Fork me on GitHub

Ventilomatic

home

Bill of Materials

13 Jun 2014

OK, let’s start with the hardware design.

The system will involve three sensor modules and one central controller module interconnected by the existing house LAN and some pre-existing X10 home automation gear.

To keep the designs as simple as possible, all three sensors will share these characteristics:

However, as this is being prototyped from supplies I already have, I’m going to need to build three heterogeneous sensors.

Note: As I don’t regularly assemble these things from memory, this list may be revised as I actually do the assembly work.

Sensor #1

This sensor gets first pick of the components, so it gets all the ones that are easiest to work with and/or least overkill spec-wise.

…plus standoffs to keep it from electrically contacting the floor since this is the sensor which will be used where condensation is a concern.

(Specifically, three 10mm hexagonal female-female standoffs to act as legs and one stacked pair of shorter male-female standoffs and a twist-tie for the remaining mounting hole since I lack a machine screw with a head small enough to fit nicely next to the SCL header on the Arduino revision 3 layout.)

Sensor #2

This sensor gets the remaining network-capable components.

The telephone wire is necessary for running the floor humidity and window light level sensors off the same node due to a shortage of Ethernet modules.

No standoffs are necessary for this sensor since it will be resting on an unfinished wooden surface.

Sensor #3

With no network adaptor modules left, this sensor connects to the central controller directly via the USB serial interface. Also, having run out of DHT11 sensors, this one uses a more precise but more expensive DHT22 instead.

This sensor node will have the additional task of monitoring whether the door needs to be manually opened to allow the fan on X10 channel 4 to have the desired effect.

While the power supply isn’t strictly necessary, the Raspberry Pi is known for having a tendency to corrupt SD cards when supplied with insufficent power and I don’t currently have the tools and skill to determine whether that’s a concern in this configuration.

No standoffs are necessary for this sensor since it will be resting on an unfinished wooden surface.

Central Controller

Note: Depending on how this project turns out, I may end up either using a CM17A with a USB-Serial adapter or just using my planned network-transparent architecture to pipe X10 control commands to my desktop PC which already has an X10 transceiver connected.

After all, I’ll probably be running a system node on my desktop anyway to log its internal thermal sensors and the CM17A is easier to work with on the software side… it just requires an RS-232 serial port and is very picky about which cheap Chinese USB-Serial adapters it will get along with.