unique entry in the tracker.
[1] https://github.com/flukso/fls/issues
I added them to the issue tracker.
Could you create some tags in github to be able to specify some priorities?
Eg:
prio:low
prio:normal
prio:high
My extra kwh meter is working in the mean time, and the difference isn't gigantic, but it is significant. As they are working both now, it is easy to compare both:
- at night the +-53 watt of the clamp dropped to 0
- during the day the clamp reports between 50 and 150 watts more than the kwh meter
- the kwh meter reports the same as my "official" meter for subsidies
BTW: got my 1 fase kwh meter from www.kwhmeter.nl for 31 euro including transport costs.
And how about putting some kind of Faraday cage around the clamp?
I have the same problem at home (59 W residual during night).
I will (carefull) try this night with aluminium paper.
I advise against putting aluminium paper in the switchboard. I'd consider it a safety risk.
It also won't work particularly well. You'd do better with a mu-metal box around the clamp to exclude external magnetic fields, but then that might influence the reading in other ways.
@ The_Roggy Switching off your inverter dropped the clamp detected power from 50w to 0w.
This sounds like your inverter is drawing 50watts of standby power. If it was caused by induction into the clamp from nearby power circuits in the switchboard it wouldn't drop to zero.
Does your inverter documentation list its standby power consumption because 50watts sounds like a lot to me.
Putting in the kW/h meter will fix the overnight false generation figures. But will that 50watts consumption be included if you are monitoring house consumption? Probably not. This will mean your consumption figures may be out by however much the inverter draws on standby.
You may end up with a hidden 50watts of overnight base load.
How about getting a timer added to the solar power circuit to only have it connected during the daylight hours? Too extreme?
The 50 to 150 watts extra during the day that the clamp reports, is that instantaneous power or daily total power? If its daily total and you have a large system then it could be caused by the kW/h meter compensating for voltage sag and the clamp not.
In Germany we have a proverb that says "wer misst, misst viel Mist" - who measures, measures a lot of crap...
That is actually an indicator for the fact that computing the power from measuring just the current of an Alterating Current circuit and assuming all other parameters constant (voltage) or being "in phase" (phase of current with phase of voltage) is true only on an average base respectively when integrating over a well-defined time interval T that makes a "transformation" towards the "what would it be if it was a direct current" (damn physics) - this does not make the FLM "worthless" and as there is no claim of the FLM that says something different, the computed averages will always be a very fine approximation of the truth.
See http://wotid.com/blog/night-time-sleeping-load-consumption-of-my-grid-co... for an explanation of what really seems to happen when inverter power over night does not drop to zero. My PV inverter's power consumption is as high as 50W shown on the dash and it also drops to zero if I switch off the main switch to the inverter - explanation is "simply" that the inverter is full of capacitors and inductances that make measuring and computation on AC rather tricky. If this was a real issue, I guess there would be "something done" by the vendors; the spec says "consumption in operation standby/night <10W/0,25W" and I believe that, thus require to think on how accuracy of the PV measurement could be increased (I am not too confident as there is the sentence in Wikipedia http://en.wikipedia.org/wiki/Current_clamp "in conjunction with more advanced instrumentation the phase and waveform are available. Very high alternating currents (1000 A and more) are easily read with an appropriate meter; direct currents, and very low AC currents (milliamperes) are more difficult to measure")
So, please keep off any aluminium paper from the switchboard if you are not sure what you are doing and if the root cause is truly somewhere else...
These are my 2 Euros on this ;-)
Best regards
Markus
Thanks for the link to the explanation. Makes it much clearer.
So the kW/h meter will eliminate the pesky VA figure that the current clamps measure on some peoples inverters overnight.
And that is all that is required as the VA figure is not what people are charged on.
It doesn't register on the kW/h meter so it also wont register on the utilities kW/h meter.
So no problem. Good to know.
Ok, i am not charged for the 59W continuous difference as measured by Flukso clamp on my PV circuit.
But how could I do to correct this measurement?
This 59W delta is also there during the day.
Should it be possible to substract those 59W from the clamp measures, by entering -59 in the FLM or Flukso.net configuration panel?
Subtracting reactive from apparent power will indeed net you real power. The subtraction is however not a scalar but a vector one. See this wikipedia article. Line voltages can also change one day to the next. Just saying that trying to get sub-1% accuracy from a current-clamp-only sensor is not realistic. Inserting a DIN-rail kWh meter with an S0 pulse output is a far more sensible way of obtaining that sub-1% accuracy.
@pirlouwi: According to the inverter specs the power factor during operation is 1; thus, the 59W should not occur during the day - at least for my SMA inverter...
Hi , I'm in the UK , and I've adjusted the default mains volts on my Flukso to my house actual , so changed from 230 volts to 242 volts. I'm certain this will give a more accurate to my SMA overnight reading. However i'm looking to change over to pulse output soon.
Go pulse. Only real solution for accuracy.