Page 1 of 1
Format of rawdata
Posted: Sun Nov 21, 2010 2:22 am
by gdfde
Hello,
i have a power meter with RFXCom connected (1000 imp = 1 kwh).
Would it be possible to make the resolution of the data in meteohub bigger?
Currently acc. to doc:
Data data# system data [1/100 value]
It would be really great to implement 1/1000 values instead of 1/100.
At least in the sensor configuration i am able to configure this, but it´s not considered (only 1/100).
Re: Format of rawdata
Posted: Sun Nov 21, 2010 1:55 pm
by skyewright
gdfde wrote:i have a power meter with RFXCom connected (1000 imp = 1 kwh).
It would be really great to implement 1/1000 values instead of 1/100.
At least in the sensor configuration i am able to configure this, but it´s not considered (only 1/100).
What advantage do you see in 1/1000?
The sensor (I have one too) is only reporting integer values (watts), so neither of the existing decimal places are used. I'm puzzled why would you want a third unused decimal?
Re: Format of rawdata
Posted: Sun Nov 21, 2010 10:11 pm
by gdfde
skyewright wrote:
What advantage do you see in 1/1000?
The sensor (I have one too) is only reporting integer values (watts), so neither of the existing decimal places are used. I'm puzzled why would you want a third unused decimal?
The sensor is sending the impulses to Rfxmeter, 1000 for 1 kwh.
With the current implementation the smallest resolution is 0,01 kw within a certain timeframe.
I want to monitor the power consumption on a 5 minutes base (or even less) to create a kind of realtime monitoring.
If i calculate this value with 12 (for kwh) the smallest possible power consumption value is 120 Watt.
1/1000 would make the "realtime" statistics more accurate...it´s also possible to configure this in the sensor settings in metehub, but it´s not considered.
Re: Format of rawdata
Posted: Mon Nov 22, 2010 12:37 am
by admin
Why don't you go with storing Watts instead of kW? Then precision should be fine.
Re: Format of rawdata
Posted: Mon Nov 22, 2010 12:18 pm
by gdfde
admin wrote:Why don't you go with storing Watts instead of kW? Then precision should be fine.
This would not change anything with the granularity.
Meters with S0 output only support sending pulses (i.e. to pulsemeter of RFXmeter and their pulse module).
With the current implementation Meteohub only considers 100 pulses for 1 kwh, even if i configure in the sensor settings 0,001.
Re: Format of rawdata
Posted: Mon Nov 22, 2010 12:53 pm
by skyewright
gdfde wrote:admin wrote:Why don't you go with storing Watts instead of kW? Then precision should be fine.
With the current implementation Meteohub only considers 100 pulses for 1 kwh, even if i configure in the sensor settings 0,001.
I have this same equipment (i.e. RFXMeter & RFXPulse), and I record meter usage to the level of 0.001 kWh just as you require.
The Meteohub stores the data Wh (i.e. each LED pulse adds 1 to the count stored in Meteohub). If I want kWh all I need to do is divide by 1000 where I want to show the value as kWh.
e.g. The sensors page currently shows a value of 419963.00 for my import meter, so that is 419963 Wh, or 419.963 kWh.
To plot a graph showing consumption in watts with 5 minute resolution I use Meteoplug (which has very nice graphing facilities) with a chart definition sensor line like:
Code: Select all
sid4=data37 sfactor4=12 scolor4=FF00FF sname4=Import ssel4=suminc sfill4=50
i.e. I chart the increment in the counter over a 5 minute period multiplied by 12.
The axis on that chart is in W, but that's fine for me in this case.
Re: Format of rawdata
Posted: Mon Nov 22, 2010 2:35 pm
by gdfde
skyewright wrote:
The Meteohub stores the data Wh (i.e. each LED puklse adds 1 to the count stoerd in Meteohub). If I want kWh all I need to do is divide by 1000 where I want to show the vale as kWh.
e.g. The sensors page currently shows a value of 419963.00 for my import meter, so that is 419963 Wh, or 419.963 kWh.
Well, thats true.
But where can you configure to divide the Wh by 1000 in Meteohub?
]
Re: Format of rawdata
Posted: Mon Nov 22, 2010 10:22 pm
by gdfde
i found the solution.
i configure the pulse meter data with original data and created a virtual sensor (which must have come with one of the latest software updates) with the necessary granularity.
thank you for bringing me on the right track
