House alarm PIR issue

Talk Electrician Forum

Help Support Talk Electrician Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

FunkyCatUK

Member
Joined
May 28, 2021
Messages
19
Reaction score
0
Dear all,

We are having a problem with our house alarm, as follows:

On setting the alarm - Accenta G4, professionally fitted around 8 years ago - (when leaving the house), the main control panel is showing solid red lights on two of the PIR sensors ('as if' they are sensing movement in the rooms where they are located). There is no movement in those rooms. We have also looked at the actual sensors at this point and the red light on the sensors themselves is not 'on' (to indicate the seniors has picked up movement), but the red lights for those sensors on the panel remains solid on. The actual sensors themselves do seem to work as the red lights on them does come on if we move in front of them. I've changed one of the affected PIR sensors,  but this hasn't fixed the issue. I've also checked the backup battery in the actual alarm unit and it's showing good (constant 13plus volts). The issue is only resolved if we 'omit' those sensors and leave those rooms unprotected.

Any input, advice or suggestions from those forum members in the know would be much appreciated.

Thanks in advance.

 
What load did you put on the battery while you were reading the voltage?

If none, then the reading doesn't mean much.

When was the battery last replaced?

(normal average lifespan is around 5 years.)

Have you tested the actual zone wiring between panel & sensors.?

What results does a walk test give?

I would have probably swapped a couple of zones at the panel first to see if the fault stays with the wiring & sensor or something on the panel itself?

Guinness

 
It sounds to me that you have done a good job of narrowing the fault to the panel, or just maybe wiring damage.

If you have access to the engineer functions you can do a "walk test" on each sensor. 

A panel factory reset may solve something.

 
Do you have door contacts? When the door is opened the light will be on solid until the door is closed on your way out? One of the pir’s will show red if in room where keypad is, as you will be moving? 

 
What load did you put on the battery while you were reading the voltage?

If none, then the reading doesn't mean much.

When was the battery last replaced?

(normal average lifespan is around 5 years.)

Have you tested the actual zone wiring between panel & sensors.?

What results does a walk test give?

I would have probably swapped a couple of zones at the panel first to see if the fault stays with the wiring & sensor or something on the panel itself?

Guinness


Trailer Boy,

Many thanks for your reply.

Just disconnected the battery and tested in isolation, not under any load (not sure how I'd do this in the context of testing the alarm).

The wiring runs partly under plaster and floorboards, so is in most parts inaccessible.

Tried a walk through test and as soon as I enter this mode the same two sensors noted above illuminate on the panel 'as if' I had walked in front of them (the others just come on as I walk in front of them as they are supposed to in the test).

Swapping things round at the panel is beyond my level of expertise and competence I suspect.

It sounds to me that you have done a good job of narrowing the fault to the panel, or just maybe wiring damage.

If you have access to the engineer functions you can do a "walk test" on each sensor. 

A panel factory reset may solve something.


Geoff1946,

Again, thanks.

See above re walk test.

Re a panel factory reset, have checked the guide I have for the alarm and it doesn't tell me how to do this.

Do you have door contacts? When the door is opened the light will be on solid until the door is closed on your way out? One of the pir’s will show red if in room where keypad is, as you will be moving? 


Thanks, Sharpend,

No, just wired PIR sensors.

 
Just disconnected the battery and tested in isolation, not under any load (not sure how I'd do this in the context of testing the alarm).


That will not correctly test the battery..

To verify the condition of any battery you need to test it under load conditions...

The easiest way if you have no special battery testing gear,

is to measure the voltage with it connected to the panel..

and disconnect the mains supply from the panel..

See how much the voltage drops after 10, 15 & 30 minutes while also working the panel & sensors etc..

A failing battery can cause all sorts of strange fault conditions..   

You didn't say how old the battery is?

If it is older than 5 years then its due replacement anyway!.

(there may be a manufactures date code somewhere on the battery..

could be just a week no/year production date..  which may be a hint)

Damaged cables can cause all sorts of strange operation...

especially if they have been installed in the "less professional" method of tucked under the carpet between the carpet gripper & the wall!!

For the "suspect" sensors..

I would join all the conductors together at one end of the cable..

and then measure the continuity between every pair of conductors..   (they should be equal)..

Then open all the conductors, and verify there is no continuity between any pair of conductors..

I may also get one of the "suspect" sensors and wire it direct into the panel with just a short length of cable <1m..

See if that sensor actually works connected via a different cable on a different zone in the panel..

It may be that you have reached the point of needing some professional help...

As random swapping components you have not proved can be an expensive and time consuming exercise..

Where about are you?

There may be a forum member local to you who could assist at a reasonable price without ripping you off for unnecessary work.

Guinness

 
That will not correctly test the battery..

To verify the condition of any battery you need to test it under load conditions...

The easiest way if you have no special battery testing gear,

is to measure the voltage with it connected to the panel..

and disconnect the mains supply from the panel..

See how much the voltage drops after 10, 15 & 30 minutes while also working the panel & sensors etc..

A failing battery can cause all sorts of strange fault conditions..   

You didn't say how old the battery is?

If it is older than 5 years then its due replacement anyway!.

(there may be a manufactures date code somewhere on the battery..

could be just a week no/year production date..  which may be a hint)

Damaged cables can cause all sorts of strange operation...

especially if they have been installed in the "less professional" method of tucked under the carpet between the carpet gripper & the wall!!

For the "suspect" sensors..

I would join all the conductors together at one end of the cable..

and then measure the continuity between every pair of conductors..   (they should be equal)..

Then open all the conductors, and verify there is no continuity between any pair of conductors..

I may also get one of the "suspect" sensors and wire it direct into the panel with just a short length of cable <1m..

See if that sensor actually works connected via a different cable on a different zone in the panel..

It may be that you have reached the point of needing some professional help...

As random swapping components you have not proved can be an expensive and time consuming exercise..

Where about are you?

There may be a forum member local to you who could assist at a reasonable price without ripping you off for unnecessary work.

Guinness


Thanks for this.

Will try to test the battery under load as you suggest (the alarm has it's own fuse at the fusebox, so presumably flicking this off would be ok, rather than physically disconnecting?). It is 8 years old so might just need changing, as you say.

The short wire (will have to get a bit from Ebay or something) to the panel is a good idea. Again, will try to give that a go. Will look inside the panel to make sure I know where to actually connect the sensor - presumably there's no way, without special equipment, to know which connection at the panel links to the affected room PIRs (they're not 'marked up' in any way at the panel).

Your conductors suggestion seems beyond my capabilities, as I've no idea what these might look like, where they are etc. I might give it a go if I can figure out how to do this.

Suggestions much appreciated.

 
When you remove the front of the panel, the zones (pir’s etc) are along the bottom -left to right. They should correspond to the zones at the keypad, so providing you know which leds light up then you should find the corresponding zone in panel.  
there will be two cores per zone these are for activation, there will then be two cores 3/4 of way along for +ve and -ve, these will be bunched with all other zones +ve’s and -ve’s. 
there will then be two cores that form the anti-tamper circuit. These are daisy chained with all other zones anti tamper cores, usually in and out of terminal blocks. 
 

just be sure to power down and disconnect battery whilst you remove or add cables in as you can potentially short out and blow a fuse. 

 
That will not correctly test the battery..

To verify the condition of any battery you need to test it under load conditions...

The easiest way if you have no special battery testing gear,

is to measure the voltage with it connected to the panel..

and disconnect the mains supply from the panel..

See how much the voltage drops after 10, 15 & 30 minutes while also working the panel & sensors etc..

A failing battery can cause all sorts of strange fault conditions..   


Just to update. Tested the battery under load, as advised (sorry, had the panel open for the first pert of this). Times tested shown on left:

18:00 12.4v disconnected from mains

18:15 12.36v

18:25 closed panel

18:40 12.26v

18:55 12.22v

 
Just to add to the above,

If it's of any relevance, on the sensors there are just 4 wires:

Black: -0v
Red: +12v
Green and Yellow: Both alarm

 
Just to update. Tested the battery under load, as advised (sorry, had the panel open for the first pert of this). Times tested shown on left:

18:00 12.4v disconnected from mains

18:15 12.36v

18:25 closed panel

18:40 12.26v

18:55 12.22v
A bell / sounder test is the best way to load up the panel while running on battery power

 
Dear all,

Many thanks for your various comments and advice on the above.

Just to update, I've now performed the following tests, with the outlined results.

1. Changed affected PIR sensor (assuming that was the fault). Didn't fix the problem.

2. Did a walk through test (see above)

3. Tested battery away from panel (see above)

4. Tested battery under load (see above)

5. Disconnected each of the sensors from the main panel / linked out affected zones (3 and 4) one by one. In each case the panel then behaved normally (i.e. no solid red light for that sensor on the panel when I went to set it).

6. Connected a short piece of alarm cable to the corresponding sectors on the panel with one of the 'affected' PIR sensors attached) then set the alarm. All normal (i.e. no solid red light on the panel for that sector as above and the sensor on the short cable worked normally / set the alarm off when I walked in front of it (which it hadn't been doing in its original position in the house)).

7. Did the same for the other affected zone on the panel. Same normal result as above.

8. Reconnected the original PIR wiring for the affected sensors, went to set the alarm, and the same problem (solid red light for each zone).

To me as a novice this would indicate some problem with the wiring to each of those sensors. I will change the battery, as I'm not sure if the distance between the panel and the sensor can be affected by this (see the above - the sensors worked ok on a shorter length of cable), but any final input would be appreciated.

Again, many thanks for all your help.

 
Thanks for update, everything points to a damaged cable, have you had any other works done recently which involved drilling into the wall? 


No recent work. Not ruling out mice / rodents (large Victorian semi). Will change the battery in any case as this has been a repeated suggestion. If this doesn't fix it will just run some new cable to the affected sensors, trying to 'hide' as much as possible.

Cheers again.

 
Thanks for update, everything points to a damaged cable, have you had any other works done recently which involved drilling into the wall? 


Dear all,

To add to the above:

....

9. Replaced backup battery with new one.

Problems still there.

I'd assume this means damaged cables (as Sharpend suggests). Just to clarify, two of the four sensors are affected. The original alarm fit was done pre-plastering / flooring completion, so all the wires are now either buried under plaster or finished flooring (so in accessible). Aside from get a completely new wireless alarm system I guess the only way to resolve this is to run fresh cables to each of the affected PIRs. I'll try to hide the cables as much as possible, but I don't see any other way? 

One final thing. I haven't actually looked inside the alarm keypad (next to the front door). This seems a little trickier to open (no screws - juts appear to be 'clipped' shut) and I don't want to break it whilst doing so. Not sure if looking in there would gain anything.

Thanks again.

 
Cant remember, have you actually tested each cable? Done continuity check?


Thanks, Sharpend.

No. Would this involve simply putting a multi-meter on the (disconnected) green / yellow wires at the main panel end to see if they are someway creating a 'short circuit' along the wiring somewhere?

Also, am I right in thinking that if I do this and have the PIR activated it should open the PIR circuit, which'd show up as no continuity on the multi-meter (in fact, wouldn't I have to do this anyway as the circuit is by default closed on the PIR until it detects movement). I'd assume it'd be ok to cut the main power to the alarm from the battery when doing this.

Thanks.

 
To do the continuity test you will need to disconnect all power inc battery. 
remove suspect cables one at a time from panel and attached sensor. Connect two cores together at one end.

Set multimeter to ohms (resistance symbol) touch a test lead to each of the two corresponding cores at the panel. 
You have now completed the circuit with your tester and should get a test result, ideally you are looking for “0” or a continuous beep sound. This would indicate a closed circuit. Now separate the two cores at sensor end and test again at panel. 
This time you should get no sound and a high reading on your multimeter. This is a result of an open circuit. 
this concludes the test of the two cores, you can reproduce the test on any two cores to prove that the cores are continuous when joined and separate when not. 
 

whilst not definitive in itself it gives a good indication if the cable is faulty. 

 
To do the continuity test you will need to disconnect all power inc battery. 
remove suspect cables one at a time from panel and attached sensor. Connect two cores together at one end.

Set multimeter to ohms (resistance symbol) touch a test lead to each of the two corresponding cores at the panel. 
You have now completed the circuit with your tester and should get a test result, ideally you are looking for “0” or a continuous beep sound. This would indicate a closed circuit. Now separate the two cores at sensor end and test again at panel. 
This time you should get no sound and a high reading on your multimeter. This is a result of an open circuit. 
this concludes the test of the two cores, you can reproduce the test on any two cores to prove that the cores are continuous when joined and separate when not. 
 

whilst not definitive in itself it gives a good indication if the cable is faulty. 


Thanks for this.

Test(s) done.

Tested each of the zones at the panel (multimeter set to test continuity (200 Ohms)); a) with PIR not set off, b) with PIR set off (walked in front of). Bold are the problematic ones). Power on (12v).

Zone / reading not set off / reading set off (tested the 'good' PIRs just to make sure working normally showed up as working normally):

1. 28.7 / 1.

2. 28.9 / 1.

3. 1. / 1.

4. 1. / 1.

5. 30.8 / 1. 

Then twisted green / yellow wires at PIR end and tested:

3. 1.

4. 1.

So, wires to sensors it is!

Many thanks for the tips and info.

 
Top