Mike, I like reliability for my house for an important reason. Whenever it breaks, my wife points out to me that if we had bough a regular system this wouldn't be a problem. :(
I tried X10 for some other stuff, and think is just a terrible solution for almost anything. The X10 lighting is the thing my wife hates the most about the house. I have globe valves and dampers for all my zones, so on/off settings don't work. I have used nudam analog out modules on 2 RS485 buses to drive the valve and damper actuators. I used nudam relay modules for contact points for both power and low voltage, nothing I need is over a couple amps. There is one zone per room, so everything is a bit of overkill. The zone dampers do dual duty for A/C and ventilation and there is continuous HEPA filtration for all zones. The boiler takes an analog in to set the desired water temperature and the 2 A/C zones take low voltage dry contacts to start the compressors (replaces a thermostat). My plan is to read the sensors (one process per bus) every 30 seconds. I check for failed reads then run the readings through a median filter to toss any outliers. Cooling loop runs every 1 minute and heating every 2.5 minutes (much longer zone damping time for heat.) There are hold on/hold off timer for the AC and the boiler to prevent fast cycling. The damper actuators have a minimum voltage change limit, so to make small changes I need to make a bigger change and then back. The biggest control problem I have is that I live in California and the lowest heat output of my variable heat boiler (15kBTU/H) is often too much and the boiler has to cycle itself when it overshoots the target temp. I don't have to control that, but it does mean more wear and tear on the boiler than I would like. I played a havc system designer in a prior life. jerry BTW, if someone is looking for a project to do something interesting, I have one related to my heating system. Water flow sensors are either hard to get accurate readings, expensive or both. I have looked into it, and it seems like one could build an ultrasonic flow detector that straps onto a PEX plastic pipe and gives a velocity readout. It turns out what makes regular ultrasonic flow sensors is that the coupling of the transducer energy to metal pipes makes for expensive transducers and electronics. PEX has a sound velocity fairly close to water, so it should be possible to couple a simple depth-finder transducer to the PEX with sufficient energy transmission for good readings. I have had this on my list for over 5 years, and realize I will never get to it. So if anyone wants to run with this, I am happy to pass on what I know and people who might be willing to test the systems. I would be happy to test units as well. The goal would be a setup with a piece of PEX with the sensors mounted and a box that can read 1 or more sensors. Price target would be $50 parts and assembly for the sensor side. This would ultimately be something that someone like Danfoss would be interested in. Why you need this is if you want to measure (sell) the heating within a zoned building with hydronic heating, you need to know the heat differential from supply to return and the flow rate for each zone. You probably want ~2-3% accuracy on the flow plus ~1% on the temperatures to get a 5% system accuracy. On 12/13/2011 01:59 AM, Mick Sulley wrote: > Hi Jerry, > > Your system sounds interesting. I also have a control system based upon > 1-wire and Python which controls a solar (hot water) heating system > heating the domestic hot water and a swimming pool. My plan is to > extend it to control the whole of the central heating system as well. > > Reliability is key to this sort of project and each time I have had a > failure (there have been a few!) I have tried to identify the root cause > and prevent it happening again, but this is an ongoing task. > > I use a mix of home brew outputs based on DS2406, and X-10 to drive > valves and pumps. X-10 seemed the easy way to go but has not proved to > be very reliable and so I am moving to the DS2406 solution. What do you > use for driving outputs? It seems to be quite difficult to get > something that is fail safe, i.e. fails to off rather than on. > > In terms of timing, my system reads all the temperatures, currently 26 > sensors, then sets the outputs according to the logic and logs all the > data to a MySQL database. This takes around 20 seconds and it then > loops round again. > > User interaction is by a local web page in PHP which reads and writes to > the database. > > It runs on a server built from an old PC, located in the loft. I have > also built a second similar server, so that I can easily swap the plugs > over if the first server fails. > > I have notices that many people use hubs to split their 1-wire networks, > currently mine is a single network with everything on it. I would be > interested to hear views on the use of hubs, do they improve reliability > or speed? Are there other advantages? > > Cheers > Mick ------------------------------------------------------------------------------ Systems Optimization Self Assessment Improve efficiency and utilization of IT resources. Drive out cost and improve service delivery. Take 5 minutes to use this Systems Optimization Self Assessment. http://www.accelacomm.com/jaw/sdnl/114/51450054/ _______________________________________________ Owfs-developers mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/owfs-developers
