RFID Door Opener

The purpose of the RFID door opener project is to manage access to the Space by using RFID tags or cards.

Requirements

Basic

RFID reader - Needs to be able to read cards through the door

Microcontroller - Needs to interface with RFID & Actuator, store IDs for authentication

Actuator - Needs to be able to turn deadbolt both to unlock and lock it

Power supply - need a hardwired power supply

WTH (wonderful to have)

Microcontroller

interface with ethernet / wifi and logs when and who comes into the space

Some sort of voice system for debugging / saying hello "user" when they walk in

Acuator - unlocks and opens the door as well

Power supply - Backup battery supply that can run for at least 24 hours

Helpful URLs

The cons are of course debatable

http://bear24rw.blogspot.com/2008/12/rfid-door-unlocker.html

Uses

Basic stamp

Parallax RFID reader

Simple Servo

Pro

Finished product

Decent writeup on blog, lots of detail

Con

Looks like it is battery powered

Can't lock the door again

http://appdelegateinc.com/blog/2010/10/04/arduino-rfid-security-access/

http://appdelegateinc.com/blog/2010/10/06/rfid-auth-using-arduino-and-python/

Uses

Arduino

Parallax RFID reader

Pro

Has code posted

Arduino (already have one for the project)

Con

Not a finished project, IE the actual door opening mechanism doesn't exist

For authentication using python the controller would have to be hooked up to a server

Uses

Keypad

Arduino

Easy Driver board (drives the stepper)

Stepper Motor

Bike Gears / Chain

Pro

Uses Arduino

Complete design

Con

No code to leverage

No text writeup that I could find... just youtube videos

Physical Design

After some discussion we've decided to pursue two paths for the physical construction

Use existing lock

This path uses a servo / DC motor / stepper motor as the actuator and uses the existing door hardware

Use a new lock

This approach uses either an electronic lock (Rob is looking into this) or possibly an electronic strike.

9/11 - the consumer-grade electronic deadbolts won't fit that door. So now using the existing lock is looking like the easier approach - Rob

Software Design

Accepts the RFID tag that is read, checks it against a list and then responds accordingly. We've decided to use the code located at [1] as our base and we will be adding to it for more functionality.

Ross - Delete key's without a master delete key and preserve all the keys in memory

Joel - Interface with a stepper motor instead of a Electronic Lock

Successfully integrated stepper motor and Easy Driver in with our RFID system.

           Potential Problem: The stepper does not move freely when it is powered, so using the key will still be an issue. 11/18/10
           * Did you try raising enable\ to turn off the motor drivers? this should make it much easier to move, but not free-spinning. Morcheeba 22:18, 19 November 2010 (PST)

Contributers

Rob

Joel

Ross

Dan

Log

11/23/10

I measured the torque required to turn the lock handle: 4.7 inch pounds (AKA 0.53 newton meters; 5.4 kilogram centimeters ; 0.39 foot pounds)

Note that the measurement was taken by finding the hardest part of the unlock motion, wrapping a headphone cord around my shape-lock adapter with a small tool bag hanging from the end, and filling the tool bag with an assortment of heavy objects until the lock barely turned in response. Then I weighed the heavy objects (7.5 lbs) and measured the radius of the adapter (.625 inches). So we're talking high precision.

Between the errorof the measurement and likely high friction losses in whatever pulley scheme we devise, I figure we should shoot for 2-4 times the torque. It sounded like the stepper wasn't necessarily working as expected anyway (?), but assuming it was the SparkFun web description says "Holding Torque : 2.3kg*cm" so figure we'd want to step it down somewhere between 4:1 and 10:1.

-Rob

11/16/10

We managed to get all the electronics and some basic software up and running at the Tuesday meeting

• Got a sketch running on the Arduino that was able to
  • Read RFID Tags
  • Know if the Tag was the Master add or delete tag
  • Add / Remove Tags based on which Master tag preceeded
  • If the tag was normal and in the DB on the EEPROM then a switch was triggered

Things we've got to figure out / work on

• Software
  • Interface with Stepper and not a single switch
    • This will require us figuring out how far the stepper should turn
    • Figure out how to know what state the lock is in

• Hardware
  • Took some snapshots of the door with a measuring tape next to it so I can make a better 3d model of the door and what we are trying to do
  • Because of where the lock is we are going to have to do some sort of pulley system.

9/6/10

We messed about with this some on Saturday (9/6/10). No success but some lessons learned...

• Software hacking without internet access is a pain. I still assume the (arduino) software side is easy, if one could look up any docs or download any drivers.
• attaching anything to the existing mechanism looks hard. I messed about for quite a while with some bits of wood and a couple erector sets, and couldn't make it happen.
  • There isn't any room between the bit you turn and the door jamb;
  • attaching something while preserving the manual function of the lock is complicated
  • attaching anything without damaging the existing lock is hard.
• as we were giving up, Kurt happened by, and we asked if we could simply replace the whole lock. He said "Sure, if you give me a key". I felt dumb. He added that he should probably tell the landlord, but they were generally OK with anything that didn't destroy the place.

Upon reflection I came up with some more thoughts:

• it would be nice if we could still unlock the door by key if our mechanism fails for some reason.
• based on my laymans understanding of the fire code, the door needs to be unlocked while anyone is inside, or unlock based on a handle you just push on. Just unlocking it when we're there is easier, but also makes it easier to forget and leave it unlocked. Optimal would be always-locked from outside, always-unlocked from inside. Not sure if there is a great answer here; maybe a big obvious led sign indicating the lock status so we don't forget.
• this link: [2] is interesting. He hacked a powered deadbolt, making it RFID controlled instead of keypad. The one he used appears to be among the cheapest: $65 at Amazon. I'm not yet sure if the dimensions of that and the door are compatible, but if you follow the article through, the board in that lock apparently has solder pads ready to go for open & close switches.
• also interesting in that same article is the distinction between an electronic deadbolt and an electronic strike - it hadn't occurred to me to make the wall side of the mechanism (the strike) move in response to our electronics. A mechanism on the wall that was totally separate from the key-based mechanism on the door would simplify multiple complications all at once.

Spring 2013

RFID Reader

Implemented with a donated DMP XR500 Series Access Control Panel sampling an Weigand RFID reader and controlling an electromagnetic locking system. The system is administered via a network connection using DMPs Remote Link software. Another Weigand RFID reader was attached to an Arduino Leonardo to act as a card reader for the software. It works by simulating a USB keyboard and entering the RFID card code when the green button is pressed. A card is swiped over the reader to get it's code.

Arduino RFID reader sketch: media: w26usbReader.ino.txt (strip off .txt suffix)