Thanks for your support!
To your questions:
Piwoslaw, Fornax is right; the moist hot air cools down in the HX, reaches 100% RH, and condensates. You can tell some water is removed from the air as the dewpoint (which is more of a measure of absolute humidity, keeping pressure constant), drops significantly. But then again this dewpoint also changes for the incoming air, which i cannot explain... apart from the fact these sensors are quite cheap and not precise. Or there is some air mixing. As I wrote before, the entry and exit of air outside my house are not separated well yet. In any case: there is quite a bit of water coming out of that HX!
The appartment is about 180m3 indeed. The replacement rate would be way to low if 1/ there would be no leakage and 2/ would the appartement be occupied for a larger part of the day. Now it is only moderately too low, I guess. But before there was no "real" ventilation at all, save from opening a window.
On the arduino, I am not sure how much details you would like. Also, I have no background whatsoever in electronics beyond what I learned the hard way, so i am not sure I'm the one to listen to on this. But anyway:
The DHT22 temperature sensors can be setup as described here (times 4)
learn.adafruit.com/dht/connecting-to-a-dhtxx-sensor
The fans get their main power from a 12V, 1A rated power supply, which also feeds the arduino.
The fans are 4-wire, the system is described here:
formfactors.org/developer%5Cspecs%5C4_Wire_PWM_Spec.pdf
I bought these el-cheapo ones (not so sure why) sharkoon.com/?q=en/content/silent-eagle-se
For a replacement (if ever), I was thinking of some sanyo-denki fans, possibly water-proof ones. I like these axial fans. Somewhat more powerful ones to push the air through some filter (I already notice there is a lot of dirt going through those tubes, good thing they can easily be cleaned or (for the interior one) replaced).
What is great about that 4 wire system is that one of the wires gives back a "tacho" signal (see point 2.1.3 in the document), and one cable is used to control the speed of the fan (point 2.1.4).
To connect the tacho signal, I connected that to one of the digital inputs, and then (using a resistor!) to the 5V supply on the arduino. As the document states, the fan will connect the 5V to ground once per half revolution (open collector), the resistor makes sure only a small current can flow; the arduino will detect the voltage drop. I use the "pulseIn" command to count the pulses.
For the pwm, that works in the opposite way: the fan provides 5V, this needs to be connected to a digital input (using a resistor!), which will connect this to ground at a certain frequency. To get that frequency, a small hack is required. IIRC I run mine at 32khz or so, out of the specification range, but it works just fine.
This all seems quite a bit simpler compared to soldering up everything yourself to do the pwm-ing to the main power of the fans, at least to me.
To talk to my router, I use this shield:
arduino.cc/en/Main/ArduinoWiFiShield
with this module
rovingnetworks.com/products/RN171XV
An alternative (perhaps cheaper?) would be sparkfun.com/products/9954
This wifi part was actually the hardest to setup. Its not fail-proof yet, if the connection is lost, the arduino freezes. I have to add some code to handle that (possibly with a watchdog timer resetting the arduino in the worst case).
Am i forgetting something? not sure how much detail anyone here is actually interested in, so just ask. I seem to be writing such long contributions it feels like therapy somehow.
I've attached a picture of the arduino setup before it was completely finised and disappeared into my wall (that was back in may before i found this job in spain).
|