Wire gauge on smokes

Daisy chaining thermal with thermal/smokes is fine.

This is not the case if you have tandem ring with a reversing relay. If you wire heats on the same loop as smokes and a reversing relay, you will let out the magic smoke on the power supply, or in a best case scenario, hit a PTC on the supply.
 
Where I come from 22/4 is multiconductor cable. Single conductors may or may not have jackets, but are "single". We don't do single conductor wiring around here. Maybe they do that in Chicago where conduit is required everywhere.
So again, I say the code does not require 18AWG, but I can see that the NEC reference you mentioned may cause a lot of arguments.

Like I said, it's been argued before with my state's AHJ's, and in our company's case, we had to get the NEC/NFPA involved on a specific project, so depending on your AHJ's interpretation of how it reads, however the NEC doesn't spell it out clearly with their intent. I follow the NEC's definitions for conductor and cable, and agree that the code section is ambigious, however in practicality, the more prudent thing to do would be to factor in the cable length and voltage drop, which 9/10 times leaves anything smaller than 18 AWG out of the selection.

I stopped using 22 AWG years ago when I saw a loaded loop of 4 wire detectors and the voltage drop at the end of the loop.
 
Like I said, it's been argued before with my state's AHJ's, and in our company's case, we had to get the NEC/NFPA involved on a specific project, so depending on your AHJ's interpretation of how it reads, however the NEC doesn't spell it out clearly with their intent. I follow the NEC's definitions for conductor and cable, and agree that the code section is ambigious, however in practicality, the more prudent thing to do would be to factor in the cable length and voltage drop, which 9/10 times leaves anything smaller than 18 AWG out of the selection.

I stopped using 22 AWG years ago when I saw a loaded loop of 4 wire detectors and the voltage drop at the end of the loop.

If the intent of the code is to ensure voltage within spec at all detectors then dictating heavier wire guage really isn't a great solution. It would ensure many systems would have wire gauge that was superfluous while simultaneously failing to ensure that all systems are adequate. It would make much more sense to simply state in the code, "all detectors must receive voltage within spec under max current draw".

Wire guage is only one part of the equation when determining voltage drop. You must also know length of wire, how much current each device is drawing, and where on the length of wire each device is located.
 
If the intent of the code is to ensure voltage within spec at all detectors then dictating heavier wire guage really isn't a great solution. It would ensure many systems would have wire gauge that was superfluous while simultaneously failing to ensure that all systems are adequate. It would make much more sense to simply state in the code, "all detectors must receive voltage within spec under max current draw".

Wire guage is only one part of the equation when determining voltage drop. You must also know length of wire, how much current each device is drawing, and where on the length of wire each device is located.
Having worked in a refinery years ago, I did some testing on the life of wire in a fire. It may be surprising to know that wire guage has more to do with the life than insulation does. I suspect that is why larger guage is required in many jurisdictions.
 
Having worked in a refinery years ago, I did some testing on the life of wire in a fire. It may be surprising to know that wire guage has more to do with the life than insulation does. I suspect that is why larger guage is required in many jurisdictions.

I am not at all surprised to hear that. In fact, I would suggest that you might want fire detector wire to have insulation that easily burns off. This would allow for a short which would trip the panel to fire alarm. It would turn the wire itself into a de-facto fire detector.
 
Having worked in a refinery years ago, I did some testing on the life of wire in a fire. It may be surprising to know that wire guage has more to do with the life than insulation does. I suspect that is why larger guage is required in many jurisdictions.

If you have ever seen the cable used for fire pumps you would definitely agree with this statement.
 
If the intent of the code is to ensure voltage within spec at all detectors then dictating heavier wire guage really isn't a great solution. It would ensure many systems would have wire gauge that was superfluous while simultaneously failing to ensure that all systems are adequate. It would make much more sense to simply state in the code, "all detectors must receive voltage within spec under max current draw".
Wire guage is only one part of the equation when determining voltage drop. You must also know length of wire, how much current each device is drawing, and where on the length of wire each device is located.

I didn't mention ampacity or voltage drop, specifically, and the NEC does not dictate such, NFPA does, and allows for percentages for allowable voltage drop, and simply stating that it needs to be within spec can't be easily adhered to.

I know I wouldn't want a 4 wire loop whose cable was easily burned off, because a dead short on the power side would defeat the whole system quickly.

We have high hazard areas on one of our sites that use protecta-wire specifically for this purpose, on cooling towers and the sprinkler risers and runs that surround the oil tank and day tank sections of the facility.

Gatchel- I think you're referring to the CI or MI cables?
 
I didn't mention ampacity or voltage drop, specifically, and the NEC does not dictate such, NFPA does, and allows for percentages for allowable voltage drop, and simply stating that it needs to be within spec can't be easily adhered to.
I think there-in lies the answer. It is easy to inspect. To check voltage at the last detector would require knowledge of the daisy chain order, a ladder, opening of the detector, a volt meter, and knowledge of tolerance of that unit. It would also require putting the units into alarm mode since the amps would be higher with sounders/relays turned on.
 
Is there a limit as to how many smokes you can put on a zone? or is it just the capacity of the panel (and wire run of course).
 
Is there a limit as to how many smokes you can put on a zone? or is it just the capacity of the panel (and wire run of course).
With 2-wire, yes, very much so. It will be specific to every system and the answer will be found in your system documentation.With 4-wire, it is more complex. Assuming a large enough power supply and low enough wire resistance for the power supply side, you could keep adding indefinitely. In practice, you would probably want to start splitting things up if you were installing that many. Electrically, you could split up the units onto more than one power supply (with an end of line relay for each power supply) and keep them all on the same alarm zone. I am not 100% sure that code allows that. Of course at some point your zone alarming wire becomes so long that it's resistance would make it fail to work properly. I would say, however, that for residential purposes, this is just theory since very few homes would be so large that any of this would come into play.
 
This is not the case if you have tandem ring with a reversing relay. If you wire heats on the same loop as smokes and a reversing relay, you will let out the magic smoke on the power supply, or in a best case scenario, hit a PTC on the supply.

I'm confused.

How would putting a mechanical heat detector (unpowered device that only attaches to the alarm leads) affect the powered side of a 4 wire system?
 
I'm confused.
How would putting a mechanical heat detector (unpowered device that only attaches to the alarm leads) affect the powered side of a 4 wire system?

I was referring to a 2 wire circuit. Putting a dry contact device that maintains a short after polarity is reversed is very likely to blow up an IDC or cause a power supply to trip a PTC or pop a fuse.

On a 4 wire loop, you would use a single supervision relay per zone, otherwise it won't work properly. I doubt that anyone would come close to exceeding some of the largest supplies out there (10A) for an analog fire zone, but if that's the case, normally the move is to get away from 12V and go to 24V. Most analog panels will allow up around 300 ohms of tolerance in addition to the EOLR value (and most panel EOLR tolerances are either 5 or 10%) so that works out to typically around 6.5 ohms/1K feet. If you size up to 14 or 12 to factor in a loss, the resistance will drop to 2.5 or 1.5/1K feet. Factor in splicing and similar, you're still in the MILES of cable before you would need to worry about cable resistance, however voltage drop would affect the detectors operation downstream in addition to the supervision relay.

This is all hypothetical, because honestly, if you get to a fire install that size, and honestly even up around 20 or more devices, I'd start to move towards addressable and if needed, interconnect the two.
 
I was referring to a 2 wire circuit. Putting a dry contact device that maintains a short after polarity is reversed is very likely to blow up an IDC or cause a power supply to trip a PTC or pop a fuse.
On a 4 wire loop, you would use a single supervision relay per zone, otherwise it won't work properly. I doubt that anyone would come close to exceeding some of the largest supplies out there (10A) for an analog fire zone, but if that's the case, normally the move is to get away from 12V and go to 24V. Most analog panels will allow up around 300 ohms of tolerance in addition to the EOLR value (and most panel EOLR tolerances are either 5 or 10%) so that works out to typically around 6.5 ohms/1K feet. If you size up to 14 or 12 to factor in a loss, the resistance will drop to 2.5 or 1.5/1K feet. Factor in splicing and similar, you're still in the MILES of cable before you would need to worry about cable resistance, however voltage drop would affect the detectors operation downstream in addition to the supervision relay.
This is all hypothetical, because honestly, if you get to a fire install that size, and honestly even up around 20 or more devices, I'd start to move towards addressable and if needed, interconnect the two.
Not that your going to do this, but, I think you would want an eol relay for every power supply, even if they shared a zone.  Take the theoretical case where each 4 wire detector on a zone had its own power supply.  If any one of those detector lost electricity, you would want the zone loop to open.  So you would want a relay on each unit that breaks the zone loop should power be lost at that unit.  It would pretty much be just the same as the units that break the loop for the "clean me".  But anyway, if you've got a 20,000 sf house, these are problems you might have, the rest of us don't need to worry.EDIT: So, it just ocurred to me why you wouldn't want the multi relay for power, If one unit lost power and the relay opened, it would stay open making the downstream detectors go offline.
 
Personally, I don't yet have a 20,000sf house. And If i had that much space, I would probably put a warehouse on the site :)

So I finished up wiring. I ended up leaving the 22/4 wire i had homeran to each smoke in the case that I would make it addressable in the future (not sure how that would work?) or to use with other devices if need be. I am looking forward to installing the system once drywall is all finished.
 
Personally, I don't yet have a 20,000sf house. And If i had that much space, I would probably put a warehouse on the site :)

So I finished up wiring. I ended up leaving the 22/4 wire i had homeran to each smoke in the case that I would make it addressable in the future (not sure how that would work?) or to use with other devices if need be. I am looking forward to installing the system once drywall is all finished.
With 4 wires home run to each location you can either use 4 wire detectors with each one on its own zone, or you can daisy chain 2 wire detectors onto the same zone."Addressable" refers to a specific type of detector that you would not have. These are meant for very large structures where you might have dozens or even hundreds of detectors. Each detector has a "serial number" of sorts and reports its status to a controller that identifies each one individually even though they may be on the same run of wire. I don't know the details, but it would function much like having multiple keypads on the same data line on your Elk. Whether they use rs485 or some other communication protocol I don't know.
 
Back
Top