System sensor 4-wire smoke with built-in relay

Ira

Active Member
I bought a couple of System Sensor 4WTA-B smokes to get familiar with, i.e., desk check 'em. Understanding that wiring them as separate zones creates issues, I've decided to daisy chain them to a single Elk M1G zone and forego the ability for the panel to tell which smoke tripped first.
 
However, it looks like if I use 4WTAR-B (built-in relay) smokes, I can connect each smoke's relay to an M1G zone to detect which smoke tripped first. Am I reading this correctly, or is the relay for some other purpose? If I'm correct, does this cause any other issues with regard to reset, etc.? I know it's additional wiring for the relay home runs, but they are all less than 20', with easy access.
 
Thanks,
Ira
 
The relay is intended to release a self closing door to help compartmentalize the fire.  You can use it however you wish!
 
I'm using HAI Omni Pro II, but I don't know why it would be different on the Elk.
I have 7 smokes all on different zones.  Plus heat sensors in several locations, attic, garage, shop, garden shed, all on separate zones.
I'm using the GE smokes.
I use the relays of the smokes as the inputs to the zones.
They all sound when any of them detects fire, but the relays operate independently and don't close unless the detector itself is tripped.
That helps identify the location fo the fire.

I also use a smoke in the equipment room whose relay does trip whenever there is an alarm on any other detector. 
I use this to verify a fire, since the signal to activate this detector is sent down the communication wire. 
If the relay failed or the wiring got cut to one of the room detectors, the verify detector wouldn't be tripped.
 
GE 350CC has the relay that does not operate from signals on the comms line, it must detect smoke to operate the relay.  I have these in my rooms and hallways.
 
GE 350CX has the relay that operates when a signal is transmitted on the comms line. This is the one in my equipment room.
 
The both have both form A and C relays
120VAC with 9V battery backup
 
The 350 is not intended for connection to an alarm panel. Not listed for purpose. Tread very lightly.
 
Read the second paragraph of the installation documents. Very clear that they are not to be connected in this manner.
 
I have the system sensor units and have one relay unit to shut off my HVAC fan if there is an event. Is the Red led light not enough to tell you which one tripped?
 
The red LED is good enough, but for the small additional cost for the smokes with relays, it seems like a relatively cheap "additional feature" (as long as the M1G zones are available).
 
I think you need to pay attention to the ZT's here IRA.....you wouldn't be able to use FA, because of the reset issue. It's going to be a real balancing act here to choose a ZT that performs as you wish but doesn't have anything to do with the FA....that is unless you put a few rules and text into the M1 that is triggered when it hits an input. I'd put it as a latching type with a * to clear, but that's me.
 
DELInstallations said:
I think you need to pay attention to the ZT's here IRA.....you wouldn't be able to use FA, because of the reset issue. It's going to be a real balancing act here to choose a ZT that performs as you wish but doesn't have anything to do with the FA....that is unless you put a few rules and text into the M1 that is triggered when it hits an input. I'd put it as a latching type with a * to clear, but that's me.
Not sure I understand. My plan was to daisy chain the smokes "normal connections" to a FA zone, then connect each of the smokes' relays to its own M1G zone, defined as a non-alarm zone. Then use either an Elk rule or CQC to trigger when the zone becomes not secure to generate a notification somewhere (iPhone, iPad, keypad, etc.) denoting which smoke caused the trigger. If I want to send the notification every x seconds/minutes, I would probably turn on a phantom output and start checking it at the desired interval. Is that not possible? Does the M1G going into FA state at the same time cause other events to be held up or not triggered?
 
The System Sensor i3 manual says "The detector that initiated the alarm condition will have its red LED and Form C relays (if applicable) latched until reset by panel." Am I wrong in thinking that means when the alarm condition is reset by the panel, the form C relay is also reset, so I could use the zone going back into secure state to do any cleanup work needed, e.g., turn off the phantom output?
 
Thanks,
Ira
 
I was referring to where you were driving the input from the smoke. Assuming the M1, you wouldn't be able to use most ZT's, so then you're looking at driving it via automation rules and possibly text or pushing the notifications out elsewhere....it was unclear that other peripherals were being connected to the M1.
 
I'm not sure if the FA would prioritize rules or cause them to stop running. I'd probably make a call to Elk and propose how I was driving the inputs and a sample rule and see how that interacts with the logic and hard coded panel events.
 
@ Desert,
 
The docs don't state that. Any system, once it has a host panel involved, meets the first line item "system with control panels". The rest is boilerplate nomenclature. To be more specific, if an AHJ was diligent and looking at the specifics combined with your statement, they could actually pull the UL listing out and see what it covers and if it is acceptable in such an installation, however the installation instructions state exactly where the units should not be installed. These are smoke alarms and not smoke detectors. I've done this dance before with AHJ's, FM and UL and "creative" interpretation. These are not detectors intended for monitoring purposes. If you'd like to know specific design criteria of when and where these units are legally allowed to be installed and connected, shoot me a PM.
 
The specifics here are that the instructions do not state the units are only acceptable in single family residential applications and not commercial or narrow the scope to such an installation. The verbiage is not there. Remember, part of the largest item with both electrical code and fire alarm code...literally one of the first lines is all products must be installed in accordance with manufacturer's installation instructions.
 
These style of smoke detectors are not listed for connection to any alarm panel, even for ancillary notification purposes unless the AHJ specifically allows it or recognizes it, which they generally don't. There's usually a specific criteria when they will allow this style of unit to be installed with a control panel and it's not for fire alarm notification, it's for monitoring purposes on a required system and not for the purpose of evacuation or notification.
 
That said, what you choose to do in your own home, if that makes you feel comfortable, so be it, but in the event of dealing with AHJ's, legalities and what could also be construed as negligence by an insurer, that is valid information. Whether or not an insurer chooses to negate payment on a loss, that's another conversation, but when you're talking money out of an insurer's pocket for a large claim...you bet they'll look at it. I've had to testify in court regarding cases and installations such as your own and it's black and white as to what is allowable and what is not, which then becomes a question for a judge to consider in the bigger picture.
 
Desert,
 
Rather than beat the horse....UL listing is only part of the equation. The listing process, which you seem vaguely familiar with, involves the testing of the equipment to a known configuration and that it functions as intended. That also means that the connection in the reverse also "does no harm". In the case of what is allowable per NEC and NFPA, that is simple...the devices must be listed as acceptable for their installation, so the fact that they are installed against their installation documents is cause enough for them to not be compliant.
 
The residential application and listing, as you are reading it, is not the application you are installing it nor how these units are intended to be installed. I'll just state I know the UL listing on the device and what it's design criteria is and how they function and how they're intended on functioning.
 
These units and others with the same listing are not intended for fire door release...never have and never would be. They could, if they were in the appropriate application (generally OEM) be used for AHU shutdown or in the case of these units, they are not intended for general alarm purposes...connection to ancillary fire protective equipment or required systems. They could, however, be connected to strobes or other similar devices in the application of ADA or similar requirements.
 
The listing on these units applies to an installation in a application similar to a dorm, hotel, or similar where there is already a required system and these units would be intended to notify someone remotely but not dump the building....they are allowed in a software configurable FACP application for monitoring purposes, but not alarm purposes...if that makes sense to you. Their failure to operate or sound does not negate a required fire system and they still function as necessary for code (IE: interconnected living space and sleeping area in a dorm/hotel application).
 
You may not like the answer and it may not be as compact and sweet as you believe, but they are not listed nor designed for connection to an alarm panel no matter the location for alarm purposes.
 
Contrary to popular belief, nowhere in code is it specified that all smoke alarms and detection systems in residential MUST be 120VAC, interconnected and battery backed up...nowhere. That would only be a local restrictive covenant and if you can provide it in all the national codes, I'd love to know the article, section and item it is. Again, I've done the dance with AHJ's on this before....the only reason why some locales forbid LV units only is due to their inherent reliance on an alarm panel and the AHJ showing concern of what happens if the panel is down, unplugged, what have you....same reason why some AHJ's want the EC to wire the smoke alarms with a circuit with other general light and power devices instead of on their own....no code requirement, only the belief it makes it more difficult for someone to negate the smoke detectors.
 
Update...
 
I bench tested everything to make sure I understood how the devices work, and for the most part, it works as expected. For the testing, I had...
 
1. Two System Sensor 4WTAR-B smokes daisy chained to the M1G Z1 (Fire Alarm).
2. The EOL relay and EOL 2200 ohm resistor.
3. The System Sensor RRS-MOD reversing relay with the purple wire connected to output 3.
4. RRS-MOD yellow and orange wires (detector power) connected to SAUX power.
5. RRS-MOD red and black wires (RRS-MOD power?) connected to VAUX power (may move to a different PS later for final install).
6. Everything wired according to Figure 4 in RRS-MOD installation instructions.
7. A rule that turns on output 3 whenever any fire alarm turns on.
8. A rule that turns off output 3 whenever any fire alarm cuts off.
9. The NC relay for the first smoke connected to Z15 (non-alarm, normally closed).
10. The NC relay for the second smoke connected to Z16 (non-alarm, normally closed).
11. A rule that writes a message ("1st smoke detect") to all keypads when Z15 becomes not secure.
12. A rule that writes a message ("2nd smoke detect") to all keypads when Z16 becomes not secure.
 
Everything pretty much works as expected when a smoke triggers. The associated relay's zone triggers, which causes the associated rule to run. What I didn't know is whether or not the associated message would go to the keypad. Based on my limited testing, it looks like the keypad message area is "locked" and only displays the standard fire alarm message until the alarm is silenced. At that time, the messages start going thru the normal round-robin, and the triggered smoke's relay message is displayed on the keypad in sequence with the others. It makes sense that Elk would lock out the keypad messages until the fire alarm is silenced, so I'm happy with the results thus far.
 
Back
Top