Hey! So… it appears that Samsung has changed their firmware in the latest release of the SHS-3321 lock so it no longer supports NTAG213 chips. This strikes at the heart of the issue surrounding compatibility - it relies 100% on lock manufacturers.
I think it’s time to start talking about open sourcing a deadbolt lock design.
Let’s focus on mechanical design first, then electronics second. My personal preference is for a lock that kinda looks like a NEST thermostat on the outside… RFID/NFC only… battery box on the inside… no extra holes that need to be drilled into the door, just use the holes drilled for the lock itself.
After a quick google search session, most of the existing DIY designs involve 3d printing and assembling the lock from PLA or ABS and relatively weak servo motors.
Material-wise, depending on thickness, we’d have to go with steel or something of equal strength and rigidity, possibly aluminum if we can make sure it’s strong enough.
Maybe we start with an electronic deadbolt/solenoid already available on the market and modify it? Perhaps a modified metal faceplate for it that houses the scanner?
When it comes to hardware, we can go down the solenoid path, or the motorized path, solenoids require a boat load of power though compared to a small, DC motor with a high gear ratio.
Testing to make sure it can’t be easily broken into is one of the more important things to keep in mind. Security is something people will need to be sure of if they’re going to put one of these between the outside world and their families.
I like this idea a lot. Seems feasible from a manufacturing perspective as well. I’d love to contribute however I can.
Agreed, some form of steel would be the best option for at least the outward facing side of the lock.
I would argue for the motorized path, as that seems to be the most common on the market and the simplest to produce and have operate.
For security, having the only option be RFID cards/implants would reduce the chances of having a weak backup lock core be a security threat. You would also want there to be no access to in internals from the outside, with the antenna and faceplate being solidly attached, and perhaps potted in resin to prevent the easy access to the internals/mechanism by simply breaking the front off of the lock. This being the case though, you would want to have some form of emergency battery connection on the front facing side as well, like the samsung locks do.
Downward facing battery terminals would be out of direct view on the faceplate, and ideally, it should have a full-bridge rectifier and short circuit protection to protect the electronics from reverse polarity or someone trying to kill the lock by bridging the contacts with metal/wire.
How about a retrofit/piggy back board for existing locks. If we can’t change existing firmware, maybe we can bypass it (esp32 based solution) while using existing reader, hardware & locking mechanism. This approach could work on different brand locks. If we are happy injecting chips into our hands, voiding a lock warranty shouldn’t be too much of a problem (still unsure about insurance companies in the event of burglary though)
If it was a piggyback, the insurance company wouldn’t ever really find out that it’s modified, like they aren’t going to go probing around the internals looking for unauthorized circuitry lol
That being said I love the idea of using existing hardware, maybe we make an open-source circuit design that taps into the cable that the Samsung SHS locks use to connect the two halves of the lock together?
We could even make it solderless and have the exact same connector and wiring harness the Samsung lock uses so you wouldn’t need to cut the cable. The question from there is whether Samsung or any lock manufacturer would get hostile and/or take legal action. Does anyone know the legality of this?
You’re right, I can think of plenty of companies selling similar products in different verticals. Examples: car tuning devices, console mod chips, hardware hacking devices, etc.
First step would be to work out the screen/reader/numberpad connector and pinout. Getting this connected and working as a standalone device would be half the battle won.
Like @MikeH1021 says, a solution wiith a solderless board using existing connectors would be perfect.
My choice would be ESP32. Small form factor, standardized architecture, Bluetooth and wi-fi, deep sleep with touch to wake. Arduino coding so nothing esoteric. Cheap and readily available.
Why do you want Bluetooth and WiFi capabilities? Would the lock not be user friendly enough with just NFC? I feel like you’re sacrificing security. What about the PN7462?
PN7462 sounds excellent thanks for the info I hadn’t come across that microcontroller before! I was just going with what I was familiar with and able to replicate functionality of locks currently on the market, not just the Samsung ones.
Lots of good ideas here. I think though that going with our own design is the best idea because buying locks from another producer still means we rely on that producer to keep producing that lock, and it also means the cost is potentially higher for the end user who must first buy a retail lock from someone else, then buy a mod kit. The mod kit will probably need to be a little more expensive to recover production costs quicker because it will always be at risk of becoming obsolete once the target lock goes out of production… then you need to get to work designing a new mod kit pcb…
If all this effort is going to go into something like this, I think it should be something we can reliably produce indefinitely. For things like the bolt/latch bit that functions as the actual lock that slides into a recess in the door frame, I don’t think there’s a point in designing that… I’m pretty sure you can buy those parts off the shelf since they are so standardized at this point… but the lock mechanics themselves… if we keep it super simple… no key cylinder… no pin pad… no screens… just an NFC antenna behind a simple plastic face on one side and the business stuff on the other, we should be able to make this happen.
My biggest question right now is - do we know anyone in China who can walk around lock manufacturing district to find off the shelf parts we can cobble together to cover most of the working basics? For example, the bolt part I’m sure we can just find somewhere. This isn’t exactly what I’m talking about, but my point is that this is an off the shelf latch you can just buy;
There must be other parts like this we can just buy.
I’ve nearly fully automated my home using BLE, WiFi, zwave, and zigbee. The more I can incorporate NFC the better. I would totally love to be able to wave my hand at my door to unlock it. Right now it operates with a keypad and BLE. I’m down to help with this project in any capacity I’m able.
Sounds like the main components are fairly cheap yeah? The next big part is the actual design of the outer casing of the lock? And as amal ststed something similar to the nest thermostat would be ideal. I can look around my wearhouse tomorrow at work, and see if I have anything that can assist in making a prototype