Tomasz Korwel
programmer, administrator, engineer - my everyday fights with reality

One of the nicest features of our new HVAC zone system is module called balanced controller. Basically it’s an algorithm calculating heat demand in every zone and opening and closing registers accordingly.

Here you can see that one of the zones (middle one) was fully open while remaining two were just barely moved, Then when this zone was satisfied in it’s heat demand it was closed and the first one was fully open to direct all hot heat into this zone to shorten furnace running time to the minimum.

I don’t have any statistic data to prove my theory yet but it seems to me that such behavior shortens greatly overall running time and simply saves us money.

Looking at the graph (and trying to keep my fingers from freezing):

I noticed interesting thing - the temperatures in my office tend to go down a bit every day (look at those last four green peaks). Keep in mind that my thermostat is set to the same temperature as it was before, yet those temps really go down. Looking for explanation I tried to remember what changes in weather did we observed during those last 4 days. Did it become colder outside? Not really. Maybe hotter? Naa… Rained? Snowed? Was it windy? And then it strike me - we got more and more sun each day. So apparently upper level of our house acts as greenhouse absorbing energy from sun. And because the thermostat is in fact mounted there, lover level rooms, not exposed to radiation heat became colder and colder.

One additional reason why giving those rooms more decision power as when to turn the furnace on is not that bad idea.

I installed 2 motorized registers and set the system to run in passive mode.

Here you can see:

how it nicely flattened the temp curve by closing the registers when temperature reached setpoint.

Time to finish the rest of the house.

A comment posted under one of my older posts reminded me that it can be a good idea to resurrect and finish my HVAC zoning project I shelved some time ago. Main cause I did so was that I couldn’t get dz compiled. I didn’t have time to mess with it any longer then, but I decided to give it a try this time. It turned out that over last two years Ubuntu project went forward so much that this time dz compiled successfully the very first time I tried. So right now I have a spider web construction on my desk consisting of two 1-wire temp sensors and two servos run by phidget boards. And they work. Servos are opening and closing according to changes in the temperature of my office. They have no influence as they aren’t running registers yet, but they proved to work. Now it’s time to run wires around the house and motorize registers.

Stay tuned.

Last fall while replacing our water heater the technician made a quick look and said that our furnace is going to need some attention in the near future because of cracked igniter. It was still working then but he said it may need replacement any time. Of course I didn’t think about it much then but, as Mr. Murphy says, things always go wrong when you least expect them. So our furnace decided to start dying when we needed to reheat our house after last week ice storm power blackout. This time I didn’t wait long and ordered needed parts from Arnold Service. As quick as small business can be - they were sent immediately and arrived in two days.

Below you can see “before” state - flame sensor on the left and igniter on the right.

It’s not that hard to see that igniter is not looking ok, although it’s much harder to recognize that flame sensor is not good either - it’s loose and not always correctly sensing flame which results in unnecessary furnace cycling.

The hole operation was quick and easy. First I replaced the flame sensor - disconnect one cable and unbolt one screw (both visible on above photo). When I took it out I could see that the ceramic holder is cracked and the sensor’s rod is free to move about 60 degrees resulting in incorrect readings.

A little harder task was waiting for me in with the igniter - first I had to disconnect the odd plastic connector, then two bolts.

After that I had to transfer metal bracket to the new one and carefully place it there without cracking the new one hitting other parts of furnace. It took me… well 30 seconds?

Below new and old one together and the new one installed in the furnace.

Some additional cleaning and our furnace works like new.

Technorati Tags: furnace, cracked igniter, replacement, flame sensor, cleaning