I work on the dryer every now and then, and at present I am working on the chimney and the ‘chimney pot’ which is design to extract air passively by the Bernoulli effect. I have also removed the light and will add a secondary fan instead. This fan will move air around inside the cabinet. I am also going to add a power switch and a dry switch (to initiate a drying ‘cycle’). This second fan will push air down from just under the roof and I hope this will mean the humid air distribution will mean that the sensors will see a humidity rise and then the computer can be programmed to handle this humidity and turn of when it drops, which hopefully means the clothes are dry. A secondary benefit of the internal fan is to stop the bad smell which gets into the clothes because of the lack of air movement. It turns out I think that just sucking out the humid air up the chimney does not work to dry clothes, air must be required to flow past the clothing.
I haven’t really done any work on the dryer at all, none of the software is done and all I did was test that the new dual channel humidity and temperature sensing would work. By the way to do the dual channel I had to use a quad analogue switch to switch the I2C since the ebay sensors I’ve got don’t allow a choice of the I2C address.
What I have been doing is designing a PCB for the controller part, and tonight I sent if of to Seeed Studio to get fabbed. This cost me just over $20 Australian, although I have to wait from between 15 – 30 days for the boards to arrive (I get 10 although I doubt I will ever use more than one). I am quite nervous that the fab will fail or that my design can’t be fabbed, or, of course that I haven’t transcribed the circuit from my hand wired version.
Kikad takes a little bit to get used to but it wasn’t that difficult to design this board; not that it’s a very big board. The last time I had a board I designed made I had to lay it out with black tape, I hardly recall, but it must have been painful. Although neither method is as painful as doing point to point wiring, even at this small scale. I’ve done a board with point to point soldered connections using wire wrap wire, where each end needs to be stripped, that had hundreds of connections. I can’t believe I did it now. That was the graphics terminal, I wish I still had it, I would frame it an put it on the wall.
I am not motivated at present to do much else at the moment even though I have had some good ideas about the software; once I can measure the ambient humidity then I think I can make the thing work more sensibly. I have again found that clothes that are hung on the line still dry faster and clothes dryed in the dryer sometimes have a faint smell about them. I think it depends on how much sun is available. I really need to design and build the new air extracting chimney hat and of course create at least version 2 of the software ( to control the fan by comparing inside and outside humidity ).
Just awaiting my board now.
I have had a long time interest in natural heating and cooling for houses. I hope to be able to have my own place one day, and I’d like this home to be passively heated and cooled, but that situation is out of my reach at present.
To partly satisfy my interest I been building a clothes dryer that is located outside and uses the wind and the sun, but protects the clothes from getting wet when it rains. My hope is that this dryer will be more effective that an ordinary clothes line by trapping infra red and capturing the wind more effectively. Other advantages are that there is no need of pegs, and UV light is blocked. I also intend to power the electronics (which includes the controller, fan and a door activated light) with a battery that is charged with a solar panel.
My current design is only partly successful and I am in the process of doing some improvements, however the first version had the following features that were my first attempt to meet my goals.
1. I used used poly carbonate sheeting which blocks UV but lets IR through.
2. I tried to make sure there are weather seals at every join so air is trapped
3. I made a chimney with hat to utilize the Bernoulli effect to provide a way to draw we air out
4. There is a control system that measures temperature and humidity so that a fan will turn on when the humidity rises above a set point.
5. The controller has Wifi enabling remote monitoring / graphing.
The first version was only partially successful. I do see a temperature and humidity rise above ambient when wet clothes are put into the cabinet. The fan does come on and the humidity then drops. Clothes do dry and come out warm – if there is enough sun to warm the system. The cabinet does protect dry clothes if it rains, and I don’t need to worry about forgetting to bring in the washing.
However for any reasonable amount of wet clothes the humidity overwhelms the fan and the clothes do not dry more quickly than clothes hung out on the ordinary clothes line. In addition it appears that the chimney does not draw very well (or at all). Additionally, measuring the internal humidity is not a good way of determining whether or not to turn on the fan since at night the humidity can often rise quite sharply (probably when the dew falls).
I am in the process of modifying the system to improve it:
1. I have put in a large fan with more airflow and made a proper seal for it.
2. I am modifying the controller by adding an extra sensor channel so I will be able to measure the ambient temperature and humidity so that I can control the fan when the internal humidity rises when the external does not.
3. I will add a detector that senses whether washing is hung inside the system.
4. I will modify the passive chimney extractor to move with the wind and have a double cone shaped wind collector to increase the chimney draw.
My next task is to finish the modifications to the controller software and then do more experiements in order to tune the way the fan works. Once this is done I will make the new chimney pot and again experiment to see how that affects the drying time.
Notes: the embedded controller is a Weimos D1 Mini.