 |
 |
| View previous topic :: View next topic |
| Author |
Message |
MiniMe
Joined: 17 Nov 2009
Posts: 50
|
| One hour long outdoor temperature buffer |
 Posted: Mon Apr 15, 2013 3:02 pm |
 |
|
Hi!
Device is 18F4550 + ROM 24LC64
Temperature probe is Microchip's voltage out sensor interfaced by "olympic averaging" filter to eliminate signal noise.
What could be reasonably accurate (fixed time step) solution for this kind of buffer ?
Buffer means "getting 1 hour old data"
I'm not sure about rate of outdoor temperature collection. (i believe fixed time step could be much simpler) rate? ~ once per minute? OK?
Application in indoor open loop temperature control.
Should I get the data every second and use some kind of integrator filter to get more accurate "1 per minute" data?
How would you solve this problem? any particular examples?
Thank you in advance! |
|
 |
temtronic
Joined: 01 Jul 2010
Posts: 9241
Location: Greensville,Ontario
|
|
| Posted: Mon Apr 15, 2013 3:33 pm |
|
|
general comments..
1) what's 'inside' that you're keeping warm ? People or equipment or stored items?
It makes a big difference.People can sense 1*C temperature change, equipment not a real issue, and well 'storage' boxes don't care!
2) what's the heat loss of the building? A 'tight' building (well insulated, no busted windows) won't change much due to outside conditions, A 'lossy' building will need better control.
3) based on the past 25 years of doing remote energy control systems,you can easily get by with reading outdoor temperature about every 5 to 10 minutes, maybe 30 if a 'tight' building.
4) placement of the 'room' thermostat has a bering on the 'comfort' level experienced by the occupants and will override any outdoor sensor information.
5) using 'outdoor' sensors you can achieve about a 5-8% fuel savings.
6) placement of the 'outdoor' sensor is critical (can't be in the sun, in the wind, get wet, etc).
hth
jay |
|
|
asmboy
Joined: 20 Nov 2007
Posts: 2128
Location: albany ny
|
|
| Posted: Mon Apr 15, 2013 3:57 pm |
|
|
| Quote: |
Temperature probe is Microchip's voltage out sensor interfaced by "olympic averaging" |
Got a part number for the sensor in question? Curiosity in play here.
BTW: there may be TWO levels of averaging that you might consider, depending on the accuracy you seek and the short term stability of the sensor:
1- default "equal parts" olympic is great for a basic cluster of short timed ADC readings
but
2- where there is trending to consider in a high accuracy result,
then groups of "equal parts" individual time point reads may benefit by being combined in an unequal "biased" olympic fashion ....
Showing one posible example,trend positive, you take 16 readings, sort and throw away the top 2 and the bottom 4, then average the remaining 10.
(where EACH of these time points is a cluster of equal parts individual ADC reads equal averaged over a short amount of net sampling time).
This decision, would be taken based on how your sensor system responds to the direction of change.
For REALLY SLOW temp changes, ( on the order of a minute or more between resolved data points) as Temtronic says, none of this matters much, IMHO too.
Do you REALLY have so much noise that you need to go to this extreme? |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
Powered by phpBB © 2001, 2005 phpBB Group
|
FAQ 
Search
Register
Profile 
Log in to check your private messages 
Log in
