I installed an Ecolink door sensor on my wrought iron gate. The magnet is on the gate and the sensor is on the 1.5" square post. It worked for about two weeks and then stopped working. I went out to troubleshoot it and found that when I remove it from the gate, and move the magnet over it several times, it begins to work again. I can “open/close” the sensor without any issues at all. The moment I put it back up against the gate, it stops working and it remains in whatever state it was at when I placed it against the metal. I have tried so many variations to this process that I am convinced it is the gate metal that causing the problem. The weird thing is that it worked perfectly for the first couple of weeks.
I have tried taking a 1/4" piece of wood and using it as a backing for the sensor to keep it away from the metal and that is not working either.
Can anyone think of a way to solve this or is there a contact sensor that doesn’t rely on magnets to determine its state?
Yeah, that can be a problem with magnetic sensors. Commercial buildings with steel doors run into this issue all the time. Usually the magnet part ends up magnetizing some part of the metal and this confuses the sensor part. Remember those old school experiments where you could extend a magnet’s field with bits of metal? This process takes time, which is why there’s a delay to failure.
Anyway, there are a couple of things you can try.
first, make sure the sensor does still work. Sounds like you did that.
if this is a reed sensor with a magnet in only one piece, that’s the piece you have to worry about keeping away from the gate/door metal. If the post is wood, sometimes just installing with the sensor on the gate and the magnet on the post avoids the problem. But once the problem has happened, it takes time to go away.
usually the answer is a thin piece of plastic (not wood) between the magnet piece and the metal of the door/gate. So you had the right idea, just not the ideal material.
it takes time for the problem to occur and it takes time for it to go away. Unfortunately once the other metal has become magnetized you can’t just get a new sensor or sensor position or barrier and start over-- you have to wait for the charge to fade in the gate. So you may just have to remove the magnet for about two weeks, then start again. And be aware that it might take a week or two for the problem to come back, making testing tedious.
there are door sensors made for steel doors that use stiffer reeds and stronger magnets but they cost a lot more and you may have to move the antenna to a third piece so gets complicated. The barrier method is better.
often weatherproofing the gate with a plastic seal will help a lot. It depends whether you like those aesthetics. But there’s no question that flakes of rust on a wrought iron gate will make the problem worse.
Anyway, it sounds like you had the right idea, you may just need to give the gate metal time to demagnetize. If you don’t want to wait, lifting both sensor and magnet piece up off the gate surface about 4" with temporary blocks will likely work, but most people don’t like the aesthetics. Might be ok for a temporary solution. Then in a couple weeks, replace the blocks with a thin plastic sheet.
Here’s an article that mentions the issue.
Yeah any wireless technology and large amounts of metal is not a good combination and magnets will only make it more interesting. The 1/4" wood or plastic buffer would of been my plan of attack as well. I can think of a few possibilities, but ALL are long shots. Your Ecolink door sensor should have contacts for wiring. If this is the case here are some options
- Order an external magnetic sensor and wire it up to the contacts, give yourself at least 1ft from the magnets and the sensor.
- Use some left over wire and wire up the gate so that when the gate is closed the circuit is complete. i.e. the 2 wires touch either directly or through the metal latch.
- Order an external pressure sensor and wire it up to the contacts. Place the pressure sensor in a place so that when the gate is closed pressure is applied and the sensor knows the gate is closed.
@JDRoberts do you think a good Silicon calk might work in this case to provide insulation from the magnets to the metal?
Good idea. I think this will work well. I can place the sensor on the brick away from the gate and use the contact sensors. I just need to find an external reed switch or wire one up myself.
Sure, anything nonconducting that moves the magnet off the metal.
The problem with this issue is it takes time for the gate metal to get magnetized, so it can be a pain to test. Most field engineers use plastic sheets since you may have to remove them several times to get the right thickness.
As you mentioned, there are potential multiple problems with a sensor and a metal gate. If the problem is the signal not traveling through the door/gate, then moving the antenna piece further away from the gate can help a lot. Same thing if you switch to a superstrong magnet and stiff reed and the magnet is interfering with the antenna.
So in one of those cases, you’re moving the antenna away from the metal of the gate, and in the other you moving the antenna away from the magnet.
However, both of those are problems that you would see on day one, they don’t build up over time.
When you have a metal door/gate with a sensor that works fine for a few days and then the sensor start failing, it’s more commonly that part of the gate itself is becoming magnetized over time, and this is what is confusing the sensor. So you can leave the antenna where it is, but you have to get the magnet up off the metal of the door/gate so that it won’t magnetize it.
That’s why sometimes just swapping the position of the magnet piece and the sensor piece is enough to avoid the problem.
But once the magnetization has occurred, it takes time for it to go away again. So you may solve the problem by moving the magnet, but it will take a couple of weeks before the sensor is reliable again.
I pulled out an old M3 Tac-Light reed switch and spliced it into the unit. I put the sensor on the wood portion of the gate and the reed switch in between the gate and the post so when it closes it presses on the switch. This solution is working well so far. Also on a side note, I sprayed the whole thing down with Never-Wet before I put it outside. It has rained a few times since and I haven’t had any issues with the electronics getting messed up.
I’m looking for open/close sensing for a metal storm door and wondering if you could make recommendations: (easiest, cheapest but reliable):
External contact sensor connected to Ecolink or GoControl contact sensor
Plastic buffer between metal and smart contact sensor
Seems like the external sensor would be more reliable, but smart contact sensor is easiest and cheapest. Which would you do?
Any sensor that uses a magnet will have the same issue, it’s not because of SmartThings. Because magnets eventually magnetize whatever ferrous metal they are attached to.
If the storm door is aluminum, you won’t have a problem of the type in this thread, it’s just a question of getting signal through.
If the storm door will hold a magnet, then any magnet sensor can cause the problem. So you need the plastic buffer.
So start by just taking a kitchen magnet and seeing if it will stick to the storm door. If it will, then you need to lift any sense of that uses a magnet up off the door or it will stop working eventually.
The storm door must be aluminum because the magnet didn’t stick. You brought up the issue of signal so my other question is whether there’ll be another issue with signal getting through my steel front door into my home?
Yes, almost certainly the steel door will block signal. Is there a window nearby? Or does the steel door have a rubber weatherstrip or a wood frame? Is the wall that the door is set into brick or wood?
Yes to all of the above. There’s two windows side by side two feet to the left of the door. There’s a rubber gasket on all around the door frame. The door frame is installed into a wooden door frame.
Then you’ll probably be fine. But you’ll just have to test it to see.
It’s not going to cause any harm to the contact sensor? I’ve seen the stories warning against using contact sensors with metal and thought for some reason it’s because it causes them harm, but from what your saying it seems like it’s just because of the interference issues, correct?