Monday, March 10, 2014

Problem In Interconnecting New Smoke Alarms



Hello:
Have really been going around the circle with this, and am truly stumped. Thought I'd try here, as I imagine there are many true experts on this forum. Would be most grateful for any help or thoughts. Moved into a 30 year old house recently, and decided to replace the old smoke detectors with new wired in to the house power types (as were the original old ones). Want to use the interconnect feature where if one goes off, they all do; old units were apparently wired this way also.
This relates to the problem I've been having in getting these 3 new Kidde PI 2000 (wired in) smoke dectors to talk to each other. They replaced 3 very old (probably about 30 yrs old !) detectors by a Rittenhouse Company which were apparently wired together also. When my new ones were hooked up to the third wire (red) they all just cycle on and off.
So, I guess there are two main possibilities:
a. that there is perhaps another detector on the line somewhere that I am not aware of. But, I really looked all over, including the attic, and cannot find any other. Still, it is just about impossible to trace the wires, and maybe there is one that was plastered over, or... ? 30 yr old house, so anything is possible.
b. I disconnected the red wires from Each detector to the red interconnect wire interconnecting the 3 new detectors, and measured the voltage appearing between the interconnect wire and the white neutral.
Couldn't believe it. Seems there is about 2 volts AC appearing.
Of all the weird voltages, this is a tough one.
It certainly isn't inductive coupling, but might be leakage from something
Was wondering if any one might have any thoughts on this.
What uses 2 V AC ?
Sure has me stumped. Becoming an intellectual challenge by now.
Any thoughts or suggestions regarding this 2 V AC, or the problem in general would sure be appreciated. Would really like to interconnect them, if I can only figure out what is preventing it.
BTW: don't want to complicate my questions any, but let me offer a little more information on an experiment I performed. i took one of the old units to the bench, hooked it up by itself, and was surprised to see that there was 110 V AC on the white neutral to the unit's interconnect line while the unit was, of course, not triggered.
Might this shed any light on how the house wired for the original unit's interconnects might have been configured ?
Sure was surprised to see 110 V there; can't imagine how it could had worked, or been configured originally.
Perhaps some end of line resistor or relay hidden away, somewhere (there's no control panel anywhere).
Much thanks,
Bob

Put away your digital voltmeter and do not use it for ac testing. The voltages you are reading are not really there. Continue looking for another detector, and verify the wiring you have is what you think it is.

Racraft,
I've been around electricity for a very long time, but I am really stumped by your statement Put away your digital voltmeter and do not use it for ac testing. The voltages you are reading are not really there. Could you explain this to me? Until about 9 years ago, I used almost exclusively analog meters. Since I obtained a Sperry DMM I have not found testing results to be any different, testing the same sorts of things and using the same methods I always used. I'm puzzled by your advice. (Not doubting you or criticizing, I just want to understand.)
Robert111,
An end-of-line resister is typically found on a Class B fire alarm system, using 24 volts for the detector circuit, loop-powered from a fire alarm control panel. This is an industrial/commercial application. For hard-wired detectors in a residential situation, I have not seen them used.
I'm sure you have a wiring diagram with your packaging and have reviewed it. I went to their website and observed the data sheet at http://www.kiddeus.com/ssPI2000.rsf. I checked the wiring diagram and observed that detectors in the middle of the line must have the red wire pigtailed (parallel) and not fed-through (series). Not sure if this was followed in your installation.
I assume that you are assuming the route of the wire runs which are concealed in your ceilings and walls. You sort of have to sometimes. If, for example, you have a black/white/red set sticking out of box A, and you believe that the black/white/red set sticking out of box B is the other end of these wires, you must know that they could be - or not. To test you would first turn the power off, then twist a black and red together. Now go to what you assume is the other end of this set of wires and do a continuity between the red and black. If you have continuity, go back to the other end and separate the two you had twisted together and test again. If you now have no continuity, you have confirmed that those are two ends of one continuous section of wiring, so you know where they come from and where they go to. Repeat this procedure along the entire circuit until you have verified that the wiring is as you believed it is. Also, if you get a resistance that is not infinity and is not a near-short (as you found when testing wires twisted together), then you have found a portion of the circuit that may well have a hidden resistor or detector somewhere.
I sure hope that helps.
Juice

Juicehead,
I believe what Racraft is referring to is the tendency for DVM's to show what seems to be voltage on an otherwise unpowered wire. Compared to your average analog meter, the input impedance of a DVM is entremely high. This means that the DVM is almost invisible to the circuit it's measuring. Any piece of wire of sufficent length can be slightly charged by the normal electromagnetic interference in the air. This interference impresses a slight AC signal on the wire. The DVM, presenting virtually no load at all to the wire, reads these stray voltages and gives a low reading. Your typical analog meter presents enough of a load that the miniscule current in the wire is drained and the meter shows virtually no voltage present. Granted there are better quality analog meters with FET inputs and such that also have high impedance inputs, but they are probably less commonly found in your average toolbox.
My guess is that a wire is improperly connected somewhere and there is actually no power at all between the two points you are measuring.
Apologies in advance to Racraft if I've misinterpreted what you meant....

Are all the smoke detectors powered From the same line ?
If not, that may set them off.

I replaced a Kidde brand smoke detector this week that was simply bad. Have you tried hooking up the smokes without the interconnect wire?

Juice,
Bubbamill has addressed the issue, as have other posts, so I am not sure that I can say much more. When power is actually present, I have found a digital voltmeter to work. It will show 120 volts or 240 volts. However, when an open wire exists, the voltmeter rends to show phantom voltage. This phantom voltage is not really there. The open wire could be an open neutral, or an open hot, or simply an open wire, such as a switched wire with the switch turned off or a smoke detector interconnect that isn't powered because no alarm exists.

Thanks to Racraft and Bubbamill for your explanations. I have not encountered problems with this on de-energized wiring, but if I ever do I will now be prepared with this knowledge to interpret the readings as false, and pull out the old analog meter.
Still curious as to what Robert111 has found, or what further troubleshooting was done or may have revealed.
Juice






Tags: problem, interconnecting, smoke, analog meter, digital voltmeter, have found, interconnect wire, open wire, really there, with this