I've been threatening to do this for several months and now I'm getting started. I want to control my fridge during peak power usage. I live out in the country so I have separate fridge and freezer (two different appliances) The shortcomings are:
It will run the compressor during peak periods
The stupid defrost timer on the freezer sometimes kicks on during peak periods
Both appliances kick on at the same time
My peak demand period is from noon to 7PM so, the defroster firing during that period is just silly, but I can't change the darn thing. Setting them up such that they can't be on at the same time during peak period shouldn't cause any significant problems with temperature, and if it does, a backup temperature sensor that lets it run anyway shouldn't be too hard.
I already have an XBee network in the house and know how to work them so communication isn't a problem and I'm familiar with controlling power so I have no concerns there. The idea is to measure the power actually used by the appliance and transmit it over the XBee network as well as grab the time of day from the same network. I also have the total house power usage available on the XBee network so I can even shut them down during high loads if that becomes an option. This thing could also monitor the internal temp of the appliances and alert me if they get too warm.
Where I would like y'all's help is in choosing a couple of transformers. I will need two of them. One is to actually supply power to the devices I add, and the other is to supply a voltage reference for power measurement. They have to be separate devices because sucking current from the transformer affects the voltage measurement and closing a relay will do that (experience talking here). I'd like this thing to be small with a plug on one end for 110V (US standard) and a socket on the other to plug the appliance into. There should be a couple of relay outputs to hook the compressor and maybe something else into as well.
I just can't seem to get past the way they rate the transformers. I'm stumbling around through the various choices on digi and mouser and it's a bit overwhelming. The voltage sense transformer can be very low voltage and current since I'm only grabbing enough to measure the voltage, but the power supply one should be big enough to power an Ardweeny (the processor board I want to use for this) and a couple of relays with a low power XBee module for comm.
I know there are tons of people on this board that can solve this kind of problem in seconds.
I once had the thought to freeze my freezer to -24C during offpeak hours and completely switch it completely off during peak hours. In a test with a time-switch(?) the temp did increase to -17C during the day. Very acceptable, however getting to -24C took quite some energy I recall...
I would use an Arduino with timer and a relay that just prevents the defrost sensor from giving a signal during daytime (ok 12:00-19:00)
Alternating the freezer and the fridge can easily be done with the help of two temp sensors, different priority can be included. A simple state machine with one state var that shows which motors are on. Some pseudocode to get the idea
// state = { none, fridge, freezer, both }
#define FREEZER -15
#define FRIDGE 4
void loop()
{
a = getTempFridge();
b = getTempFreezer()
newstate = none;
if ( peakhour() ) // function to be elaborated.
{
if (a > FREEZER) newstate = freezer; // freezer has priority over fridge!
else if (b > FRIDGE) newstate = fridge;
}
else
{
if (a > FREEZER) newstate = freezer;
if (b > FRIDGE) newstate = fridge;
if (a > FREEZER && b > FRIDGE ) newstate = both;
}
if (state != newstate)
{
state = newstate;
if (state == freezer || state == both) SwitchOn(freezer);
else SwitchOff(freezer);
if (state == fridge || state == both) SwitchOn(fridge);
else SwitchOff(fridge);
}
}
Thanks for the snippet, I didn't think of using state variables. And, I was going to try exactly what you described for the freezer, glad you did it for me. It never occurred to me that it would use more energy to suck the temperature down, however it did occur to me that opening the freezer during the peak period (to get a frozen fudge bar) could blow the entire experiment. I think a little ice ballast might be a good idea though.
Any ideas on which transformer I could use for powering the devices? I have limited wall plugs behind the appliances.
If efficiency is the goal, make sure you have as much mass inside the fridge and freezer sections as possible. A full fridge/freezer is typically a more efficient fridge/freezer. Having the extra mass, once cooled, will make it far less susceptible to large temperature swings from opening and closing the doors and removing items like your fudge bar. And when you do experience swings, it will recover much faster.
Also, make sure all your coils are cleans. Clean coils are efficient coils. If you don't know, on most fridges, the coils are accessible from the grid-like panel on the front-bottom of the unit. Usually you can also access them from the rear but its typically not as convenient and use of a screw driver is frequently required. Your coils, just like on your AC, should be cleaned no less than once per year.
And if you have an ice maker and still have copper tubing, go ahead and spend the couple of bucks on a modern hose replacement kit. Compared to new walls, flooring, and mold, its an extremely cheap investment. There's a good reason people have moved away from that copper tubing. And of course, be super careful when moving a fridge if it does still have those pesky copper tubes.