Shop OBEX P1 Docs P2 Docs Learn Events
How to detect this door latch? — Parallax Forums

How to detect this door latch?

T ChapT Chap Posts: 4,198
edited 2018-02-08 23:21 in General Discussion
I want to check if a patio door is locked or not remotely. I have a Prop that will check a switch/sensor and communication to the home automation system whether the lock is engaged or not. There are two latches that flip out and go into a receiver slot. I want to put a switch or sensor of some sort inside this plate to know when it is latched. I suppose the options are a plunger or lever switch, optical interruption(emitter below and detector above), proximity or inductive sensor. Optical interruption may be less reliable as the light can bounce around and still hit sensor. The sensor needs to have a long lifespan, the simpler and dumber the better.

I use this Avago Gesture sensor APDS-9960 which measures distance and outputs distance in a value in the Prop, so I can set a threshold to check if there is a reflection in less than 5":

https://www.sparkfun.com/products/12787

McMaster sells metalic sensors like this which I assume are inductive:

https://www.mcmaster.com/#73635k81/=1bhiahp

https://www.mcmaster.com/#7364k24/=1bhibzj

The photo I uploaded shows the latches in the Extended Lock Position.

I can remove the plate with two screws and add something inside the plate. Any ideas on a good way to do this?

Comments

  • Don MDon M Posts: 1,647
    edited 2018-02-08 23:26
    You could have your system send a text to the person inside and ask them to see if the door is locked... ;)
    Sorry I just had to say it.
  • How about a slotted optical switch?
  • kwinnkwinn Posts: 8,697
    An optical interrupter glued to the plate is probably the simplest and most reliable method. If reflected or ambient light leakage are a problem the plate and door area behind it can be painted black. The current to the led can also be made low enough that only a direct emitter to detector beam will activate it.
  • https://www.sparkfun.com/products/9299

    Here is one I have somewhere in my shop. This could easily be glued in.
  • An optical interrupter glued to the plate is probably the simplest and most reliable method.
    This could easily be glued in.

    Surely you are both joking. Get a slotted optical switch with the proper configuration for screwing it to the strike plate. Drill and tap the strike plate, and truly make it reliable.
  • Yeah, that has a nice wide throat. It shouldn't collide with the latch.

    -{hil
  • I'd use an analog hall sensor module. They are about $5.
    Then I'd mount it to the door frame, couple screws will do nicely.
    Then glue a neodymium magnet to the door face, even with the sensor.
    Close the door read the sensor, open it a crack and read the sensor again. That's your threshold.

    That's the simplest and least invasive.
  • Are you sure the latch is steel and not aluminum?

    -Phil
  • All good ideas. I would buy the same latch if it's cheap enough and tinker with it . Maybe you could get away with using a simple micro switch.
  • ercoerco Posts: 20,244
    edited 2018-02-12 03:03
    DigitalBob wrote: »
    Maybe you could get away with using a simple micro switch.

    +1. K.I.S.S.! Nothing to go wrong and good for a million cycles.

  • kwinnkwinn Posts: 8,697
    erco wrote: »
    DigitalBob wrote: »
    Maybe you could get away with using a simple micro switch.

    +1. K.I.S.S.! Nothing to go wrong and good for a million cycles.

    Simple, yes. Nothing to go wrong for a million cycles? Maybe....or maybe not. I replaced an awful lot of micro switches used in low voltage light switches for building automation systems every year. Far far more than should have been replaced if they were actually good for anywhere near a million cycles.
  • kwinn wrote: »
    erco wrote: »
    DigitalBob wrote: »
    Maybe you could get away with using a simple micro switch.

    +1. K.I.S.S.! Nothing to go wrong and good for a million cycles.

    Simple, yes. Nothing to go wrong for a million cycles? Maybe....or maybe not. I replaced an awful lot of micro switches used in low voltage light switches for building automation systems every year. Far far more than should have been replaced if they were actually good for anywhere near a million cycles.

    +1 give me something without moving parts any day of the week and twice on Saturday
  • I found a 14mm wide version of the photo interrupter like the sparkfun linked above. Easy enough to install and replace if needed. A switch is hand to install as there is very limited space. Thanks for the various suggestions.
  • T ChapT Chap Posts: 4,198
    edited 2018-02-12 16:43
    Thinking out loud.

    I have a similar application that will not allow for wires. I want to know if a latch is present but there can be no wires ran to the switch or sensor. My home alarm has door contact reed or hall sensors with a single cr2032 that has not dropped hardly any over 2 years since it was installed. It wirelessly talks to the home alarm lcd brain. My guess is a PIC with a very low clock speed watching the contact and it turns on full speed if at all just to send data to the main box. Anyone got an idea how to watch a contact and get years of life like this example? I have the Nordic NRF24L01+ working well for other applications on the prop but that board only gets a few days of life with 2 AA but at 80mhz. One option is use wires but run the wires to an area where there can be a through-air power transfer. This might even possibly be preferred over replacing batteries every few years. What to search for to find this type of power transfer to end up with 5VDC.
  • ercoerco Posts: 20,244
    kwinn wrote: »
    Nothing to go wrong for a million cycles? Maybe....or maybe not. I replaced an awful lot of micro switches used in low voltage light switches for building automation systems every year. Far far more than should have been replaced if they were actually good for anywhere near a million cycles.

    OK. At the extreme end of the Weibull probability curve, you get the Smile switch which fails at 1% of a million cycles. That's 10,000 cycles. Activated once a day, that will last over 27 years. I could live with that.

    Yikes. I could die with that.

  • T ChapT Chap Posts: 4,198
    edited 2018-02-12 17:21
    How long does an LED last if on 24/7? 7 years before 50% lumens? Optical has a life expectancy right?

    Newhaven Display LCD datasheet: 50,000 hours Backlight lifetime, rated as Hours until half-brightness, under normal operating conditions.

    An LED for a photo interrupt will have a life span too.
  • ercoerco Posts: 20,244
    Door sills & latches can get grungy after many years of dust, weather & spider webs. Hard to know for sure which would fare better, a mechanical switch or a photointerrupter. A sealed reed switch & magnet is a solid weatherproof/dirtproof combo (used on bicycle wheels/computers forever) if you can mount the magnet reliably on the latch.
  • I am trying to determine if a steel latch has flipped out from one door into the receptacle on another. No way to attach a magnet. This is not to see if a door is closed. A sealed reed and magnet is for door close detection, not good for latch detection.

    BTW my xfinity alarm front door reed/hall and magnet(whatever type it is) goes off randomly every day. It is mounted perfectly. Something is wrong they have to replace it.
  • ercoerco Posts: 20,244
    With a bit of drilling, you could countersink a tiny neodymium magnet in the lip of the latch, to trigger a reed switch.

    latchmagnet.png
    342 x 282 - 2K
  • kwinnkwinn Posts: 8,697
    erco wrote: »
    kwinn wrote: »
    Nothing to go wrong for a million cycles? Maybe....or maybe not. I replaced an awful lot of micro switches used in low voltage light switches for building automation systems every year. Far far more than should have been replaced if they were actually good for anywhere near a million cycles.

    OK. At the extreme end of the Weibull probability curve, you get the Smile switch which fails at 1% of a million cycles. That's 10,000 cycles. Activated once a day, that will last over 27 years. I could live with that.

    Yikes. I could die with that.

    I think the early failures of the light switches has more to do with with the use case rather than the quality/cycles of use of those switches. By that I mean that the lights tend to be turned on or off with a tap or slap rather than a more gentle press. The same type of closure action is probably going to happen with that style of latch.
  • I have some Magnasphere switches that would be perfect. They enclose a magnetic ball that connects the two pins until a piece of metal is placed near them to attract the magnetic ball the opposite direction. Super simple, very reliable. Let me know if you want one.


  • T ChapT Chap Posts: 4,198
    edited 2018-02-13 02:07
    WBA that is very cool. I'll get some off Digi right now and test it tomorrow. I believe that latches are steel, not aluminum... Should work. So you think this is more reliable long term 10+ years vs the Opto Interrupt method?


    Edit:
    The issue with this part looks like with low carbon steel you need a distance of around .1" to get it to switch. That's a tight tolerance in this case but I want to check them out for future reference anyway.
  • I just got in one of those magnasphere devices. Its very interesting. You need a magnet for a target to get any distance with it, otherwise with carbon steel it has to be almost touching it, around .1". It wont work in this application but I am glad to see this device for future use. There are cases where adding magnets aren't a good option and if you are needing to check a tight proximity with steel it may be a nice solution.
  • Wow, that was quick! Yes, they are slick devices if your project can support them. I have one mounted in a plastic enclosure for a client. The enclosure is part of the original equipment, but I have added in an extra controller for the device. Rather than drilling holes in the enclosure for a switch, I have one of these mounted in a space that is about 0.080" thick. To trip the override program that my controller performs, all he has to do is hold a piece of metal to that part of the enclosure. It works great perfect when he holds his car keys up against the side of the enclosure.

    In regards to your reliability question, yes, very reliable. I know of a product used by firefighters that utilizes these switches.
  • WBA,

    I have a sliding door with a similar latch and also wondered what kind of sensor I could use to make sure it was fully latched...

    Those are impressive - I briefly looked at the specs and thought the air gap distance was about 1 inch but a second look showed that was with NEO 35 magnet...

    With ferrous metal it is closer to .1" as T Chap found out. Like he said - might be handy in certain situations.

  • Ron, yeah, if you look at the datasheet, there are two banks of specs; one for a magnet target and one for a ferrous metal target. When I set up a test circuit with one on my PAB, I could get longer ranges with large magnets. In this case, I agree that these specs are a little tight at the 0.128" trip distance.
  • T ChapT Chap Posts: 4,198
    edited 2018-02-16 23:41
    Here are the two options, one is photo interrupt and the other is a lever switch. I made some little adjustable aluminum mounts for fine tuning position.
    640 x 480 - 113K
Sign In or Register to comment.