I have configured a new Flukso with current clamps. I have noticed that there appears to be a minimum value reported once the current flows, of around 400 Watts. So, the graphs jump from 0 to 400 and never show values less than 400. Is there some sort of minimum startup voltage on the A/D converters?
The current clamp, which provides 0 to 5V proportional to the current measured, is attached to a voltage divider internally (impedance of clamp via 1k91) the FLM. This divider "maps" the 0 to 5V to the Vref range, which is 1V1 for the full 10 bit accuracy - so basically the accuracy is clamp size divided by 1024. You may calculate the expected accuracy for yourself, as you lacked to denote your clamp's size...
HTH - the circuit is part of the source repository; the accuracy of the ADC is part of the used uC's spec...
Please note: Make sure that the clamp is tightly locked around the mains wire - if that is not given you may experience rather weird readings...
Thanks for that information. There is something screwy going on then. These are 50 A clamps, so that should give 50 mA resolution, or 12W accuracy. Two of the circuits (solar generation and demand) both seem to jump from 0 to around 400W, and then rise above that.
As for locking the clamp tightly, I assume you mean making sure the clamp is closed (with a click)? because the cable itself is way smaller than the opening, and is not by any means tightly held by the clamp.
I am going to try an calibrate this against a known load and see what reading I get … any other suggestions are welcome.
This reminds me of a wire not sitting tight at the Flukso connector side because wires can slip behind the small empty space just behind the screw. You then tighten the screw but there is no wire end underneath it, the wire end just sits - intermittently making contact - behind the screw. Check and double check ...
Last week, I also noticed these screws have a life on their own. They seem to come loose after a while. So, it's a good idea to check once or twice a year and give them some extra twists.
The "life on their own" is related to the tinned ends of the wires; by tension the tin "melts", thus loses its friction for the screw becoming loose. For that purpose any electrical wire should have end sleeves instead of tinned ends...
The connections are ok as far as I can tell.
The Hot Water service is currently on and should be drawing 1800 W (according to the specs). I measure 165 mV across the clamp. If I understand your numbers above, that should actually deliver 1.1 V * 1800W/(50*240) = 165 mV? So that's spot on!
I need to test the same thing at the low end …. because both the house demand and the solar production seem to only fall to 33mV, after which they then drop to 0.
Ok, a little more homework. I have put a known 72W globe and measured the current. Interesting. The voltage shown on the clamp is 6 mV, which seems right. The Flukso shows this as 42W only.
I also have a clamp on the whole house, so turning the lamp off and on adds to the base load of the whole house. When the lamp is on, it jumps by 6mV as well, and the Flukso shows an additional 65W. Close enough for me. .
In other words, there is some strange non-linearity down the bottom end of the D/A.
Thoughts?
There's also the sample-rate and averaging to take into account: with a calculated accuracy of around 12 Watt (for a 50A clamp at around 240V), I can see fluctuations sometimes as small as 1 or 2 Watt, so clearly there is either a lot of fluctuation (or noise) going into the final 'measurement'
I have been going around in circles, but learned a little more today. I think it is a co-incidence that the minimum solar and demand figures are similar. I have found most of the demand load … 200W of it is having our air conditioner on standby! Looking at the net a bit suggests this is not a reliable measure.
Still need to identify why the solar power figure never goes below 300W, but I am guessing it is also not reliable?
Davida
It has been reported afew times on here that if you want really accurate power readings you have to install din rail pulse meters. I and lots of others have.
Clamps are still fine though to use as an averaging and performance measure.
I fitted 2 off these, House consumption and Solar input.
example:
http://www.schnap.com.au/kilowatt-hour-meter/single-phase-electric-kilow...
Yes, thanks Bazzle. I have read this. I didn't need really accurate but some of the readings are way off. and I guess I really needed to prove it to myself ;-)
http://en.wikipedia.org/wiki/Power_factor , http://en.wikipedia.org/wiki/Current_transformer and http://en.wikipedia.org/wiki/AC_power#Reactive_power; quote: "any phase shift in the measured current can affect the accuracy of the target measurement" - this is the trade-off between measuring principle (clamps don't need the circuit to be split up) and the accuracy...