Jump to content
Security Installer Community

Wiring Old Alarm System Into Computer For Monitoring


dinki

Recommended Posts

IIRC, the modules for MH are written in Perl so I don't have much choice in the matter. MH can handle the arm/disarm/alerts (if I program it correctly, that is).

I haven't seen an alarm module, but I've only started looking it it.

I'll probably go with the serial port rather than LPT as I should be able to get a PIC that does serial out. One less thing for me to fiddle with.

?? Tell that to all my home and work pcs that run 24/365 . Yeah, I am not trying to compare what I'm attempting with a 'real' monitoring system, but what I propose is much better than what I currently have (nothing).

Perhaps I'm oversimplifying things too much, but really, at it's lower level, it's just a bunch of switches and a mechanism to determine state change.

hi dinki,

well perhaps i could have qualified my statement by saying 'most home users's pc's', but then you being pc aware at least technicaly i could assume you would know exactely the validity of what i was saying.

as to being 'much better than what you already have', i totally dis-agree with you. at the moment with no system at all if a neighbour see's something suspicious he/she may take action, call th police or whatever. but if you install your at best 'cobbled together' alarm system and it false alarms while you 'get it right', you will have far less protection from your to date observant neighbours, as they will think "don't worry it's only dinki's dodgy alarm going off again", so when you get fed up and fit a proper panel your level of protection will be 50% at best of what it could have been, had you done it properly with the right equipment in the 1st place.

door contacts are as you say 'simple' devices offering 2 states but they cause the most false alarms, poorly fitted or faulty reeds are some problems, but not securing door and windows is another. you will walk into the trap of thinking the alarm set so they all must be closed - WRONG!

all it means is the magnets are close enough to the reeds for the system to set at that time, if a door that was not latched blows open and then slam's shut causes a false alarm, so after the false alarm you check it and find it secure and so are totally misled even baffled by a 'simlple' door contact, and this is only one scenario.

regs

alan

If you think education is difficult, try being stupid!!!!

Link to comment
Share on other sites

i think you could get a panel for less than 200 bucks (napco, dsc, fbii or so).

if you want to fiddle with PIC you need to feed detectors with 12VDC and they consume from 10mA to 20mA (usually) each. door contacts don't ofcourse consume any current. something you need to think is the reed vibration when opening so you should start with times such as 500ms open until it should be recognised open. also you might want to measure the actual current flowing through the loop - not just open/close loop.

if you want to do it hard way make a small chip" of some cheap PIC so that is capable ( 7805 should do that ) to take 12V power and measure the contacts fitted inside the detector. PIC could also measure the feeding voltage from the detector and communicate with your computer directly over RS485 or if you want to make it harder then over TCP-IP. lantronix TCP/IP module is thou approx $50 but has serial interface at TTL - level... this way you could supervise your detector actually is at the end of the line.

prepare for a long project thou...

also i wouldn't expect EN - graded or U.L. approved installation afterall.. :roflmao:

Link to comment
Share on other sites

  • 3 weeks later...

I have read that you no longer want to do it, but for reference here is what I would do if I wanted to link sensors to a PC without a microcontroller:

- If you have less that ~10 signals, you can connect them directly to the printer status feedback lines and to the data lines of a bidirectionnal parallel port.

- If you have more signals, you can etch a small PCB with some parallel in/serial out shift registers that you read from the parallel port.

Your PC can then regularly poll the status of the inputs and do the necessary actions if something changes. You can use parallel port lines such as STROBE to trigger the siren.

Using a microcontroller takes more time but allows for more features, especially it can run without the PC being on, and it will continue to run if the PC crashes.

I made a very simple alarm system a few weeks ago using a PIC microcontroller. Adding status reporting over a serial port wouldn't be hard once this is done.

The alarm firmware is basically a state machine with debouncing on the inputs, and status reporting on a LED display (last event memory, armed or not, ...)

[as you can see on my current avatar, I love microcontroller based projects ;D]

With a microcontroller or not, your circuit will usually work with a 5V power supply that you need to either take from the PC or generate from an external voltage. Using a 7805 is a good idea for this purpose.

With simple NC zones, you simply send a current limited (with a resistor) voltage to the loop, and receive it back from the sensors. Always protect your inputs using 10K resistors in series with them. Most microcontrollers have what is called clamping diodes that prevent the voltage on their input from exceeding the circuit's supply voltage. But as these diodes are small, you need to limit the current flowing through them and this is what the resistor does.

Protecting the input this way is important for two reasons:

- Wiring error when fitting the system

- Bad guys trying weird stuff if they have access to one wire (...)

If you're paranoid you can provide a full galvanic isolation to the system using optocouplers for the inputs and 1:1 transformers for the supply voltages.

If your PC is far away from the circuit, you can use differential wiring such as RS485 (as suggested) to transmit the data; Maxim makes good, small, and simple to use RS485 transceivers integrated circuits.

Don't forget the backup power supply. A PC would need an UPS whereas a microcontroller-based alarm can run off a small battery for months... especially if you put the microcontroller in sleep mode and use the pin change interrupt to wake it up when something happens (just an idea). Your microcontroller can monitor the battery voltage quite easily as most of them have an ADC.

Now, about fiability - be careful with crystals. Always use a small series resistor with them, or they can simply "blow". Don't ask me why, but a lot of electronic engineers I know have reported problems with unprotected crystals. If you do not need a precise oscillator, that is if you don't do serial transmission or real-time time-keeping, you can use a RC oscillator.

I'd love to have more electronic talks in this forum !! :D

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use.