Migrated from here:
I’d really like if resist could include when the fan is running. While the fan may affect some areas of the house, the overall structure should be relatively unnafected by the fan running (since there are no btus being injected into the house)
We run our fan 247 to help a little with balance and filtration and to keep the lower level from getting stagnant.
If a gas furnace fan is running on auto for, say, 10 minutes an hour, ie 5 minutes every half hour, will this not prevent effective data collection for calculating the resist line? The fan has not heating effect, so why not ignore it? Thanks, Simon
If you run the fan for 5 minutes every half hour then there are still 25-minute ranges where resist samples are being gathered. This is plenty accurate, especially because there are a lot of resist samples in general so everything averages out nicely.
The original idea was that the fan could circulate air around the house and adversely affect the data, especially for multi-zone systems. If warmer upstairs air circulates around with cooler downstairs air, both profiles will change. Maybe that’s a desirable change, though?
I’ll take another look at this and see if allowing for fan usage has any negative effect.
I notice that my heat profiles have a lot of volatility behind them (before you fit a straight line through them) and am wondering whether this could be caused by solar radiant heating of the house on cold but sunny days here in Alaska. If this is the case, maybe it would be worth limiting sampling for these graphs to when the impact of solar radiation is low, ie between a few hours after sunset until sunrise or, if that’s too difficult, between 9pm and 6 am?
I actually removed daytime hours from the heat profile calculation a couple months ago but it ended up causing more problems than it solved by reducing the amount of data available and preventing many heat profiles from even generating.
Solar radiation is certainly a factor on the profiles. I am already removing statistical outliers, but if the data is generally noisy then there’s not much that can be done.
I may end up making this a setting so users can see if it would be helpful.