meteohub.de

Hello,

i installed a separate power meter with S0 output, because the powermeter from the energy company does not support S0.
This additional power meter is connected with RFXmeter to Meteohub.

Now i have the problem, that the difference between the 1st powermeter, which is relevant for my energy bill, and the S0 powermeter is exactly 0,3 kwh per day, independently from the power consumption.

Is there any possibility to consider this difference in the statistics?
I haven´t found anything to configure in the sensor settings.
Thank you!

calibration at the bottom of the settings page will maybe do the trick, if I got it right, what you wrote

Hi wfpost,

the calibration will not work.
There it´s only possible to configure a fix offset, which is only considered once.
In my case, the offset of 0.3 kwh should be considered on a daily base.

No, you can provide an offset and a factor to meteohub calibration.

I still didn´t get it.
In the calibration section i can configure an offset, which is a fix value, and a factor (i.e. 0.001 for 1000 impuls for 1 kwh).

My problem is, that my power meter counts 0.3 kwh per day too less.

What do i need to configure?
Thx.

ah, I see. No idea. Hopefully Meteohub doesn't drop the 0.3 kwh.

gdfde wrote:Now i have the problem, that the difference between the 1st powermeter, which is relevant for my energy bill, and the S0 powermeter is exactly 0,3 kwh per day, independently from the power consumption.

I can't offer a solution, but can I ask for a little clarification in case that helps someone spot a solution.

Is it:
a) The extra meter (the SO one) disagrees with the energy company meter (and Meteohub is just out because it is showing what the extra meter shows).
or
b) The extra meter and the energy company meter agree, but the figure seen in Meteohub is out.

It´s case A...the extra meter disagrees with the energy company meter.

gdfde wrote:My problem is, that my power meter counts 0.3 kwh per day too less.

When measuring electrical power, the sensor reads the current, so what's actually measured is kw. When accumulated for one hour, the reading is kwh. If the offset is 0.3 kwh per day, you need to determine frequency of data samples during the day. Suppose the sample is once each minute, therefore 60 minutes per hour x 24 hours per day = 1440 samples per day. If the daily accumulated offset is 0.3 kwh, then the calibration factor is 0.3 / 1440 = 0.000208333 offset. I'm not sure what precision meteohub can use. This value would be a fixed offset for each sample.

If you cannot determine the number of samples per day, then the closest you can get is an approximation by determining what percentage 0.3 kwh error is in a *typical* day. Suppose that typically you use 31.7 kwh per day. Calibration factor would be 1 + (0.3 / 31.7) = 1.009463722. When consumption is higher than typical, your reading will be slightly lower. But on a typical day, it will be very close to actual.

If your consumption pattern is seasonal, like mine is, then you would need to recalculate the percentage factor periodically.

Hi Birdfeedr,

thank you for this excellent hint.
In my case i just need to find out the frequency of sending data from the RFXReceiver (on which the RFXmeter for the power consumption is connected) to Meteohub

There is only one very small disadvantage...in case, when there are some data samples lost, the offset will not be 100 % correct...but i think i can live with this.

gdfde wrote:In my case i just need to find out the frequency of sending data from the RFXReceiver (on which the RFXmeter for the power consumption is connected) to Meteohub

The RFXReceiver will pass data to Meteohub each time data is received from the RFXMeter.

The RFXMeter can be set to send data at various intervals. I think the default is 30 seconds? If you are not sure how yours is set then taking a look in Meteohub => Inspect Data is probably the simplest option.

skyewright wrote: The RFXReceiver will pass data to Meteohub each time data is received from the RFXMeter.

The RFXMeter can be set to send data at various intervals. I think the default is 30 seconds? If you are not sure how yours is set then taking a look in Meteohub => Inspect Data is probably the simplest option.

Well, i guess, you are right, the RFXmeter sends every 30 seconds.
When i check the raw data in Meteohub, the timestamp tells me, that data is received/recognized every 31 or 32 seconds.
So i have here the problem, that somehow the processing time of meteohub also needs to be considered...

The best thing would be, if RFXmeter would include a timestamp and Metehub would only consider this timestamp...

Do you have any idea on how i could set the offset somehow reliable?

20100810145310 data10 125244300
20100810145342 data10 125244600
20100810145414 data10 125244900
20100810145446 data10 125245200
20100810145518 data10 125245500
20100810145550 data10 125245800
20100810145622 data10 125246000
20100810145654 data10 125246200
20100810145726 data10 125246400
20100810145757 data10 125246700
20100810145829 data10 125246900
THx.

gdfde wrote:The best thing would be, if RFXmeter would include a timestamp and Metehub would only consider this timestamp...

(so far as I'm aware) that is simply not how Meteohub is designed to work, i.e. the timestamp refers to when Meteohub processed the data.

However I don't think that minor variations in the Meteohub time stamp are significant, or even minor variations in when RFXMeter sends. More significant is that the number sent over a substantial period (e.g. a day) remains reasonably consistent. You have 11 readings between 14:53:10 and 14:58:29, that's 319 seconds which suggests that the actual interval of sending is probably closer to 32s than spot on 30s. It would probably be an idea to use a bigger sample to check up on that.

At 32s there would be 2700 in 24 hours.

I think a problem you may have is that the calibration factor you want to apply may be closer to 1 (e.g. 1.000111111) than Meteohub allows for? When I look at the calibrations page here I only see 3 decimal places.

Hmm, a would be a bit careful, occasionally I'm missing the rfxmeter readings for some time, since the transmission is overlapping with other sensors on 433 Mhz. So adding a constant throughout the day might give some misreadings. The likelyhood depends, of course, a bit on the number of sensors you've got deployed.

I'm still wondering where the misreading comes from, I'm pretty confident that the RFX stuff is doing it's job, so my 1st step would be to consult whomever did provide the meter, since there is something funny going on.

YJB wrote:Hmm, a would be a bit careful, occasionally I'm missing the rfxmeter readings for some time, since the transmission is overlapping with other sensors on 433 Mhz.

Any particular ones? I ask as I'm awaiting delivery of an RFXMeter (plus 3 x RFXPulse) and I have a lot of OS 433 Mhz sensors...

YJB wrote:I'm still wondering where the misreading comes from

Me too. Solving that problem would be better than calibration factor that is at best likely to be an approximate fiddle factor.

gdfde: How are the 2 meters arranged? Is it possible that there is something between them that is consuming 12.5 watts (12.5 wats x 24h => 0.3 kWh). Presumably the meter must use at least some power itself, though 12.5 watts would seem unlikley (you could run a couple of ALIX-1D's on that!).