[hackerspaces] Open Access Control for Hacker Spaces v1.0

[john arclight]

Your architecture sounds like a good approach for expandability.  We
were looking for something that could run stand-alone, without a PC or
net connection if necessary.  There are a lot of commercial products,
but they fall into two categories:

1. Small, self-contained reader+microcontroller for one entrance.
Price: US$50-300. Functionality: Limited
2. Networked access control systems from a reputable security vendor.
Price:  US$1000+

The commercial products are all proprietary and have expensive
accessories. Wiegand26 is a bus protocol that has been around since
the early 1980s. It uses a 2-wire, TTL-level signal that is
asynchronous (there is a zero-line and a ones-line, plus
ground/power/etc).  Lots of knock-off prox card, keypad and other
readers are available from Hong Kong on the cheap.

It's not a particularly secure protocol, but it's adequate for out
semi-public space, especially if the reader and associated wiring is
behind the glass storefront.

We were able to pick up 3 readers and baout 50 tokens on eBay for
under US$100. We thought it would be better to focus our limited
resources on the control system, rather than the readers which at
least nominally work.

One major drawback to Wiegand26 is that the readers do not have any
kind of unique addressing. Your bus approach seems like a sensible way
to reduce the number of pins needed and complexity.

I believe that the larger commercial access controls usually have
small controllers based on a PIC or similar uC, which aggregate a
number of doors or sensors and then attach to a "master control" panel
via RS485.

So maybe a goal for the board would be to make it flexible enough to
be used as a "master" or "slave" system, depending on how it's
configured.

Arclight

On Fri, Apr 9, 2010 at 2:06 PM, Royce Pipkins <royce.pipkins at gmail.com> wrote:
> Interesting. I am coming close to being finished with my own access control
> system that will initially be used at Bucketworks and then later at
> Milwaukee Makerspace (If Milwaukee Makerspace ever gets a space to access).
> As soon as I get it all going I'll share all my stuff as well. (I'm in a
> rush to just plain finish for right now, hopefully in another week)
>
> I went with two Arduino's. One on the outside that runs an 2x16 LCD,
> Sparkfun's ID-12 RFID reader, and a 3x4 keypad in case of a lost card or a
> special event with a special non-member access code. The exterior Arduino
> talks via RS-232 to an interior Arduino that runs the door strike (or
> whatever) via relay and also talks to a Drupal server via RS-485. An
> addressed master-slave protocol is used on the RS-485 side so that you can
> have about 30 access control units on a single bus before you need a
> repeater or a separate bus. The user list is kept in Drupal, not in the
> Arduino. A program running on the same server as Drupal will be the master
> of the RS-485 and will turn access requests on the bus into web calls to
> Drupal and relay the access granted/denied response back down the bus.
>
> Anyway the thought behind the RS-485 was to be able to have centralized
> control of access to more than just doors. Perhaps we'd control access to
> heavy machinery as well so that only those who'd been trained on a given
> piece of equipment could turn it on.
>
> So, I guess my suggestion is to pull the user list out to a web server of
> some type.
>
> I gather this Wiegand 26 thing is a pre-made module that can collect an RFID
> tag or code from a PIN pad?
>
> Regards,
> Royce
>
> --
> The real problem is not whether machines think but whether men do.
> B. F. Skinner
>
> _______________________________________________
> Discuss mailing list
> Discuss at lists.hackerspaces.org
> http://lists.hackerspaces.org/mailman/listinfo/discuss
>
>