Questions about constructing a switch

Discussion in 'The Projects Forum' started by someonesdad, Jul 7, 2009.

  1. someonesdad

    Thread Starter Senior Member

    Jul 7, 2009
    I'm building an automatic door opener for a chicken coop. I'm using a 3.6 VDC motor and planetary gear reducer from a Harbor Freight cordless screwdriver. A spool is turned by the screwdriver which winds up a cord, lifting the door. Closing just unwinds the thread. Opening the door requires between 3 and 3.5 A of current at 3.6 V, measured using a DC power supply.

    I'm using some Omron 1 A 125 VAC ceramic and stainless steel microswitches I had on hand as limit switches for door movement. While the power being switched is less than the 125 W as indicated by the rating, I worry about the switches' lives at 3.5 A. The door takes 3 seconds to open and less to close; there's one opening and one closing per day.

    This coop door has to reliably work in temperatures from -20 to 100 deg F for years; I'm keeping the manual system we now use in place because I know all mechanical and electrical things will eventually fail. Because I'm making this thing with parts from the junk box or that I can make in my shop, I'm using a transformer that steps the 120 VAC to about 10.7 VAC; then I'll get the required 3.6 VDC by using two LM317's in parallel. They'll have to dissipate about 20 W, so I've mounted them with heat sink grease on an aluminum heat sink extrusion that has room for two TO-3 devices. My intuition tells me three seconds of operation twice per day will be fine.

    However, I suspect I don't have enough heat sink to reliably remove 20 W of continuous power. Thus, I want another switch in series with the microswitches.

    Here's a description of what I'm planning on making. The door lifting cord passes through a 0.75" diameter hole in a 2x4. On the top of this 2x4, there's a 1.25" diameter counterbore that's 0.2" deep. I'm planning on putting a metal washer 0.1" thick in this counterbore (the washer will have a 0.75" inside diameter so as not to impede the cord). On top of this washer will sit a 1.25" diameter chunk of metal bar stock. The cord passes through this chunk of metal, but the hole through the chunk will only be about 0.1" in diameter (about twice the diameter of the cord). A small washer on the cord sitting on top of a knot in the cord will be perhaps an inch below the chunk when the door microswitch shuts the motor off. If the door microswitch fails, then the knot/washer will lift the chunk off the 1.25" diameter washer, breaking the circuit. The only ways I can see this second "high reliability" switch failing are: 1) corrosion reducing good contact, leading to the door not opening (this is a "safe" failure) or 2) the cord breaking. The second failure probably OK, since the no-load current of the motor is 1.5 A or so, meaning about 10 W needs to be dissipated -- and the heat sink may be able to handle that.

    I'll make the washer and chunk of metal from either stainless steel or brass (I prefer brass, as I have it on hand, but I'll have to go out and buy the stainless). If I use brass, I'll either have to gold plate it or resign myself to going out and removing the tarnish periodically (probably a couple of times per year). I'll machine the mating faces flat and smooth by using up to a 400 or 600 grit sandpaper.

    The current density for the washer/chunk at 3.5 A is 6.9 kA/m^2. This compares to 6 MA/m^2 for 20 A of current in a 12 gauge wire insulated with rubber insulation, a NEC value from my 1969 Handbook of Chemistry and Physics.

    Phew! Thanks for wading through the description -- I hope it's clear enough. My questions are:

    1. Do you see any problems with this design?

    2. This is potentially a "dry" type of contact, handling a low current density over a long period of time. Is there any possibility of the chunk and washer welding together so that the switch stalls the motor before opening? Remember, this thing will likely sit in one position for many years unless I deliberately go out and disturb it.

    3. If you could choose any material for the washer and the chunk, what would you pick and why?

    4. Are there better (easier, lower cost, more reliable, etc.) ways of doing this?

    Comment: this has been an interesting design problem. I've thought about it on and off for years and wanted something cheap and reliable. I considered things like garage door openers, filling a counterweight with an antifreeze/water mixture from the house, a custom-made hydraulic cylinder, etc. The best solution is not terribly obvious.

    By the way, here's the problem this solves. We usually have 10 to 15 chickens at any time that roam free on our property during the day. But we want them to be shut into the coop at night. If we forget to close the door, there's a good chance a raccoon or fox will get in and kill many chickens (foxes usually don't kill a bunch, but we've had raccoons wipe out 4 or 5 chickens at once). A big problem is that once this happens, those chickens will never go back in the coop to roost again. That means we have to be there or make arrangements with a neighbor to remember to close the door at night. If the door isn't opened during the day, the chickens can die on a hot day because there's no water in the coop. Thus, reliable operation is a priority.
  2. jpanhalt

    AAC Fanatic!

    Jan 18, 2008
    DC and AC ratings for switches are different. DC is usually quite a bit lower, as AC tends to self clean the contacts when making and breaking. You cannot rely on the wattage or VA.

    I am not sure you need the back-up switch. My fist impression was to pick copper for at least one of the contacts. Copper is high melting and does not self weld as easily as steel. However, since the switch is not really needed, maybe stainless steel or nickel would be a better choice. They won't corrode as much.

    Don't know. Never had an automatic door. The entire nesting coop and exercise area was enclosed.


    Senior Member

    Jul 1, 2008
    The general practice is to use a high current relay that's controlled by the low current switch.