Alarm triggered by switching a light on.

[aptsys]

Based on your description, it sounds like there is a run of alarm cable parallel to the lighting circuit. If it's a basic alarm panel with a global tamper circuit, you'll see a spike coupled in when something like a fluorescent tube strikes as the panel has high impedance inputs. One solution would be to get a ferrite core and wind a few turns of two tamper wires going into the tamper circuit around the core. The windings must be in the same direction and same number of turns and this'll give you some common mode rejection (which is what you'll see from the parallel run). Otherwise you can probably use that off-the-shelf suppressor, or even better would be to move the alarm cable away from the mains circuit. It's highly unlikely this is anything to do with the AC supply to the panel as it would not meet EMC regulations if this was the case and there is heavy capacitance around the voltage regulators anyway, so HF spikes won't make their way through.

 

[winston1]

1 trace the feed of original wire back to the lighting ring and see where it runs and looks for interference.

There is no lighting ring. Lights are not wired in that way.

 

[shirley83]

There is no lighting ring. Lights are not wired in that way.

Agreed, I used the wrong phase. Ultimately, the 240v source for the alarm comes from the lighting circuit.

 

[sparkymarka]

So what reading did you get from the panel earth and 0v on the panel ? Meter set on AC low range if it's not auto ranging one probe on incoming Earth the other probe on 0v on the panel ?

 

[shirley83]

So what reading did you get from the panel earth and 0v on the panel ? Meter set on AC low range if it's not auto ranging one probe on incoming Earth the other probe on 0v on the panel ?

Hi Mark

If I did it correctly I got 36v. I've have attached a picture, I put one probe (black) on the earth at the connector and the other probe on the 0v. I've put some red circles on picture of where I measured and attached a picture of multi meter output. The Alarm is now back on its normal spur from the lighting circuit. Also which the panel cover off, I'm noticing how much of a buzz the alarm transformer is giving off.

http://s1293.photobucket.com/user/ashhammond1983/media/AC_zpssobvzx54.jpg.html

 

[sparkymarka]

That's too high for a start ....induced AC fit a act 1313 for a start , transformer buzz is common , try nipping up the screws that hold it in (4 of )

 

[shirley83]

Had I put the probes in the correct place?

As a quick test, I have just rewired the alarm to a socket via a plug to test what AC is introduced via that route. The reading still came out to high 28v. Does this indicate a general problem with the house wiring?

I've order a filter as a matter of course, £10 is not bad if that is the fix.

Thanks Again.

 

[aptsys]

Had I put the probes in the correct place?

As a quick test, I have just rewired the alarm to a socket via a plug to test what AC is introduced via that route. The reading still came out to high 28v. Does this indicate a general problem with the house wiring?

I've order a filter as a matter of course, £10 is not bad if that is the fix.

Thanks Again.

No, the reading is completely meaningless. You have a transformer based power supply in that panel, so the secondary side is completely floating with respect to mains earth.

 

[shirley83]

No, the reading is completely meaningless. You have a transformer based power supply in that panel, so the secondary side is completely floating with respect to mains earth.

Are you saying that because I have a transformer based supply the introduced AC test wont prove anything of any value?

From the previous post, I took earth to be earth in the supply wire and 0 volts to be the earth of the DC 12v system. Can neutral also be called 0 volts, if so do the probes go across earth and neutral.

 

[aptsys]

Are you saying that because I have a transformer based supply the introduced AC test wont prove anything of any value?

From the previous post, I took earth to be earth in the supply wire and 0 volts to be the earth of the DC 12v system. Can neutral also be called 0 volts, if so do the probes go across earth and neutral.

Yes, the whole idea of a transformer is it provides isolation as well as transforming the voltage. The low voltage side of the transformer is floating, and that particular panel has no secondary side connection to mains earth so the AC test means nothing.

What you may find is that the ESR on that big blue capacitor above the battery terminals has started to rise given the panel is around 15 years old. If you're able to solder it would be worth replacing that. Remove AC from the panel so it runs on the battery, and try replicating the fault. This'll highlight if that cap is at fault.

If it still triggers, then I'd look for the alarm circuit that is triggering the tamper. You could disconnect the global tamper (black and blue in your picture) and connect each zone in turn until you find the zone that is at fault. If it's repeatable then it's differential mode noise and a choke will likely fix the false alarm as explained in my earlier post.

 

[shirley83]

I feel this is my longest ever forum thread and I maybe making progress. Induced AC of 36v was found at the panel so a 1313 was fitted and the got rid of the induced AC along with a 2323 mains filter for good measure. The problem still persisted. I contacted ACT meters as I got my filters from there and they have been a great support. After unwiring all the zones and rewiring them all individually. It looks like the back door key pad is the issue. As a light switch is very close it and is on the same circuit as the suspect lights. This key pad is currently unwired and we are waiting to see if it does not drigger over the next weeks. It might be a case of running a shielded cable and adjusting the position of the keypad to stop any spikes on that cable. We are now only waiting on this issue before we decorate so it has her in doors full attention now.

 

[bernardgreen]

The voltage between a floating 0 volts and Ground is in almost all cases a capacitively induced voltage. The impedance of the meter used to measure it will affect the voltage as a lower impedance meter will drain to ground more of the limited current available via the capacitive coupling.

Using the same meter but with a 10 K ohm ( 10,000 ohm ) resistor between 0 volts and ground will give a lower reading.

If the floating ciruit is fully isolated then the voltage 0 volts to ground could be much higher than 36 volts (*) when there is no meter draining current to ground. If the ciruit is completely isolated then referencing it's 0 volts to ground will often solve the problem. A 1000 ohm ( 1K ohm ) resistor between 0 volts and ground is a good starting point.

If the Live end of the primary winding in the transformer is closest to the secondary winding then the secondary winding and everything connected to it can be pulled up to more than 200 volts AC by capacitive coupling between windings in the transformer. The coupling capacity is very small and this limits the coupled current to a few micro-amps at most. Not enough to be a hazard but enough to produce a voltage on a meter. The voltage indicaed will be less than the voltage when the meter is not connected Exchanging the Live and Neutral connection to the transformer can remove that source of capacitive coupled voltage.