My DIY Security System

A while back (during construction of my new house), I wrote an article about the capabilities of the DSC 1832 alarm panel. Well, time got the better of us and I wasn’t able to wire the house myself. Instead, we had a third party alarm company provide “a la carte” wiring service. This included (all 22/4):

  • Phone line to the outside utility area
  • Front + Side door
  • 4 motion detectors
  • “Basement” wire (which I used for freeze)
  • Siren Wire
  • Keypad

And a larger gauge AC wiring from the outlet over to the alarm panel

I went ahead and purchased the following items:

Tools:

  • CATV Stripper (for outer sheath)
  • Wire stripper
  • Tone tester
  • Screwdriver set
  • Drill (and drywall screws, drill bits etc)
  • Multimeter (highly recommended)

 Basic System

To get booted up, you’ll want the following:

  1. Snap the board into the alarm panel using the stand-offs provided.  They punch through the back of the case.
  2. Mount the alarm box. If your installer has all the wires come out of a single hole, mount the box over that hole. Use heavy drywall screws and do the top two first, then the bottom once the box can rest on the wall.
  3. Attach the AC line wire to the alarm panel. The screw terminals require a bit of fiddling, but they can hold multiple wires (which is convenient for the Aux power bus). You have to come up from below at a 45 degree angle. If you’ve mounted them correctly, when you tighten you should not be able to remove the wire.
  4. Attach the other end of the AC line wire to the adapter. For the time being, you can remove the center screw that holds the adapter into the socket.
  5. Plug it in.

You should see the panel boot up. After you’ve got lights on you can unplug it.

Keypad

Mount the keypad on the wall using four drywall screws over the keypad wire. Then shorten the wire down to a few inches, leaving enough room to clean them up.

Use a cable stripper to remove the sheath around the bundle. Adjust the blade until it barely rests upon the sheath of the wire and rotate.

Use the wire stripper to strip each of the leads and screw each into the terminals on both ends.

Once it’s all connected you should be set to go.

Settings worth fiddling

  • Remove the TLM (Telephone Line Monitoring)
  • Enable the AC Power Indicator on the Panel (disabled by default)
  • If you want to disable the siren, you can enable the keypad beeper for alarm condition.

Troubleshooting

An alarm zone, by default, has three states:

  1. Open Circuit (Violated)
  2. ~6.5k loop resistance (Secure)
  3. Short (Tamper condition)

You can diagnose these zones with an ohm meter. The other trouble codes you’ll encounter are low DC battery voltage (plug in the battery, duh) and Bell trouble (either attach the bell or run a 1k resistor in between the contacts to get rid of the trouble).

Wiring

Each alarm zone should be wired with a loop and a series resistor. For this purpose, you will either need a soldering iron and electrical tape or some end caps to butt join the resistor lead and wire.

Some sensors require power in the form of 12V. It’s best to keep the convention of using the Red and Black for 12V and the remaining wires (Green and Yellow) for the loop.

Monitoring

Using the Envisalink IP Module, I was able to set up remote monitoring. Because we planned to snooze the cable, I used a prepaid h2o GSM sim, a USB adapter, and a TP-Link router to get internet over cellular. Combined with a backup UPS, this should provide moderate resilience to power failure. In addition, the Envisalink online service provides network monitoring and will alert on network failure and I get AC failure alerts through the alarm panel.

The Envisalink IP module can be addressed directly over IP, but unfortunately the interface is basically straight TCP. The result is that it would require an idled TCP connection to receive updates along with a stateful monitoring TCP program. There are several open source implementations available but their service is sufficient for most use cases. It’s a bit unfortunate in its more proprietary nature and lack of a more interesting API, but it does exist should you wish to do some local work with the device.

4 thoughts on “My DIY Security System”

  1. In these parts, the cellular or as we like the call them mobile broadband providers no longer provide public IPs to most of their users, so accessing the device remotely could require the use of a VPN service and a router with that capability. If a public IP is available, but not a static IP, a dynamic DNS capability in the router is an alternative,

    1. You are correct. The implementation of Envisalink appears to be a TCP push style getting around the NAT issues. I didn’t find a way to reconfigure it to use a personal server (likely that there is no such support) so if you go that directly you may find yourself SOL.

        1. Ah. It is a bit confusing. There are three interfaces that I know of.

          It has a built in web server on port 80 with some minimal information and basically no apparent ability to control the system. I was more than a little disappointed that basically no effort was put in here.

          It also has a serial-over-telnet interface which allows you to manipulate the system by sending commands and subscribe to status changes by listening to the TCP connection. This requires, as with the web server, that you are able to send TCP connections into the network that you want to manipulate (impossible with the cell phone modem as it’s NATed).

          It has a third entirely undocumented interface which communicates with the Envisalink website transparently. It basically “phones home” when you power it up and then you can associate the Envisalink module by MAC address to your account. In this manner, all that you need to manipulate it is outgoing TCP capacity and you communicate through their server. The unfortunate problem with this is it leaves you dependent on a third party service which likes (but does not require) subscription fees.

          It would be possible to replace this with your own equivalent but it would require some serious hackery (DNS spoofing and reverse engineering their protocol maybe?). If that were the case, it would probably be better off using your own Arduino and reverse engineering the keybus. You may be able to hack the firmware off the chip and disassemble it.

Leave a Reply