Siren Mystery

[chunkman128]

Hello, I have been scratching my head on this issue for a long time and I have had no help at all from the manufacturer of the siren. The fire department I belong to bought a new house siren from Sentry Siren in 2019. However, ever since day one, the siren has not operated "properly" since it was installed. There is a unique issue with it as follows.

If only my station is paged out for a call, the siren will work.
If my station is paged after any other station regardless of it being 2nd inline or last in a line of 5, it will not go off. This includes the weekly siren/pager test and the monthly tornado siren test.

I have contacted Sentry Siren on numerous occasions with no help, they seem to think that it is an issue on the county 911 center's side, not their system. That is not the cause because I have talked directly with the 911 director on this issue and I work there as well.

The county uses UHF for radio communications and for setting off pagers/sirens. I have messed around with the programming of the siren a few times with no real luck. The closest I got was getting the siren to go off when it should for everything, except it also would go off when another station in the county was paged out.

The programming box is a Gen 3 from Sentry Siren. See link for photos of the programming screens https://drive.google.com/drive/folders/ ... sp=sharing

Any help is greatly appreciated in trying to get this issue figured out

[nvanw27]

It sounds like an issue with your radio unit on the siren. The google drive link is not shared properly (make sure you make the link viewable by anyone with the link), but from what I've heard, the radio units that typically ship with Sentry sirens are very unreliable. Are there any other tone sets programmed into the siren, and are the tones that have been programmed properly programmed?

[chunkman128]

Here is a link that should work now https://drive.google.com/drive/folders/ ... sp=sharing

The siren is set for 2 different tones, the first one is for any call we have rather it be a mva, tree down, ect. The second one is for the weather siren to activate.

All of the frequency numbers are correct for each of the tones.

Here is the weird thing of it though, the sensitivity setting is at the highest it can go without the siren going off with another stations tones, if I go up one more, the siren will work like it should 100% but it goes off with the other stations tone. Hope the pictures help some,
Greg

[CanadianTbolt1003]

how long are each tone in a Quik call 2 / two-tone page? (A and B tones I am talking about)

[chunkman128]

.7 seconds and 3 seconds are the tone lengths for the a/b tones. I have tried those numbers and a few other ones and no matter what I change it to it does not make a difference

[OHIO Desperado]

It has to do with the tone groups and the receiver / decoder on the siren.
If your county uses tone groups properly, the first tone is the same on a number of houses and the second tone is the specific house. So what is happening is the siren hears the first tone and it matches, the second tone is wrong and it goes into a wait state to reset and go back to listening. If that reset time is too long, it will not hear the second tone group fire that IS the correct tone group for the siren and it will NOT go off. The reason that counties use this system is typically a station group is considered A /Long B two tones with the second being 3 seconds the first being 1 second. With a properly programmed radio, if a fire district has multiple houses they will share the first tone. The radios can all be programmed with a second single tone group being LONG A that will alert ALL houses in that district. So dispatch will fire that A tone for 3 seconds and it will alert ALL the fire houses in that district. Proper radio receivers from manufactures like Motorola will listen for ALL this and immediately reset if a it hears part of a tone group but not all. It will find it's tone group out of a series of tone groups regardless of what they are. Some other decoders will get confused (like siren ones) if it hears ANY part of it's programmed tone group. Remedies for this are simple, but it requires programming. If a currently unused tone groups is implemented that has NEITHER tone in the group that is used in that siren is implemented and sent in addition to your station group then the siren will consistently sound. If dispatch puts your tone group FIRST in all multi-station runs it sound. Lastly and the BEST option is use a better decoder. Get a radio (I prefer CDM Series Motorola radios) and program it with the tone group and come out of the back on the "Horn and Lights" output of the radio via a relay and use that to signal the siren to start. That will consistently work. I have run into situations with older Federal decoders that if they heard ANY DTMF they would lockup for a time which is a problem in a county full of Whelen sirens. The Federals would need triggered first and THEN the Whelen all call commands could be sent. This was ultimately fixed by putting Whelen decoders in ALL the Federal sirens and they would trip off the Whelen wildcard "all call" command.

[chunkman128]

Alright. From the Google docs link I put above that has all of the screen shots of the programming screens. Which setting or settings would I need to change? Again sentry siren don't seem like they are willing to help, and I do have the radio company that programs all of the county radios helping me now as well

[OHIO Desperado]

The change isn't there.
It has to do with the siren decoder hearing the first tone and not the second correct tone when it's a multi-agency run. Again, Federal decoders were famous for this in a mixed environment with other DTMF controlled sirens like Whelen. The Federal decoder would hear the fast DTMF from the station commanding the sirens to sound and the Federal decoder would lock up for some amount of time. One possibility is to request as a test that dispatch send tone test to one of the other stations using the 496.6 (I believe that was your A tone) and then once the dispatch has gone through fully and the repeater has dropped to send YOUR tone group and see if the siren sounds. Or, have the two way shop send the tones via a service monitor and then drop the carrier, then turn the carrier back on and send the correct tone to the siren and see if it activates. It may reset as soon as the carrier is no longer detected. This is a shot in the dark but may work. Now if that does work then you will need to ask dispatch that when dispatching a run card with you and other agencies that either YOUR tones come out first, OR that you are last dispatched and they finish the dispatch of the other agency and then tone you out after the repeater has had time to drop.

[kb4mdz]

Late to the party, but here are some other questions, just kinda random, but thinks to look at/find out.

What was the radio and controller brand and model of the old system?

Who was the company/contractor who put in the system? (I know, this may open a can of worms)

Just a thought, are there any audio adjustments between the audio output of the radio and the input to the siren controller? My point is, if audio levels coming into the decoder apparatus are too high, the circuitry may not work properly; distortion will kill decoder accuracy.

What does the 911 center say they are sending for tones, and the timing? And the audio level, aka frequency deviation? These will be important numbers.

FWIW, I once had a conversation with a mfr's engineer about tone paging, and his basic message was 2-tone paging systems are really pretty loose, in order to accommodate all the other decoder manufacturers out there. Oh, your decoder responds to different timings? No problem, we'll make it work. Because the pageout is important, not adherence to the 'stanard' (whoever has defined that).
Just a FWIW, the BW, Bandwidth, is how far from the center frequency one tone is supposed to be allowed; example, your A-tone is supposed to be 496.8 Hz, it could go as low as 484.7, Hz, or as high as 509.2 Hz, that's +/- 2.5%. This figure is also important in whether adjacent tones are within or outside of that 2.5% either side of center.

Another important figure is the audio level, called FM deviation, of the transmitter. Industry standard is supposed to be 60% of channel peak deviation. Ever since the narrowbanding deadline of Jan 2013, peck deviation has been 2.5 KHz, and 60% of that is 1.5 KHz, deviation of the tones. That's the deviation level your 911's transmitter should be reaching. Go too hot, like 2.2, 2.3, 2.4, or 2.5 KHz deviation and the resulting distortion at the receiving radio will cause poor decoding. Likewise, if the transmitter deviation of tones is small, like 0.5KHz, will cause decode problems. But again, I've seen lots of systems with enough 'slop' in them to accept a wide range of levels.

As far as timing, a realllll long time ago I heard about an (unmanned) fire station, whose tones were like this: the A-tone was actually the B-tone of one station, (lets call it station 1) and its B-tone was the A-tone of another station (we'll call it station 3). The idea was if station 1 was paged out, then station 3, Station 2 would decode the correct A-B set, even tho the tones coming in were 3 seconds of A, and 1 second of B!

Another anecdote, City of Raleigh NC, while I was working in their own radio shop; sometimes a station would report a pageout (lights turned on, speakers opened) but they weren't dispatched. My colleague and I were working in one station one day, and suddenly the lights turned on and speakers opened up, even tho that station hadn't been paged. It had picked up the B-then-A tones from 2 other stations on the other side of the city, paged in a certain sequence. We looked at each other and said "What the hell just happened???!!!" The short version was that tones for new stations had just been picked at random over the years, and we found about 6 different stations out of 27 that had these conflicts. It was just a minor annoyance to FF's, because it was comparatively rare occurance. We did end up re-jiggering the whole tone set for the entire city and got rid of all conflicts.

The final of all this, and this is my opinion only, but you need to have a combined meeting with the siren manufacturer, the radio shop and the 911 center. Perhaps some knocking of heads, but this is ultimately solvable. there's too much finger pointing. (the bane of so many problems).

[chunkman128]

I have talked to both the county and the group that does the radio programming for the 911 center.
The radio controller was a old minitor 4 pager and amplified charger. That would trip the timer mechanism to tell the motor starter when to go. It was a fed sig siren.
The siren box with the decoder and everything came pre programmed from sentry siren. And we had a electrician wire it in.