Electric dead bolt design

[LiveWire]

I was at Menards hardware store the other day and saw electronic dead bolts. Upon close inspection they were using a small motor and a worm gear transmission to throw out the bolt and bring it back.

My question: How do they design it so the worm gear cuts off after the bolt extends? Then how does it reverse direction in order to pull the bolt back?

Thanks,

 

[zkt_PiratesDen]

never seen one
but a dc motor is reversable and switches could limit the bolt travel

 

[zevon8]

One way is to measure the current of the motor. When the motor stalls at end of travel, current rises dramatically, so you sense it then shut it off. There is likely a small micro that stores the keypad code, so it could also control the motor. It would be easy for it to operate the current sense, and when powering up the motor it checks for stall current, if the motor is stalled, try running it in the other direction. If both ways show a stall, then stop, theres a problem. You could also use limit switches, but current sensing is easier and one 2 less mechanical things to go wrong.

 

[LiveWire]

Thanks for your responses,

I figured they probably used either limit switches or some other electronic sensor, but how is this physically done?

In other words, how is this mechancially set up? How do you design the gears to push and pull the bolt in and out?

And how is a limit switch physically placed to limit the bolts travel?

Also, if a current sensor is used, how is it designed? Any examples out there of the circuit? What componets would be used? And how wired?

In otherwords, I know nothing! Other than I've worked with some bread board circuits before. But how do you combine the gears and circuitry? In reality?

Thanks for your patience

 

[dknguyen]

I'm not sure what you mean by combining gears and circuitry...yOu dont combine gears into a circuit. The motor shaft drives the gears amd the gears move the deadbolt. You control the motor with the circuit have a button-like switch device somewhere else. The deadbolt could push against it to indicate it has stopped, or there could be a notch on the gear that flicks a nearby switch to indicate it has rotated too far.

If it is current sensing then there might a very small resistor in series with a motor current lead and something measuring the voltage across it, most lilkely a comparator (an ADC is possible then you need to include an MCU into your lock which is overkill since all you need to know is when the voltage drop passes a threshold rather than its exact value)d a threshold rather than the exact value). The current going to the motor will increase once it has hit the deadbolt has hit the physical limits and stall. This increased current will cause the voltage drop across the current-sense resistor to rise triggering the comparator which will trigger something else to cause the motor to stop.

If the motor is driving a worm gear then the deadbolt is probably a toothed square rod, similar to an inch think hacksaw blade. These teeth mate with the teeth of the worm gear (the screw threads) and as the worm gear spins it pushes the deadbolt along. If you plan to build a gearbox...I wouldn't do that- very hard to get the precision required even if you are able to machine the parts. A worm gear is probably used due to high reduction ratios, but mainly because a worm gear is not backdrivable so you can't knock the deadbolt out of position by jerking/pushing/banging the deadbolt- you can only move it by turning the motor/worm-gear.

 

[BeeBop]

Perhaps you could adapt something to work for you? I think an old CD tray may do it, and you also want to do some form of error checking. Is the bolt closed all the way, or is it jammed on something? This will depend on how well and how accurate your mechanical systems function.
You may be able to do this with just a solenoid.
What makes robotics interesting is that it is a combination of the electric, electronic, mechanical and pneumatic or hydraulic systems.

 

[LiveWire]

The deadbolt could push against it to indicate it has stopped, or there could be a notch on the gear that flicks a nearby switch to indicate it has rotated too far.

If it is current sensing then there might a very small resistor in series with a motor current lead and something measuring the voltage across it, most lilkely a comparator (an ADC is possible then you need to include an MCU into your lock which is overkill since all you need to know is when the voltage drop passes a threshold rather than its exact value)d a threshold rather than the exact value). The current going to the motor will increase once it has hit the deadbolt has hit the physical limits and stall. This increased current will cause the voltage drop across the current-sense resistor to rise triggering the comparator which will trigger something else to cause the motor to stop.

If the motor is driving a worm gear then the deadbolt is probably a toothed square rod, similar to an inch think hacksaw blade. These teeth mate with the teeth of the worm gear (the screw threads) and as the worm gear spins it pushes the deadbolt along. If you plan to build a gearbox...I wouldn't do that- very hard to get the precision required even if you are able to machine the parts. A worm gear is probably used due to high reduction ratios, but mainly because a worm gear is not backdrivable so you can't knock the deadbolt out of position by jerking/pushing/banging the deadbolt- you can only move it by turning the motor/worm-gear.

Thank you for this visual illustration. That's what I need, to visualize the set up. I'm quite elementary regarding this.

What confuses me the most is assembling the componets. Figuring out how and where to place a limiting switch for example, how do you place a "notch" on a gear? with Glue and plastic knob?

Thanks for the understanding of a square rod dead bolt with teeth. Now I see how a worm gear will move it. Where would I purchase a solid bolt with teeth on it?

Do you know of any schematics that show a Comparator tripping a relay or something else that stops the motor? And what device switches the motor polarity for reverse?

Thanks for your help.

 

[dknguyen]

THe notch might be machined right onto the gear. A notch can only be used if the particular gear does not have to spin more than one rotation for complete movement of the deadbolt (since this will limit it to only one rotation). Look at notch gears for RC servos to see what I mean.

An H-bridge circuit can drive a motor in forward or reverse. You can use an op-amp as a comparator.

You'd have to get a toothed deadbolt machined since it's a custom component apart of a gearbox ($$$). Not the cheapest thing to do, especially since you are going to have to manually mate it yourself with the other gears which I mentioned before is really hard.

Ii would personally go with some kind of spring-solenoid system (or use an RC servo motor). But this results in a deadbolt that is by default always locked, or always open depending on how you set it up. It's just a lot easier and you don't have custom gears to deal with.

 

[zkt_PiratesDen]

funny how even the simplest things can get real complicated in a hurry

 

[LiveWire]

funny how even the simplest things can get real complicated in a hurry

The above is so true!

dknguyen, thanks for your suggestions. I can see now why they charge over $100.00 for a remote electric door lock.

Anyone ever mold gear teeth? What if you bought the worm gear then used fiber glass mat and resin to match the gears, then glued the fiberglass mat to a square metal bar? Without a lot of torque it might work?

Do you know of any internet sites where I could see a schematic of an H-bridge and op-amp as a comparator?

thanks

 

[zevon8]

I use linear actuators everyday at work , and I would think that this dead bolt is done the same way. Think of a threaded rod with a nut on it. Attach a regular round gear to one end of the threaded rod, drive this round gear with a worm gear on a motor, turning the threaded rod. Now, in this case the "nut" on the threaded rod is the dead bolt. Turn the rod, the dead bolt "threads" in or out on the rod. Most actuators have several gears in sets to reduce speed, and gain mechanical advantage. The most common way limit switches are done is to have 2 cams operate 2 SPDT microswitches. The cams are mounted on a shaft in the gear train that makes one revolution per full extension and retraction of the plunger. The switches are wired to disconnect the motor when full travel is reached. When the switch opens, it has a diode that connects to the other switch, allowing the motor to be operated in the reverse direction, and the cycle can repeat. Think of it like the light switches at the top and botom of a set of stairs, one turns off, but the other can turn it back on again.

 

[LiveWire]

Zevon8,

I looked at some linear actuators on the internet but the seemed to move slow. But I guess it really all depends on the gear set up and torque desired, right?

I think it would really help to see how these limit switches are placed. Know of any pictures?

I actually wanted to lock my big garage door this way. So, I have somewhat of a problem because I wouldn't want any wires on the door its self. The electronics would set on the side of the door. So, a limiting switch couldn't be at the inside end of the bolt. Anyway, a picture would be worth a thousand words.

Thanks

 

[philba]

Are you sure it's a deadbolt and not an electric strike? I've never seen a deadbolt but lots of strikes. A strike just pushes the door latch and causes the door to open - just like if you turned the knob. a deadbolt would make for a very unsafe situation in the event of a fire or other emergency.

I have 3 electric strikes in my house (it's a long story...) but didn't open them up when installing. They have 2 wires coming out and are driven by 12V with a simple relay - power to the strike unlatches the door. no polarity reversal. current is on the order of 250 mA, iirc. I think the motor is spring loaded. Power to the motor turns the lead screw and causes the strike plate (not sure that's the correct name) to push the door's latchbolt in, unlatching the door. removing power allows the spring to unwind the screw. You can hear it when the door is unlatched - about a 1 second motor+screw wirring. door pops open and is followed by a fast but soft sound. there may well be a limit switch that cuts off the current. It also has an interlock that prevents it from running if the door is open. Not sure why it matters, though.

Note that there are more sophisticated types that pivot a gate in front of the latchbolt to allow the door to be pushed/pulled open. the latchbolt isn't actually pushed in with this model.

edit: here are some strikes: https://contractorstools.com/rofu_strikes.html

 

[dknguyen]

I use linear actuators everyday at work , and I would think that this dead bolt is done the same way.

Aha! A linear actuator! I knew there was something that like a solenoid that could move at hold its position in the absence of power AND have built-in limit detection! But I couldn't put my tongue on the word since I rarely ever look at them.

This was the link to them buried in my bookmarks of parts pages:
https://www.firgelliauto.com/

H-bridge (and other parts) explanations:
https://www.acroname.com/robotics/info/articles/drivers/drivers.html

The switches in the H-bridge would be replaced with relays or transistors (I prefer using MOSFET transistors rather than BJTs). You should also place a high-speed recovery diode (called a flyback diode, also explained in the webpage) in parallel across each transistor/relay/switch so that it will allow current to flow from - to + (allow current to flow up in the diagram). If you put it the other way it will cause a short-circuit, so it's pretty easy to tell which direction they need to be pointing in. This dampens voltage spikes from switching the current on and off. Relays cannot be used if you want to rapidly switch the H-bridge on and off with PWM pulses to control speed. Transistors must be used in this case.

There are some things you should learn about transistors (like the the two opposite/complimentary P-channel and N-channel for MOSFETs, or NPN and PNP for BJTs before you use them for this. They can make it very easy or very hard to provide the required high/lowvoltages to turn certain transistors in your circuit on or off if you use the wrong transistor in the wrong place. For MOSFETs, using P-channel in the top and N-channel in the bottom makes it easy to switch them on and off (just keep in mind that a signal that turns a P-channel off will turn an N-channel on and vice versa, they are opposites of each other). This means you need to provide inverted signals or two signals to the transistors since you want an particular N-P pair to be on and the other N-P pair to be off in an H-bridge (you don't switch both P-channels or N-channels on at the same time unless you want to accomplish some kind of braking, and sending a single signal to all MOSFET transistors will do this). Using N-channels eveywhere gives you more effiency since N-channels are more efficienct than P-channel, but the voltage boosting circuitry is really complicated...so don't do it.

SO if you get the gist of the last paragraph...basically:
In your case you just want full on or full stop for maximum speed...so just use relays...they are simpler to use, more robust, and run their load capability stays more consistent over temperature than transistors. They operate mechanically, and thus will perform the same function and be operable no matter where you stick them in whatever circuit, unlike transistors which behave differently depending on the where they are in a circuit and what electrical conditions they are under.

Just don't go around unlocking your garate door for fun (the relays die out mechanically after about 5k-10k switches).

 

[LiveWire]

Thanks dknguyen for your help. I think you are correct in suggesting using a relay. For the simple fact that I'm a newbie and it would take me forever to figure out how to swtich those transisters, probably.

I'm back now to thinking I might be able to pull this off with a solenoid. Someone mentioned this earlier. If I can find one that can pull close to an inch and then would release and return to it's previous position when the power is turned off, that might do it.

Know of any good sources for hefty soleniods?

Thank you

 

[dknguyen]

I would use a linear actuator, like Zevon said, since it's like the worm gear and motor all in one already. The solenoid requires a spring for power-less standby which makes it have a default locking or open state.

Default-locked: The deadbolt is ALWAYS sticking out. Not a good idea if someone forgets to pull it in while the door is closing.

Default-opened: Requires power to stay locked and springs open when power is turned off.

No spring- the deadbolt can be jerked closed or opened in unpowered state.

A linear actuator will tenaciously maintain it's current state unpowered. You don't even need a very powerful one- just enough to push the deadbolt (I'd make the deadbolt separate from the actuator shaft to put the stress off the actuator, and it you could use a giant hunk of steel or the deadbolt that way).

 

[LiveWire]

Yes I see. Also I would imagine a linear actuator might push and pull further? What do you think?

I was actually thinking of having the powered device pull a pully approximately a quarter turn. And the pulley then pushes or pulles two pistons on both sides of the garage door to either lock or unlock.

 

[dknguyen]

A liner actuator can push and pull farther than pretty much any solenoid (for a given size). Linear actuators seem to start out a bit bigger than solenoids though.

A pulley wheel is a great idea with the linear actuator pushing one end and a stick pushing and pulling the deadbolt (kind of like the wheels of the old steam locomotives). Linear actuators can be long- too long to fit horizontally to the garage door. I was trying to think of a way to make a "right angle" linear actuator so you could place it vertically and make it much thinner, but yeah and pulley wheel with driving rods (instead of ropes so you can push as well as pull) would do the trick. Plus with a pulley wheel you can change the radius of the wheel to get the required amount of movement if your linear actuator moves too far or not enough, so then you don't need electronic motion controls to limit the movement- it's all done with the internal mechanical stops of the linear actuator.

Well, I guess you could also mount the linear actuators horizontally on the garage door instead of the wall beside them. You then have lots of room to move the actuator left and right to accomodate the actuator's throw and length. Possibly simpler too. Depends on how many garage door mechanisms the actuator can hook onto when the garage door is opening.

You want one on both sides of the door? I was only thinking one. Easy enough to make two mirror devices though and stick them on each side.

Oh here's another idea...maybe you can just use springs latches with a solenoid for your garage door. The garage door is always "locked by default" and requires power to open, but no power to close or stay closed because when the garage door comes down it just slides into the spring latches and locks on it's own whenever it's closed. Kind of like some of the large plastic buckles on backpacks and other things. It's not very deadbolt-like though.

 

[LiveWire]

"Oh here's another idea...maybe you can just use springs latches with a solenoid for your garage door. The garage door is always "locked by default" and requires power to open, but no power to close or stay closed because when the garage door comes down it just slides into the spring latches and locks on it's own whenever it's closed. Kind of like some of the large plastic buckles on backpacks and other things. It's not very deadbolt-like though."


Yeah, this is what I was sort of thinking about. Perhaps two latches on both sides which automatically latch when the door comes down and then unlatches with power so the door opener will open it.

I hadn't thought of a buckle type latch. That's something to think about. However I was thinking something almost like a fence gate latch. When a gate swings shut that bar pushes into the latch which then falls over the bar. But that kind of a latch will not work for up and down motion. So, I was trying to think of something similar that might work as it comes down.

Now it sounds like we might be back to the solenoid idea. I was thinking if I could pull a cable from the center of the door with a solenoid it might lift both latches and allow the door to go up.

But if I mount the solenoid on the door then I need to use battery power because wires coming off the door could easily get caught somehow. And if I us battery power then that complicates my wireless remote idea. I was going to use an X-10 wireless remote for power . . . any ideas?

 

[dknguyen]

I didn't mean an actual buckle. It's just the only thing I could think of that works that way. I know there are others but I can't remember any. A gate latch makes more sense, but I think a doorknob makes the most sense. Maybe something like the door knob thingy that holds the door shut (what is that called anyway?!). Kind of like a really short deadbolt with an angle so that the downward motion of the door can push it aside so it pops into a piece of metal bent to be like a rectangular tube that is bolted onto the garage door, just like a door knob. The spring can hold it sticking out and give when the garage door comes down and the solenoid can pull it out whenever you need to to open it. It would have to be very short though compared to a deadbolt length.