Hi guys,
My name is Phil,i am an electrician and now and again i work on fire alarms.
I have tested and changed Batteries,and detectors that type of thing.
I work for several landlords and most of the systems are twinflex and multipoint.
I have a fault on one system at the moment that i could do with a little help.
I was getting nuisance alarms from zone 3,there is three detectors on zone 3,1 in each of the
two bedrooms and 1 on the landing.The detector in the first bedroom had its led on solid,
I have had this twice before at other houses and had to change the detector both times.
So i changed the detector.I powered down,and fitted the new detector,powered up after about 10
seconds the alarm goes off.I thought the new one may be faulty so i changed it for another one
but still same.Basically any detector put into that bedroom sets the alarm off.Its fine with no detector plugged into the base.I have also changed the base.I have tried the new detector in another location
and it works fine.There seems to be a problem just with any detector plugged into that base.
I spoke to Rafiki but they were no help really other than saying it could be noise on the cables.
The system is 8 years old and been ok till now.
Can anybody help?
Is the smoke detector compatible with the system? What is the panel?
Hi There,
Thank you for your reply,yes i am sure the detector is compatable,it has all the exact numbers on it.
I am aware that the detectors and detector sounders are identicle in appearance.I have also tried two other
detectors from other locations in the building.The panel is a CB200 4 zone.
Photoelectric or ionization type of smoke detector?
I am not familiar with this panel, may be something used only in the UK, but typically smoke detector false alarms can be attributed to environmental conditions. Too much humidity - if they get located too close to the bathroom door and someone opens the door after a hot shower it can cause a false alarm. Located too close to an air vent - you generally want to located detectors at least 3’ from any air vents and out of the path of any large movements of air. Dirty detectors - if you get a draft coming through the wiring entrances of the backbox (especially if the detector is located on the ceiling with an attic above) you can get dust and dirt from the attic space being introduced into the detector.
I would also assume this is a conventional zoned panel? I would imagine if there was “noise” on the wiring it would effect the entire zone or panel itself, not just one detector located in one room. Maybe addressable, just don’t believe it on a conventional panel. The only other thing I can think of, is maybe there is something in the room like a wifi transmitter or some other piece of equipment that is putting out a strong signal that is effecting the operation of the detector - long shot and usually not seen on wired equipment but I’ve seem fire alarm panels do some pretty weird things if you get too close to them with a 2-way radio. Maybe just unplug everything electrical in the room and see if that clears the detector.
Hi,
Thank you for your reply,the detectors are fike Rafiki multipoint detector with sounder,model no 202/0001.
The room where the detector is mounted is empty at the moment,i have tried fitting the detector with a air tight cover
over it,which made no difference.The whole system works fine with nothing plugged in that bedroom.As soon as any
one off 3 detectors that i have tried is plugged in,within a few seconds the led on the detector lights up and the alarm goes off.
That is a real mystery you’ve got there. In looking at the documentation on line this looks to be a UK and EU system. Most of us across the pond have never seen one.
I am curious what would happen if you jumped through this detector location and connected this base and detector to it’s zone at the panel.
Is there a spare zone in this panel you could switch the wiring to? If not, you could temporarily switch this zone with another to test what happens? Seems like the results will be instant.
What kind of wiring is used in this system? Twisted pair? Twisted shielded pair? Individual conductors? In conduit? Metallic? Plastic? Open cable supported by D-rings? Other?
How is this individual detector location wired? Trunk line with an out-back run? Straight shot between devices with minimal turns and bends? (I notice that some of the Fike systems are picky about out-back runs. They say NOT to use a two pair cable. Use two one pair cables.)
According to the manual this system uses an end of line capacitor. I had minimal contact with other systems but did hear about other brands using EOLC, and a EOL capacitor failing can cause weird problems. Have you substituted the EOLC?
That’s about all the ideas I have at the moment. Good luck.
One more idea. Is there any sign at all that there might have been a leak in or near this detector location?
We had a system where the SLC would shut down. The problem was solved by replacing one 15 foot run of wire. The detector that wire ran to showed signs of there being a leak.
Its possible about the leak as the bedroom has a bathroom above it.Could be worth a try changing zones,not been back over xmas.Each zone has a break glass as the last item in the zone,the break glass has dip switches in that are set for eol.The cable looks like standard cable with foil inside,not twisted pair.
So shielded pair cable. Check that the shield is connected all the way through the circuit from panel to the last device. Also be sure the shield is not grounded anyplace but at the panel. Sometimes the shield works better if connected to power supply negative rather than earth ground.
Hmmm, perplexing.
Great questions from STR-SG, and got me thinking when you said Fike. As SG said, try the device at the panel and I bet it acts normal, also due to the EOLC, which is an odd bird, is to check the shielding as SG had mentioned. Check for not only ground faults as you may have checked for, but also checked for induced AC voltages. Being in a bed room, you may have to use an extension cord plugged into a known good grounded outlet, and use this ground reference with one meter lead, and the other meter lead to shield. Also, since it’s in a bedroom, I always carry a battery operated detector to use as a temp backup until I can get strange ones figured out.
Regarding Ac voltages - I had a women’s prison that I used to service that had a large Notifier Sys 5000 that started to have intermittent zones faults, but every time we went out, it would not fault every time, or stop before we could catch what it was faulting on the zone that was for one of the dorms. I had a guy stationed at the dorm panel, and me in the control (guard) tower with the panel, and a meter (Fluke) on the zone in DC mode. It would fault, the meter would be normal, and the tech radioed that the dorm panel was normal? But it wasn’t until I switched to AC mode that I saw the 28VAC induced voltage on the zone. As it turned out, their electricians ran 120vac power for the visitor walk-thru (sally port) gate operator in our fire alarm conduit! And each time one of two guard towers, one tower with the FACP, and the other without FACP operated the gate, it would fault the zone. You just had to catch that one just right.
philipb - Something else with your system, is that you said there are some other devices on the same line, glass breaks with EOL dip switches. Some panels are current sensitive, especially with an EOLC, or EOL switches, and your bad device might actually be the device before it, or after it. Some designs use an “Identity resistor” and the resistor drifts out of tolerance, and the bad device tricks the panel into reporting an alarm from the wrong (good) device. You can replace them all day, and no luck.
We used to have a bas_ard system like this in our Papa John’s Cardinal Stadium, and we finally replaced it with a 4100U and Miniplexes after a lightning strike, thank you “mother nature”. SG, if you had any of these in your area, it was a Grinnell/Auto-Call TFX-800, and it used English based analog/digital loop boards (ten loops) and Apollo smoke and contact module devices. The only way to catch the actual bad device was to do an analog ID current report, and it would then show you the bad device’s ID current and address, and then go replace it and not the good device that the panel falsely reported the alarm from.
Philipb – does your system have the point capability to scroll through each device, and its parameters and values? If not, test each device (without your one smoke) and see if they report correctly?
Paul
Hi Paul,
Thank you for your reply,plenty more things to check,going to have a bash at it on friday.
Willl post an update then.