I like electronic locks because they are relatively pick-proof. And if you eliminate the keypad and reduce things to a single button, that looks like the definition of "impossible". I'm working on a door unlock system based on a rhythm pattern entered with a single, unmarked button, probably a small disc on a plunger that activates a SPST momentary switch. With a 4 to 8 step press-release pattern that must be completed within a specific time *window*, I think this will be secure.

But what if ... What if an evil doer got a copy of the schematic? Could he decipher the correct input pattern? Can you? What do you have to do at SW1 to cause U1Dpin11 to change state for 0.5 seconds?

Note: power decoupling, transient protection, unused gate termination - all done.

ak

 

@AnalogKid
Super interesting...
Sorry I do not have any advice or input but very intriguing. Will definitely be following this

 

1 long
4 short
1 long

Maybe
1 long
4 short

 

Hey, a co-worked asked me if could turn the copier on. It shut down at 5:00 and you needed a 4 digit code. His request was reasonable, so I said I'll try. I had the code in about 10 minutes. I was given a briefcase with a 3 dgit number. Open in about 5 minutes. Tried 2 numbers based on who it belonged to.

I can open simple L-R-L type combo locks like Master in 20 minutes. It takes me a little longer to get the combination, Self-taught.

 

a co-worked asked me if could turn the copier on. It shut down at 5:00 and you needed a 4 digit code. I had the code in about 10 minutes.

That is why I don't like keypads.

ak

 

After looking again I'm thinking...

5 short
1 long

EDIT:

I miss counted.

4 short
1 long

 

4 short
1 long

Justify that.

ak

 

That pathetic huh?

Well my thinking is the first push will bring the 4017 out of reset, then 3 more will advance the counter to Y3, then one more will advance to counter to Y4 and must be held to overcome the RC created by R4 & C4.

 

my thinking is the first push will bring the 4017 out of reset,

How?

 

I see trouble ahead.

The clock pulse is generated by shorting a 2.2uF capacitor to Vcc?

Think this through a bit more.

 

I did. SW1 is an industrial type, not a gold-contact, 10mA max., delicate little flower. C1 is a MIL-grade MLC. I might swap D1 for a 1N4002; I have some getting dusty.

ak

 

I am guessing from your response that I'm way off base, so I will withdraw and let someone else have at it.

 

I am guessing from your response that I'm way off base

Not so; you've got the back end right. It's the stuff between the switch and the Reset input (and the complete lack of visual or audible feedback) that make this lock impossible to pick.

ak

 

Embedding the code in hardware is not a good idea. It is reasonable to believe that at some point you'll need or want to change the code. Maybe the wrong person learns the code somehow, the lock changes owners, any number of reasons. If it were my lock, I would make it possible to change the code somehow. If you want to do it all in hardware, maybe add a DIP switch that you can toggle to select the code. Or add a little micro processor and put the code in software. Any way that allows you to change the code without having to completely redesign and rebuild the lock.

Side note; for the locking mechanism itself; use a worm drive gear and solid circuitry to drive it. Anything held by a spring can be jarred to unlatch, mechanical relays and plunger style solenoids are susceptible to strong magnets.

 

for the locking mechanism itself; use a worm drive gear

I already have chain drive. The project is an outside switch for an old, domestic garage door opener.

ak

 

Ah cool! Should you want to do it in software, there is existing arduino code out there. It's used for the escape room "secret knock" type puzzle, the player knocks a pattern on a door and the arduino compares the pattern to the pre-recorded pattern. You can just replace the piezo knock sensor with a button.

 

I like electronic locks because they are relatively pick-proof. And if you eliminate the keypad and reduce things to a single button, that looks like the definition of "impossible". I'm working on a door unlock system based on a rhythm pattern entered with a single, unmarked button, probably a small disc on a plunger that activates a SPST momentary switch. With a 4 to 8 step press-release pattern that must be completed within a specific time *window*, I think this will be secure.

But what if ... What if an evil doer got a copy of the schematic? Could he decipher the correct input pattern? Can you? What do you have to do at SW1 to cause U1Dpin11 to change state for 0.5 seconds?

Note: power decoupling, transient protection, unused gate termination - all done.

ak
View attachment 233955

As a former locksmith, and very knowledgeable one the subject, while I cannot answer your specific question I will answer a heuristic one- other locking systems employ this technique, but usually with more than one button, increasing the level of difficulty. The problem most users have, which makes them dislike this, is the time aspect- even if you know the code, you can be forced to attempt it 4 or 5 times because you're just 'off' that day for whatever reason- most people don't have perfect timing, so timing related things can become a nuissance. IMHO.

 

As a former locksmith, and very knowledgeable one the subject, while I cannot answer your specific question I will answer a heuristic one- other locking systems employ this technique, but usually with more than one button, increasing the level of difficulty.

As an experienced amateur lock-person, I completely get your point. And disagree.

Nothing is more ambiguous than a single round button. Not doorbell-looking, not security-company looking, not explosive-proof looking; something with zero personality (read: clue). It is difficult to convey exactly how little information is given away by a single, small, round, plastic button.

Press it - nothing happens
Tap it a few times - nothing happens
Hold it down - nothing happens
Tap it a few times than hold it down -nothing happens

That exhausts the thinking abilities of 99.9996% of the earth. Ahh, but you are not in that group. You are a trained, experienced, professional. So, what would you do next? Note that your professional eye tells you that the button is not a part of any standard switch you know. And it isn't; it is the head of a small plunger. Internally, the switch is 1" off to the side, but you don't know that.

ak

 

If everybody was a CW fist then keying a long password in Morse code would have a definite attraction. I could even see making the paddles explicit and requiring Iambic-B keying. No run of the mill crook would have much of a prayer.

 

It is reasonable to believe that at some point you'll need or want to change the code.

Not in this case. It is an auxilliary system, I'm the sole user, and it's a rhythm patern I created.

Maybe the wrong person learns the code somehow

It's a rhythm pattern with false entries. Even if someone were secretly looking over my shoulder (in my driveway?), they would not know what they were seeing. BTW, incase someone on this forum wants to raid my garage, the above schematic is not complete. There is additional timing and gating.

the lock changes owners

When I kick, the system can be abandoned in place without affecting the normal buttons that the house came with.

ak