How to detect deadbolt?

T Chap

I want to make a way to remotely find out if the deadbolt was locked or not. On a pocket door, a deadbolt comes out of the door edge and has some flaps that go up and down inside a receiver box to secure the door. I want to be able to check if it was locked or not. Adding a magnet to the leading edge of the deadbolt for hall sensing would not look great and it might get knocked off.

https://www.parallax.com/product/555-27401

I thought about putting this inside the receiver box that sits in the jamb, so that if the deadbolt is pushed in, the sensor would detected it, however the problem is that the deadbolt with either be about .25" away if locked and 1.25" away if not locked. I am not sure how this sensor would work at such a small range of distance. I don't want to put a spring loaded switch if possible.

Any suggestions?

I tested a 38k modulated LED on one side of the box, and a 38k detector on the other side, and when the deadbolt is inserted it blocks the light. It adds some bulk on either side of the receiver box, and also over time may die and will be difficult to get access to it. Also, I am thinking the light may reflect and get around the deadbolt.

In the drawing, #20 is the receiver box that is a very thin metal, easily drilled.

tonyp12

Proximity Sensors LDC0851 Differentially Compensated and Highly Accurate Inductive Switch
http://www.mouser.com/ProductDetail/Texas-Instruments/LDC0851HDSGR/?qs=sGAEpiMZZMs3uAJYYmvlK0y75LftYGld%2baD8v0dfs43ogAWcyHIXSg==

Seairth

I wonder if you could just electrically isolate 19 and 20, then use the deadbolt "close" the switch.

T Chap

Tony, that looks like a great solution. Especially since it can be tuned with the reference metal plate. Thanks.

Seairth, the deadbolt does not really make contact with 19 and 20.

Chris Savage

I was just thinking why can't you mount a plunger switch into the back of the receiver box. A switch would be depressed when the door was locked.

jones

I like Tonyp12's idea, or if you don't want an active device inside the wall you might wind a coil of wire around the receiver box such that the deadbolt acts as a (presumably ferrous) core when it's locked. You should then be able to detect the change in coil inductance. The only thing in the wall would be the coil of wire, and since it wouldn't move or be driven with a lot of current it ought to be reliable.

tonyp12

Two ways to go about it.
1: Looking at the deadbolt head on, side by side coils with reference metal plate or stacked coils on 4 layer pcb board.

2: Make bolt slide sideways over two side-by-side coils (probably best if you want it inside door and not frame)
You tune the RED coils so it wins if both coils is seeing the dead bolt.
Green coil wins when 50% of the red coil is not seeing the bolt

garyg

Hi
I have a deadbolt viewing situation also.
I resolved it by setting up a camera to watch my front door.
I can see if it is locked and I can also see if the door is ajar.
This is a very different method than you are thinking about, but I
thought I'd throw this in there because it works for me.
I use Insteon hub and camera.

msrobots

I think a simple mechanical switch would do it, like Chris Savage wrote.

You said the deadbolt does not make contact with the receiver box. Change that. just cut the spring of a ball pen in half pull both longer, mount both insulated in the receiver and then let the deadbolt make a contact between both. Or any other form of simple switch.

Maybe a piece of soft foam and two contacts to be connected by the deadbolt?

Enjoy!

Mike

jmg

T Chap wrote: »

I want to make a way to remotely find out if the deadbolt was locked or not. On a pocket door, a deadbolt comes out of the door edge and has some flaps that go up and down inside a receiver box to secure the door.

So you want to sense the final locking state of 'flaps out', not just 'bolt in' ?

T Chap wrote: »

I tested a 38k modulated LED on one side of the box, and a 38k detector on the other side, and when the deadbolt is inserted it blocks the light. It adds some bulk on either side of the receiver box, and also over time may die and will be difficult to get access to it. Also, I am thinking the light may reflect and get around the deadbolt.

Sounds tough.
A small reflective sensor, something like
http://www.sharpsma.com/webfm_send/1267
with two holes in the back of the box that 'looked' at the flaps, may work.
You would adjust the LED drive & sync-detect, to get the levels right.

Are the flaps magnetic, or stainless ?

Phil Pilgrim (PhiPi)

Put a small spring between the door end and the frame. That way, unless the door is locked, the spring will create a gap between the door and the frame. This gap should be easier to detect than the deployment of the flaps.

-Phil

jmg

Phil Pilgrim (PhiPi) wrote: »

Put a small spring between the door end and the frame. That way, unless the door is locked, the spring will create a gap between the door and the frame. This gap should be easier to detect than the deployment of the flaps.

Sometimes these sensors are needed to prove a door is both locked and dead-bolted.
eg A common attack is simple clear tape over the door latch, so the door appears to be closed and locked, but actually is not.
The felons come back later, and simply walk in their pre-prepared door.

Actual sense of the dead-latch flaps, give better assurance.

Peter Jakacki

I've used a very simple technique for detecting objects, a simple capacitive plate fed from an pin through a resistor and a ground plate so that whatever material is in-between affects the rise time, a bit like RCTIME. I originally developed this type of sensor in a vending machine on a long PCB strip with large pads so that as cards were fed through I could detect the change in rise time and even tell how far through the card was, no moving parts and no coils etc, ultra simple. Of course this sensor had multiple columns and the other columns plus a fixed cap were used for reference but for the dead bolt even just two I/O and pads could provide a very easy to discriminate signal, one pin changes, the other fixed, then the other changes will the first remains stable etc.

T Chap

Great suggestions guys. Many thanks.

jmg wrote: »

So you want to sense the final locking state of 'flaps out', not just 'bolt in' ?

If the deadbolt is in, then the flaps must be out too, they pop out at the last bit of travel of the deadbolt. I can't put springs to hold the door open and there cannot be any additional force required to push the deadbolt into the receiver(springs, plungers etc). However, I like the idea of making the flaps become part of a switch. I can make my own receiver with isolated top and bottom parts that are spring loaded so that when the flaps extend up and down, they make contact with metal surfaces and complete a circuit.

On my next batch of PCB orders I will definitely add that circuit Tony mentioned, that would be very useful for other experiments.

idbruce

How about a reflective object sensor, similar to the now obsolete QRB1113 and QRB1114, mounted to the back of the receiver?