7273 and location of detectors

[Wiz]

I read it as the door must close in the event of any of the 12 conditions (which this unit could do).

I haven't heard from the manufacturers yet to answer my queries but I think they are going to say that their unit monitors 1 fault relay contact in the control panel ,and they are assuming that this relay contact operates for any or all of the 12 conditions. I personally think it is hyperbole to say the unit monitors 12 conditions but that isn't really the important issue.

However,in the event of software failure is a bit of a sticky one unless it is a totally self intelligent processor based I/O!!

I think the BS says 'software failure' but not explain exactly what they mean by this or how it should be achieved.

No I/O is infallible as I have had plenty that sit quite happy until called for and the actual relay fails - how do you compensate for that I ask??

I think it is fair to say that the most likely fault with a relay is that it won't 'turn on' at the critical time so to hold it 'on' in the normal condition and turn 'off' on 'alarm' seems a sensible 'fail-safe' option. But I agree with you that nothing is infallible.

[Allen Higginson]

I read it as the door must close in the event of any of the 12 conditions (which this unit could do).

I haven't heard from the manufacturers yet to answer my queries but I think they are going to say that their unit monitors 1 fault relay contact in the control panel ,and they are assuming that this relay contact operates for any or all of the 12 conditions. I personally think it is hyperbole to say the unit monitors 12 conditions but that isn't really the important issue.

However,in the event of software failure is a bit of a sticky one unless it is a totally self intelligent processor based I/O!!

I think the BS says 'software failure' but not explain exactly what they mean by this or how it should be achieved.

No I/O is infallible as I have had plenty that sit quite happy until called for and the actual relay fails - how do you compensate for that I ask??

I think it is fair to say that the most likely fault with a relay is that it won't 'turn on' at the critical time so to hold it 'on' in the normal condition and turn 'off' on 'alarm' seems a sensible 'fail-safe' option. But I agree with you that nothing is infallible.

I wasn't going to get involved in this because of the contradictions and personal perceptions but now that I have,well,I'm contradicting and arguing against myself now!!!
Polarised relay on a bell circuit anyone?

[Wiz]

I haven't really got time to do this but I am determined to get the confusion with BS7273-4 out in the open.

I'm not trying to score points off anyone or to make things difficult for anyone, although I am disappointed that some people who have the ability to clear this matter up seem unwilling to do so.

I'm just a simple fire alarm engineer who feels that it is not only I who is unable to understand and comply with this BS and I am just trying to sort things out for everybody's benefit.

Further to the above, I've just been reading the section regarding British Standards on the Fireco Dorgard X website at www.firecoltd.com/DGXBritishStandards.asp and have found the following:
(please note that this is only part of the full text)


BS 7273-4 concerns the interface (Critical Signal Path) between fire detection and fire alarm systems with forms of door hardware including devices to hold open self-closing fire doors.

The Critical Signal Path is the connection between the CIE and the door hardware. If the connection is compromised, the door hardware must fail-to-safe. It is important to note that for acoustically actuated systems fire alarm sounder circuits are not deemed to be part of the Critical Signal Path. There are three categories of actuation. A combination of categories can be used throughout a building according to the specific needs of the areas in question on a risk assessed basis.

I have highlighted part of the text above that confuses me and which I will discuss with Fireco when they return my call.

I would have thought that the fire alarm sounder circuits were the most important part of the csp on an acoustic system.

Can anyone else shed any light on what they mean?

[David Rooney]

7273

3.9 critical signal path

all interconnections and communications between fire alarm CIE and
the input terminals on, or within, (a) device(s) provided to open, release
or unlock a door or shutter, or between CIE and other control
equipment by which such devices are controlled

NOTE 1 Examples of other control equipment include the
control equipment of an access control system.

NOTE 2 In the case of acoustically actuated systems (see Clause 16),
fire alarm sounder circuits are not considered to be part of the critical signal path.


Does seem a bit strange to me.... except that these "acoustic" units can't be used on Cat A doors, so the "fail safe" requirements are far less onerous....

[Wiz]

7273

3.9 critical signal path

all interconnections and communications between fire alarm CIE and
the input terminals on, or within, (a) device(s) provided to open, release
or unlock a door or shutter, or between CIE and other control
equipment by which such devices are controlled

NOTE 1 Examples of other control equipment include the
control equipment of an access control system.

NOTE 2 In the case of acoustically actuated systems (see Clause 16),
fire alarm sounder circuits are not considered to be part of the critical signal path.


Does seem a bit strange to me.... except that these "acoustic" units can't be used on Cat A doors, so the "fail safe" requirements are far less onerous....

Thanks David.

Comments:

(1) Obviously the definition of csp in 7273-4 is different to that in BS5839-1 although the concept of it is similar. Pity they didn't use a new term in 7273-4. I can see it confusing people.

(2) I would have thought that in any Category system, if the sounders didn't operate then the acoustic devices wouldn't operate, and so the integrity of sounder circuits is pretty critical!

[GregC]

There you go again Wiz, using common sense, when will you see the light???

[Wiz]

I still hadn't heard from Fireco but I gave them a call a few minutes ago.

The guy I spoke to said that he wasn't technical enough to answer all of my questions but I understood from him that there is one fault input from the cie into their TX and that the panel needs to generate any of the required fault conditions to be able to signal their unit. Obviously if the control panel could only generate some of the fault conditions then only these could be sent to operate the TX. It would therefore appear that the TX itself doesn't actually monitor for all the fire system BS required conditions itself, as such (not that I'm saying it should) although the text on the link provided by Buzz earlier could be read as if it does.

The guy wasn't able to confirm any specific fire alarm control panels that were tested with their TX and that also definitely produced all the required 7273-4 fault warnings, although he said that someone more technical than him in his organisation might know of something and might call me back.

It sounds like the System X product is not the total solution to the queries being raised in this thread despite Buzz's previous highlight of a link to it hoping it would be.

[David Rooney]

So to recap, as we kind of thought, the fireco or the Salamander product just want a relay from the fire panel to switch to cause the transmitter to signal the doors to release...

So now we need a panel that either has a relay that will switch over under all conditions (not forgetting this applies to non addressable micro processor controlled panels as well) or an I/O that will switch over under all conditions.

[Wiz]

So to recap, as we kind of thought, the fireco or the Salamander product just want a relay from the fire panel to switch to cause the transmitter to signal the doors to release...

So now we need a panel that either has a relay that will switch over under all conditions (not forgetting this applies to non addressable micro processor controlled panels as well) or an I/O that will switch over under all conditions.

David, I basically agree with that recap. Of course, it doesn't have to be a single relay and a combination of different relays would still work. I think that they would all have to be NC with the relay powered on in it's 'normal' state. I assume products like the Salamander and System X are looking for a NC input opening on 'alarm/fault'

I going to email a lot of the panel manufacturers and ask if any of them have anything in their ranges that they say definitely complies with BS7273-4

[Graeme]

i,m fairly sure that all the above fault conditions would trigger the fault contact in the a'a panels we use (Advanced).

The fault relay in the panel is inverted to fail safe. I did a hotel two years ago with the supply to the door holders going through the panels fault and fire contacts and it did close the doors on the fault conditions i generated during commsioning.

They had a small leak last week and water got into a detector and all the doors shut when it showed up earth fault on the panel.

[David Rooney]

Not sure about fireco. but salamander has two inputs one for fire and one for fault... both inputs cause all the doors to close.

The inputs work when 24volts is applied, so the fire contact needs to be open, and the fault contact needs to be closed in "normal" state.

Sounds like "Advanced" may be the kiddy Graeme... just depends on this "software" failure bit, and if an I/O unit can react the same to save running more cables...

[Wiz]

Graeme & David, thanks for info regarding the Advanced. Although we really need a bit more confirmation than just 'fairly sure'.  

It has begun to worry me that none of the fire alarm panel manufacturers who can comply with the recommendations of BS7273-4, are not highlighting such in their literature. Is it because they can't? or is it because this issue hasn't become a big enough issue in their eyes yet? I'm hoping that what we are posting here might bring answers to these questions.

If we find panels with all the required fault outputs then that solves the issues for 7273-4 where interfacing is carried out at the panel.

I think that addressable loop outputs will be a bigger problem. As explained previously, not only will their relays have to be powered in the 'normal condition but I think they have to be able to respond to two other criteria:

a) To respond to a data signal or signals relating to any of the system faults required by the BS (may be possible through c&e in some software) and to a fire signal.

b) Loss of data signals

The first is probably easy to achieve.

I can only see the second being achieved by a method where the addressable interface is made so that it is expecting to see a regular 'O.K.?' data signal and operating whenever it can no longer see it.

[Wiz]

Not sure about fireco. but salamander has two inputs one for fire and one for fault... both inputs cause all the doors to close.

The inputs work when 24volts is applied, so the fire contact needs to be open, and the fault contact needs to be closed in "normal" state.

Sounds like "Advanced" may be the kiddy Graeme... just depends on this "software" failure bit, and if an I/O unit can react the same to save running more cables...

David, Is the bit I've highlighted above in accordance with BS?

[David Rooney]

Yes, I can't see anyreason why not.


Clause 10 note.
For Category A actuation, it is unlikely that a system that relies on operation of one or more relays to actuate a release mechanism will
conform to 5.1.1e) unless the coil of at least one relay is normally energized; in the case of a normally de-energized coil, failure of both the
normal and standby power supply to the fire detection and fire alarm system precludes changeover of the relay contacts in the event of fire.

[Wiz]

Yes, I can't see anyreason why not.


Clause 10 note.
For Category A actuation, it is unlikely that a system that relies on operation of one or more relays to actuate a release mechanism will
conform to 5.1.1e) unless the coil of at least one relay is normally energized; in the case of a normally de-energized coil, failure of both the
normal and standby power supply to the fire detection and fire alarm system precludes changeover of the relay contacts in the event of fire.

For years people have been saying that all relays used in this sort of interface must be powered on in the 'normal' state because failure of any normally 'unenergised' coil means that the system is not fail-safe.

I personally have always thought this to be overkill but believed that this requirement was now incorporated in this standard (as I've always said I don't profess to understand it all). But I agree the text you highlight could be read that it is not. But I am confused by the text you have highlighted. It doesn't make 100% sense to me. What do others think?

The drawing attachment shows both relays as NC contacts (and not a NO as you said one particular product needed!) The  fire relay contacts are described as 'normally closed' and the fault relay contacts are described as 'held closed' so i read this as being that the 'fault' relay at least is 'normally energised'. I wonder why they consider the fault relay to be more important than the fire relay? I would have thought that the fire condition is always important and that very few fault conditions would have much as much impact.

As I've always said, this BS is very confusing.