Stupid alarm systems

I suppose there may be some rules, and regulations and even logic, but it escapes me...

If I was making an alarm system* and someone tried to set the alarm and a fire escape was ajar, what would I do?

1. Beep in some cryptic way then set the alarm anyway so the bells go off and the idiot setting it (my son in this instance) realizes they have an issue.

2. Beep in some cryptic way then set the alarm ignoring the fire exit, so if someone does wander in the PIRs will pick them up and cause the alarm to go off.

3. Beep in some cryptic way and not set the bloody alarm at all.

To be honest I would probably do a bit of the first two. Set off the alarm but only long enough to make a point, and then go to a mode where set but ignoring the fire exit. Maybe if fire exit closes within a few minutes assume the problem was fixed and include it (and stop the bells).

What I would not do is option 3, not setting the alarm at all!

The alarm system in that case knows someone has tried setting the alarm and buggered off. It knows they have not canceled the setting and knows they have not gone back in the building. It basically knows they have not understood the cryptic beeps and have buggered off assuming all is well. So why on earth leave the building with an open door and alarm not set at all. That has to be the worst scenario of all.

As you can imagine, we have a nice high-end alarm system on the office and this is what the bloody thing does. Thankfuly I have extra monitoring on it and I get a text to say someone failed to set the alarm. So I had to go down there and sort it. Without that extra precaution we would have had an un-locked and un-alarmed office. At least the surveillance cameras are 24/7, but still!

* I say "if I was making an alarm system". Obviously I did, 20 years ago. I made a door entry and alarm system for my house in Newcastle. It was designed and built by hand using leaded components, wire wrap sockets and a bare breadboard. It used a lot of wire wrap. 6502 based. The door entry was mag card (credit card) using a head from an old cassette tape player. Obviously all my own software - I even wrote my own assembler. It worked well for many years until I re-made it from a PIC 16C84. Sadly, these days, double glazed uPVC doors mean I have to use a key at home.


  1. What sort of alarm lets you think it's set when it hasn't. Every modern alarm I've used goes through some process of beeping until the final door is locked and then gives a confirmation tone. All but the cheapest can and should have a clear display of the problem on a screen, like "Fire exit open". If not then shoot the installer or his salesman. Then shoot the person who can't tell a cryptic beep from the normal one and finally shoot the person who trained him in how to use it.

    Situation 2 is a non-starter as it'll cause a false alarm when a cat wanders in or the wind moves something on a desk, while a burglar will have time to grab something valuable with the inconvenience of having to jemmy a door or smash a window.

  2. Yes, it bleeps, but the display is not actually on the door in question. There are separate alarmed zones and display is only on one. And people do not understand the beeps.

    Un-alarmed is worst case. If you can convince the idiot setting the alarm they have failed, that is best, I agree.

  3. Couldn't you have additional display keypads? The popular Menvier/Cooper systems allow more than one (exact number depending on the system). I have three on my home/office as there are that many entrance/exits that can be used independently.

  4. I may be able to, yes. As it happens it is likely to change and we may have kepads near doors. The issue is that the person setting it was an idiot - it happens. The alarm system knows someone tried to set it and walked off. It's current action is the worse case scenario. It's stupid :-)

  5. I bet YourFaveTelco could sell you an alarm that does even less, for more though eh? ;-)

  6. 4) Don't beep at all and don't set the alarm at all. Where a display is available, display the message "Don't be an idiot. The fire door's open. Go and close it and then try again."

  7. Why doesn't the fire-exit have a permanent alarm on it? It should never be opened except when there's a fire, so it should tell everyone when it is, whether the main alarm is set or not!
    You need a sign on it saying "This door is alarmed (and so will you be if you open it!)" :-)

  8. The Honeywell Galaxy range will beep loudly and display the open zone(s) on the keypad, but if you ignore the keypad and beeping, you'll trigger a "fail to set" alarm. This only happens if the zone is not omittable though obviously, if it can omit it, then it will do.

    As for your uPVC doors, it's a common problem, but you can now buy electronic strike systems for uPVC doors which replace the existing locks.

  9. OK, we have a galaxy, so maybe we are missing a trick.

    And as for upvc with strikes - tell!!! URL?

  10. Sounds like a training issue I think for your staff. Make sure that the system has a confirm set beep setup on it.

    Setting goes like this:
    Set the alarm, tones will indicate to leave the area with a steady beeping, if non-entry zones are active, you'll get a very fast beeping noise. The keypad will say either the specific zone that is active, or "x Zones open" while this beeping happens. Close those zones and the exit time restarts. When you've left the area, you should have a zone set as "Final" which when you close it ends the exit delay, waits four seconds with no tones, then it will emit two long tones to confirm it has been set. Some systems have a bell switch outside which is a Terminator button, which performs the same thing (useful when you have tricky locks to engage and you often open/close the final door while the alarm is still setting).
    System Parameter 35 is the Fail to Set delay. This can be programmed from 0 - 999 seconds, with the default as 360 seconds. It's advised the Fail to Set time should be approx 5 seconds longer than the exit time to allow for the setting period of a set. Each group can have it's own setting for this.
    System Parameter 47 is the Set Confirm, but it's marked as not available, so it should be turned on all the time.
    Let me know if you'd like the engineering manuals for this, I'm sure you can work out how to email me :)
    That's the link for the multi-point upvc door lock, there's also a door strike sold for doors that just have a latch (or if you just wish to use the latch locking, with the multi-point as an overlock) here:

  11. Thanks, but the beeping thing is not near the exit door for that zone and cannot be heard when setting the alarm. The setting is instant because it is set using the fob outside the final exit door so no delay is needed.

    Even for the zones where it can be heard, people are very bad at learning patterns of beeps.

    My point was that the alarm system knows the end user has been an idiot. I am not saying for one moment that staff training is needed to do things like check the fire exit and to (where possible) listen to beeps, but there will always be exceptions. This is the first time this has happened for at least a year, so the training is pretty good.

    My point is that knowing someone has been an idiot, why take the worst possible option of not setting the alarm knowing (a) that the end user thinks they set it and (b) that a door is open!

    Thankfully the monitoring means we can tell this has happened.

  12. I meant I am not saying staff training is not needed!

  13. Why not ask your alarm company to install an additional sounder near the door for the tones. If you're setting on a Max, I'm not sure if you get a confirm set beep though. Perhaps check the fail-set delay is close enough that the staff member won't have gotten to their car before the bells go off?


Comments are moderated purely to filter out obvious spam, but it means they may not show immediately.

ISO8601 is wasted

Why did we even bother? Why create ISO8601? A new API, new this year, as an industry standard, has JSON fields like this "nextAccessTim...