I work for a large fire department in Massachusets that has a problem with audio quality in its Motorola radio system. As part of my graduate degree I am required to submit a thesis so I thought I would try to solve this problem and I need help from some experts. What I am doing now is called the "literature review" (it's part of the thesis process) to see if anyone else has heard of the problem or can suggest a solution.
We use a total Motorola system consisting of a Quantar repeater, Spectra Tac receivers and HT1000 portables ("D" version) with public safety speaker mic's. The system is UHF-T band programmed wide band.
The problem starts when someone with a portable walks into a building that has a fire alarm system activated and the alarm horns are sounding. When that radio is keyed you hear an initial burst of the alarm horn in the background but as soon as the person speaks, TX audio stops. It appears that the radio is still transmitting but no audio goes out. If the person stops talking but keeps the mic keyed you get the background noise from the alarm again. What this does is make for a choppy audio situation. If the alarms are shut off aaudio returns to normal. This does not happen in every building that the alarms are going off. I have been able to narrow it down to the new style shrill alarms, the ones that sound like a sonalert on steroids. The older style lower picth alarms do not cause the problem. I realize that the obvious soultion is to shut the alarms off but from a firefighting point of view it's not that easy.
Basically I am looking to see if any other fire department has experienced this problem or if any radio tech's have received complaints. I am currently working with the departments radio shop and the Motorola account rep on this end but this is part of the writing process so I need to ask. You can respond to the group or privately to me if you wish.
Many thanks in advance.
Joe
Help with Moto system problem. Part of my thesis project
Moderator: Queue Moderator
-
rrfd43
- Posts: 434
- Joined: Wed Mar 06, 2002 4:00 pm
- What radios do you own?: Cobra 25 LTD Classic with echo
Try lloking at the speaker mics. You noted that they were "public safety". Do they have an antenna on top? Not sure if they exist. I do know that the standard ht mics starting with "d" revision have a noise canceling feature. Could this be it? We have a number of these at our fire house but have not had a problem. Check the automatic mic gain too. Good luck!
This is a wild guess, but the voting receivers normally feed a 2175 Hz status tone to the voter when no carrier is received. When a carrier is received, this tone disappears and the voter will now consider this receiver's signal as a valid signal.
Is it possible that the tone from the alarm horn looks like a status tone and upsets the voter and prevents the receiver from being voted?
Nand.
Is it possible that the tone from the alarm horn looks like a status tone and upsets the voter and prevents the receiver from being voted?
Nand.
Last edited by Nand on Sat Jan 10, 2004 7:14 am, edited 5 times in total.
-
w7com
- Posts: 434
- Joined: Tue Apr 22, 2003 11:20 pm
- What radios do you own?: MCS2000, HT1000, Pageboy II
Although this wouldn't be part of the literature review the first trouble shooting step would be to see if the same problem happens using the mic on the radio without the speaker-mic. If the problem goes away you've just reduced the problem are to the speaker-mic and eliminated the entire rest of the system (mostly.)
You do understand that by posting your question here we will expect the thesis to be published here when you're done. We're nosey bastards.
And Ooh! Nice call Nand!
You do understand that by posting your question here we will expect the thesis to be published here when you're done. We're nosey bastards.
And Ooh! Nice call Nand!
-Joe@nethead.com, Tulalip, WA
I have a private email server in my basement.
I have a private email server in my basement.
Thanks, I just had to find a way to put this in words. Not to easy tonight. It took me four (and another one) edits to add the missing characters etc.w7com wrote:
And Ooh! Nice call Nand!
Nand.
Last edited by Nand on Sat Jan 10, 2004 7:16 am, edited 3 times in total.
I would check into the radio programming (radio wide section) and see if AGC (auto. gain control) is enabled. This audio treatment will do what you are saying in certain circumstances. School is out on this feature. Some love it, some hate it. I like this feature and use it very often in loud, noisy industrial type evironments. A fireground in my opinion, meets this criteria.
My opinion is that AGC is one of the only "audio treatments" that Motorola offers that is worthwhile pursuing. All of the compression type treatments that are offered on the Pro (HT750\1250 CDM mobile) series are a waste of time and only work to distort the actual recovered audio delivered to the speaker.....(my experience)
This is how AGC works.
In a low ambient audio environment, when keyed up AGC will actually amplify the mic audio to help increase transmitter deviation. ex. A police officer is in a tactical situation and needs to whisper quietly to the dispatcher... AGC will help boost this low level input to a higher and more intelegible level. Of course, any ambient noise is exagerated right along with it. There is a pro and con to everything...
In the opposite and more common situation that you describe, very loud ambient noise is almost instantly attenuated to near zero levels. As long as the mic audio (your voice) is LOUDER than the ambient noise around you (fire alarm), the Alarm will be attenuated and your voice will be amplified.
I'm willing to bet that if your fire fighters were told to speak VERY LOUDLY into the microphone while in the noise field, your problem would improve dramatically. It sounds to me like the level that your FD members are speaking at is on the threshold of amplify\attenuate\amplify etc. This would sound like crap at the rx end. Bells, voice, bells, noise......
The HT1000 AGC takes about 2-3 seconds to ramp up amplification and same for back down again.
This is very easy to demonstrate using receiver feedback from a co-located radio on the same freq.
Turn up the volume on any radio on the same channel (does not have to be an HT1000) Now enable AGC on your HT1000 and transmit with no audio while near the receiving radios speaker (5-10 feet) Instant, loop feedback right? Now, while still transmitting and hearing the feedback, start speaking directly into the microphone. Increase the amplitude of your voice gradually and magically, the feedback will disappear. If you stop speaking, you will hear the feedback slowly come back over the next 1-2 seconds. Same trick works in reverse in a real quiet environment using the whisper technique outlined earlier....
Real world problems and solutions using and not using AGC.
Fireground:
Volunteer FD at accident scene. Lots of nice, well meaning firemen with lots of portable radios. Low ambient noise other than usual highway traffic, tool use etc. Now, someone decides they need to turn on the PA speaker (hooked to the radio receiver) so EVERYONE can hear the radio traffic. If any one tries to transmit using their portable....instant feedback, P.O'd dispatcher, well, you know the story. Using AGC, as long as the HT1000 transmitter has a voice audio input GREATER than than the PA audio, the feedback will die. You may have a quick burst at the very beggining of your transmission but otherwise nice clean audio that everone can understand. Same story for noisy machine rooms etc. in a refinery, paper mill etc.
Police patrol car:
2 member detail, 1 mobile radio in car and 2 portables on the 2 members. The passenger usually runs the radio 1-Adam 12 style while the driver, well, drives. Portable radios have to be turned waaaay down or off to prevent feedback when the mobile is keyed up. Now the passenger decides he needs to say something over the radio and he doesn't really feel like reaching into his partners lap to retrieve the mobile microphone so he reaches for his trusty HT1000 speaker mic and hits the PTT. If AGC is enabled, instant feedback. If AGC is disabled and the normal mic gain is set moderately, you can have a very nice conversation provided the mobile speaker isn't set to warp 9.
I have found that there is a time and a place for AGC. Sometimes it works for you sometimes against. Education of the radio uses is key and the feedback test works great. The guys get to play with it and can instantly see the benefits if they would only speak up a bit. It has to be understood that their voice MUST be somewhat louder than the ambient noise or they make things worse. Sometimes, when in a real loud environment, this can feel sort of unnatural. Sort of like trying to talk while wearing those headphones on your portable cd player.
Some things to note.....
On a 25\30 KHz (wideband) channel, deviation will be maxed out at around 2-3 KHz. Often this gives the illusion that the transmitter is out of tune or something is wrong with the radio.
AGC affects only transmit audio. It does not fool with RX audio levels.
When using some audio accesories other than the speaker mic, (high quality, David Clark style headsets, surveilance accessories etc.) AGC can have either a positive or derimental effect on performance. Generally, we find it helps (AGC on) to give those "noise cancelling" accessories a much needed boost because no one ever gets those boom mics close enough to their mouth when speaking.
There are 2 types of speaker mic available from Motorola regular and noise cancelling. My opinion, regular mic works great, noise cancelling ones are not worth an extra 2 cents...
AGC and all analog noise cancelling in general, is accomplished through the use of differential amplifiers. If you look carefully at an HT1000, there is a mic "hole" on the front of the housing and another one around the side, just under the spkr mic bottom "clip". Theory goes that both mics recieve the same amount of ambient noise but the mic in the front of the radio likely gets about 5-10 dB more audio than the side mic. Simple amplifier logic looks after the rest by summing and cancelling out the competing "ambient" noise while ramping up the difference, which of course, is your voice.
What I have not taken the time to figure out is, why this works so well when a spkr. mic is used. No special noise cancelling in the non-noise cancelling spkr. mic that I can see but it still works really well. Maybe both mics are actually the same and ARE noise cancelling, but Mot. puts a different part number on and sells it for more. HMMMMMM, wouldn't put it past em'....... Oh look the silent black helicopters are coming!....
I don't know for sure if this actually relates to your problem or not but it really sounds like an AGC issue to me. Good luck, and thanks for reading.
My opinion is that AGC is one of the only "audio treatments" that Motorola offers that is worthwhile pursuing. All of the compression type treatments that are offered on the Pro (HT750\1250 CDM mobile) series are a waste of time and only work to distort the actual recovered audio delivered to the speaker.....(my experience)
This is how AGC works.
In a low ambient audio environment, when keyed up AGC will actually amplify the mic audio to help increase transmitter deviation. ex. A police officer is in a tactical situation and needs to whisper quietly to the dispatcher... AGC will help boost this low level input to a higher and more intelegible level. Of course, any ambient noise is exagerated right along with it. There is a pro and con to everything...
In the opposite and more common situation that you describe, very loud ambient noise is almost instantly attenuated to near zero levels. As long as the mic audio (your voice) is LOUDER than the ambient noise around you (fire alarm), the Alarm will be attenuated and your voice will be amplified.
I'm willing to bet that if your fire fighters were told to speak VERY LOUDLY into the microphone while in the noise field, your problem would improve dramatically. It sounds to me like the level that your FD members are speaking at is on the threshold of amplify\attenuate\amplify etc. This would sound like crap at the rx end. Bells, voice, bells, noise......
The HT1000 AGC takes about 2-3 seconds to ramp up amplification and same for back down again.
This is very easy to demonstrate using receiver feedback from a co-located radio on the same freq.
Turn up the volume on any radio on the same channel (does not have to be an HT1000) Now enable AGC on your HT1000 and transmit with no audio while near the receiving radios speaker (5-10 feet) Instant, loop feedback right? Now, while still transmitting and hearing the feedback, start speaking directly into the microphone. Increase the amplitude of your voice gradually and magically, the feedback will disappear. If you stop speaking, you will hear the feedback slowly come back over the next 1-2 seconds. Same trick works in reverse in a real quiet environment using the whisper technique outlined earlier....
Real world problems and solutions using and not using AGC.
Fireground:
Volunteer FD at accident scene. Lots of nice, well meaning firemen with lots of portable radios. Low ambient noise other than usual highway traffic, tool use etc. Now, someone decides they need to turn on the PA speaker (hooked to the radio receiver) so EVERYONE can hear the radio traffic. If any one tries to transmit using their portable....instant feedback, P.O'd dispatcher, well, you know the story. Using AGC, as long as the HT1000 transmitter has a voice audio input GREATER than than the PA audio, the feedback will die. You may have a quick burst at the very beggining of your transmission but otherwise nice clean audio that everone can understand. Same story for noisy machine rooms etc. in a refinery, paper mill etc.
Police patrol car:
2 member detail, 1 mobile radio in car and 2 portables on the 2 members. The passenger usually runs the radio 1-Adam 12 style while the driver, well, drives. Portable radios have to be turned waaaay down or off to prevent feedback when the mobile is keyed up. Now the passenger decides he needs to say something over the radio and he doesn't really feel like reaching into his partners lap to retrieve the mobile microphone so he reaches for his trusty HT1000 speaker mic and hits the PTT. If AGC is enabled, instant feedback. If AGC is disabled and the normal mic gain is set moderately, you can have a very nice conversation provided the mobile speaker isn't set to warp 9.
I have found that there is a time and a place for AGC. Sometimes it works for you sometimes against. Education of the radio uses is key and the feedback test works great. The guys get to play with it and can instantly see the benefits if they would only speak up a bit. It has to be understood that their voice MUST be somewhat louder than the ambient noise or they make things worse. Sometimes, when in a real loud environment, this can feel sort of unnatural. Sort of like trying to talk while wearing those headphones on your portable cd player.
Some things to note.....
On a 25\30 KHz (wideband) channel, deviation will be maxed out at around 2-3 KHz. Often this gives the illusion that the transmitter is out of tune or something is wrong with the radio.
AGC affects only transmit audio. It does not fool with RX audio levels.
When using some audio accesories other than the speaker mic, (high quality, David Clark style headsets, surveilance accessories etc.) AGC can have either a positive or derimental effect on performance. Generally, we find it helps (AGC on) to give those "noise cancelling" accessories a much needed boost because no one ever gets those boom mics close enough to their mouth when speaking.
There are 2 types of speaker mic available from Motorola regular and noise cancelling. My opinion, regular mic works great, noise cancelling ones are not worth an extra 2 cents...
AGC and all analog noise cancelling in general, is accomplished through the use of differential amplifiers. If you look carefully at an HT1000, there is a mic "hole" on the front of the housing and another one around the side, just under the spkr mic bottom "clip". Theory goes that both mics recieve the same amount of ambient noise but the mic in the front of the radio likely gets about 5-10 dB more audio than the side mic. Simple amplifier logic looks after the rest by summing and cancelling out the competing "ambient" noise while ramping up the difference, which of course, is your voice.
What I have not taken the time to figure out is, why this works so well when a spkr. mic is used. No special noise cancelling in the non-noise cancelling spkr. mic that I can see but it still works really well. Maybe both mics are actually the same and ARE noise cancelling, but Mot. puts a different part number on and sells it for more. HMMMMMM, wouldn't put it past em'....... Oh look the silent black helicopters are coming!....
I don't know for sure if this actually relates to your problem or not but it really sounds like an AGC issue to me. Good luck, and thanks for reading.
My two bits worth:
I think that nand has the answer.
The solution is to modify the SQM modules to eliminate the AGC sample and hold circuit.
This sort of problem also occours if you have a noisy link, such as a marginal noisy RF link, or telco line.
I have modified the old Hilo fire dept comparator, as well as the Radio call IMTS because of such problems.
The AGC "pumps" up until the status tone can no longer be detected, and then the SQM (signal quality module) times out on the inactivity timer.
This was many years ago, I may have the mods some where.
The mod requires manual gain adjustment at the comparator.
Levels are very critical for proper system operation.
In the systems that I have installed I optioned for the "alternative" status tone.(1975) The problem is that in the lab there is never any "cross talk" but in the real world 2175 may "leak" into the TX audio path causing transmitter lock up.
Another solution may be to install a 2175 notch filter on the input of each SQM.
I think that nand has the answer.
The solution is to modify the SQM modules to eliminate the AGC sample and hold circuit.
This sort of problem also occours if you have a noisy link, such as a marginal noisy RF link, or telco line.
I have modified the old Hilo fire dept comparator, as well as the Radio call IMTS because of such problems.
The AGC "pumps" up until the status tone can no longer be detected, and then the SQM (signal quality module) times out on the inactivity timer.
This was many years ago, I may have the mods some where.
The mod requires manual gain adjustment at the comparator.
Levels are very critical for proper system operation.
In the systems that I have installed I optioned for the "alternative" status tone.(1975) The problem is that in the lab there is never any "cross talk" but in the real world 2175 may "leak" into the TX audio path causing transmitter lock up.
Another solution may be to install a 2175 notch filter on the input of each SQM.
Aloha, Bernie
Thanks to everyone that has responded so far.
In asnwer to some of the questions, the public safety mics are the non noise cancelling type with the antenna on the mic and the straight cord.
Early on in the investigation the portable radios were checked for AGC activation and they are set to "disable" so I think that can be ruled out.
I like the suggestion by NAND to check the 2175 status tone to see if some how it is being replicated by the TX audio going through the system.
Someone locally also suggested checking the PL to see if it is being distorted or changed somehow. That could also account for the missing audio. Only thing is that the system is using DPL but anything is possible.
I was listening to a call last night and the alarm horn was a pulsed on/off type. When the alarm was off voice audio was perfect but when it went on while the officer was talking the voice audio got distorted.
Another thought is that it could be a simple overload of the microphone audio circuit in the radio and that it is just shutting down or going into clipping.
I have built a test procedure based on all the information suggested to date and will be working with a tech in the next 2 weeks to do some bench testing of the portabales and the accessories.
For those of you who are interested in reading the thesis when it is done I will PDF it and post a message to a link where it can be downloaded but that most likely won't be until late spring.
Keep the suggestions or comments coming.
Thanks,
Joe
In asnwer to some of the questions, the public safety mics are the non noise cancelling type with the antenna on the mic and the straight cord.
Early on in the investigation the portable radios were checked for AGC activation and they are set to "disable" so I think that can be ruled out.
I like the suggestion by NAND to check the 2175 status tone to see if some how it is being replicated by the TX audio going through the system.
Someone locally also suggested checking the PL to see if it is being distorted or changed somehow. That could also account for the missing audio. Only thing is that the system is using DPL but anything is possible.
I was listening to a call last night and the alarm horn was a pulsed on/off type. When the alarm was off voice audio was perfect but when it went on while the officer was talking the voice audio got distorted.
Another thought is that it could be a simple overload of the microphone audio circuit in the radio and that it is just shutting down or going into clipping.
I have built a test procedure based on all the information suggested to date and will be working with a tech in the next 2 weeks to do some bench testing of the portabales and the accessories.
For those of you who are interested in reading the thesis when it is done I will PDF it and post a message to a link where it can be downloaded but that most likely won't be until late spring.
Keep the suggestions or comments coming.
Thanks,
Joe
A couple of simple tests may get you in the right direction.
While TX’ing on the portable with the alarm sounding in the bay, listen on another radio that listens to the portables TX signal directly. This second portable should by outside the noisy area. This will show if the voice audio is actually transmitted from the portable. You could also do this on simplex.
A second test would be to whistle at the status tone frequency and try to fool the voter. This may be a bit difficult because the tone needs to be dead on and steady at 2175 Hz. Someone needs to listen to the repeated audio when you do this. A whistle on this frequency should not go though the system.
Nand.
While TX’ing on the portable with the alarm sounding in the bay, listen on another radio that listens to the portables TX signal directly. This second portable should by outside the noisy area. This will show if the voice audio is actually transmitted from the portable. You could also do this on simplex.
A second test would be to whistle at the status tone frequency and try to fool the voter. This may be a bit difficult because the tone needs to be dead on and steady at 2175 Hz. Someone needs to listen to the repeated audio when you do this. A whistle on this frequency should not go though the system.
Nand.
Another area to check is the adjustments of the repeater system itself. As has been discussed several times here on the board, it is not unusual for the radio shop to mal adjust the repeater audio path.
In another way of putting it, it is possible that the high audio level from the fire alarm is causing the portable TX audio to go out of the pass band of the repeater receiving receivers. If your using voting receivers, the problem becomes even more complicated.
Basically you have to start at the originating source, in this case, the portable. Make sure the TX deviation is correct. In a wideband system, this will be about 4.8 Khz. max. This is when you are bellowing into the mic.
Each and every voting reciever needs to be set on frequency. The audio output level needs to be adjusted correctly. They "ALL" must be set the same. The voter comparitor audio output level that feeds the repeater needs to be checked.
Lats but not least, we get to the repeater. You don't want to have much if any compression on the audio of the repeater. In other words, if you provide 3 Khz. in, then you want to see just a little over 3 Khz. of audio out of the repeater. This is adjusted with the adudio throughput inside the repeater.
If the audio is into compression, then it will cause all sorts of problems. This will be especially noticable when several portable radios are close to each other and one tries to talk on the radio. Instant feedback is the result. Another way to tell that the audio is set too high is to listen for how much background you can hear when the radio is keyed and no one is talking.
Most dispatch mics are set with way too much gain. You can hear people talking in the background. As with many radio shops, they will take a portable right out of the box, program it and send it out. They never even measure the frequency of deviation. So part of the issue here is just how good and caring the radio shop is.
Hope this gives you something else to think about.
Jim
In another way of putting it, it is possible that the high audio level from the fire alarm is causing the portable TX audio to go out of the pass band of the repeater receiving receivers. If your using voting receivers, the problem becomes even more complicated.
Basically you have to start at the originating source, in this case, the portable. Make sure the TX deviation is correct. In a wideband system, this will be about 4.8 Khz. max. This is when you are bellowing into the mic.
Each and every voting reciever needs to be set on frequency. The audio output level needs to be adjusted correctly. They "ALL" must be set the same. The voter comparitor audio output level that feeds the repeater needs to be checked.
Lats but not least, we get to the repeater. You don't want to have much if any compression on the audio of the repeater. In other words, if you provide 3 Khz. in, then you want to see just a little over 3 Khz. of audio out of the repeater. This is adjusted with the adudio throughput inside the repeater.
If the audio is into compression, then it will cause all sorts of problems. This will be especially noticable when several portable radios are close to each other and one tries to talk on the radio. Instant feedback is the result. Another way to tell that the audio is set too high is to listen for how much background you can hear when the radio is keyed and no one is talking.
Most dispatch mics are set with way too much gain. You can hear people talking in the background. As with many radio shops, they will take a portable right out of the box, program it and send it out. They never even measure the frequency of deviation. So part of the issue here is just how good and caring the radio shop is.
Hope this gives you something else to think about.
Jim
-
aaron_tech
- Posts: 76
- Joined: Sat Sep 29, 2001 4:00 pm
Well this is an interesting problem, and some very good posts. It would be interesting to see if the mic audio amp is going into a rail type state and shutting off. This would kind of go along with what he sai in the original post the audio drops out but the system stays keyed. I maintain a system in VA that is 490Mhz with quantar, spectac w/ steering. We have almost every version of the HT1000 and we have no problems like this. If its available try a different model portable like XTS, HT1250 ect. Find out if the repeater is staying keyed when the audio is lost. If the repeater drops out Nand is the man. The spectac is being fooled by the siren. Check what stat tone level is set for ours is set for -18 dbm.
Aaron
Aaron
Motorola System Problem: Thesis Project
Just wanted to update everybody that responded about my finddings to date. I met with two tech's from the department's radio shop for a 4 hour session Monday afternoon and conducted some tests on their equipment. This is what I found:
The audio frequency of the worst offending alarm horn (Wheelock) was 3300Hz average.
That audio coming out of the portable on the repeater input frequency was slightly distorted but perfectly understandable when mixed with voice audio. It didn't make ant difference if a speaker mic was used or not. The maximum deviation was 4.5K.
A test with the dispatcher over the air confirmed that they were loosing all audio as soon as the alarm was turned on. We didn't hold the system keyd long enough to see if the repeater was dropping, a mistake on my part.
I think it is safe to say that I can eliminate the portables and speaker microphones as the cause of the problem.
This narrows it down to satellite receivers, phone lines, SQM cards or the comparitor. The dispatcher is hearing audio directly out of the comparitor.
One very interesting comment was made about the system that I was not aware of. All the phone line configurations are "packaged" circuits. One tech said that the phone company will only guarantee audio quality in the window from 400 Hz to 2800 Hz. Anything outside of that your SOL. With the alarm tone at 3300 Hz it is already exceeding the limits of the phone line and the packages bandwidth is "rolling off" sharply at 2800 so no audio is getting through. He compared it to bare copper pairs end to end that may well pass audio beyond the guaranteed range because they dont have the limiting filters or amplifiers on the line. None of the other departments with repeaters and satellite receivers they maintain are using packages on the lines.
I will be setting up another test date where we will go to one of the receiver sites and monitor audio coming out of the receiver to test this theory. Depending on what we hear we will then go to the dispatch center where the comparitor is and monitor it there, asuming it is making it onto the phone line OK. Right now though, people are leaning to a phone line issue with possible no way to solve it.
Comments invited.
Thanks,
Joe
The audio frequency of the worst offending alarm horn (Wheelock) was 3300Hz average.
That audio coming out of the portable on the repeater input frequency was slightly distorted but perfectly understandable when mixed with voice audio. It didn't make ant difference if a speaker mic was used or not. The maximum deviation was 4.5K.
A test with the dispatcher over the air confirmed that they were loosing all audio as soon as the alarm was turned on. We didn't hold the system keyd long enough to see if the repeater was dropping, a mistake on my part.
I think it is safe to say that I can eliminate the portables and speaker microphones as the cause of the problem.
This narrows it down to satellite receivers, phone lines, SQM cards or the comparitor. The dispatcher is hearing audio directly out of the comparitor.
One very interesting comment was made about the system that I was not aware of. All the phone line configurations are "packaged" circuits. One tech said that the phone company will only guarantee audio quality in the window from 400 Hz to 2800 Hz. Anything outside of that your SOL. With the alarm tone at 3300 Hz it is already exceeding the limits of the phone line and the packages bandwidth is "rolling off" sharply at 2800 so no audio is getting through. He compared it to bare copper pairs end to end that may well pass audio beyond the guaranteed range because they dont have the limiting filters or amplifiers on the line. None of the other departments with repeaters and satellite receivers they maintain are using packages on the lines.
I will be setting up another test date where we will go to one of the receiver sites and monitor audio coming out of the receiver to test this theory. Depending on what we hear we will then go to the dispatch center where the comparitor is and monitor it there, asuming it is making it onto the phone line OK. Right now though, people are leaning to a phone line issue with possible no way to solve it.
Comments invited.
Thanks,
Joe
-
jackhackett
- Posts: 1513
- Joined: Tue Jun 10, 2003 8:52 am
The original post said it appears to still be transmitting, but no audio.. I would think this would rule out the 2175 status tone, if it lost the tone it would stop transmitting, no?
Ok, read his later post where he says he didn't check to see if the repeater was dropping... so my point may be moot
Ok, read his later post where he says he didn't check to see if the repeater was dropping... so my point may be moot
Last edited by jackhackett on Tue Jan 20, 2004 1:44 pm, edited 1 time in total.
I'm a bit late to this party, but my first thought is that the siren is overloading the receivers so as to cause the PL to roll off.
Some repeater controllers have an explicit parameter for anti-rolloff of PL tones; in essence, they will hold the key for a given period if either freq or tone is lost so long as the other stays good. You might try lengthing this period.
Now that you have eliminated the transmitters (portables) as the source of the siren-following problem, the next issue is whether the remote receivers are failing. What you want to do is force a situation that causes the audio failure at the console and note which receiver is voted. Then go to that receiver, take it off line and open its local speaker, and repeat the test. If you lose the audio the same way, the problem is in the receiver, and could be PL roll-off, or something else, but it is in that receiver. If the receiver audio does not cut-out the way the console audio does, then the problem is downstream of the receiver.
Noting the high frequency of the siren (>3kHz), I wonder if what is happening is that the siren is causing the voting selector to sense noise where it should be sensing full-quieting. I had thought that the sampling done by most voters for noise was closer to 4 kHz, but offhand I can't recall where I got that value. If this is the problem, then replicate the failure sequence and pay close attention to the voting selector annunciator; what you should see is the "good" receivers suddenly voting bad.
If the problem is that the siren is flooding the portion of the audio spectrum that the voter is looking to for quiet, perhaps the solution is some form of low-pass audio filter between the receiver and the phone line.
As I said (or should have said) at the outset, from this distance all one can do is put forth some guesses, but these are my guesses.
Some repeater controllers have an explicit parameter for anti-rolloff of PL tones; in essence, they will hold the key for a given period if either freq or tone is lost so long as the other stays good. You might try lengthing this period.
Now that you have eliminated the transmitters (portables) as the source of the siren-following problem, the next issue is whether the remote receivers are failing. What you want to do is force a situation that causes the audio failure at the console and note which receiver is voted. Then go to that receiver, take it off line and open its local speaker, and repeat the test. If you lose the audio the same way, the problem is in the receiver, and could be PL roll-off, or something else, but it is in that receiver. If the receiver audio does not cut-out the way the console audio does, then the problem is downstream of the receiver.
Noting the high frequency of the siren (>3kHz), I wonder if what is happening is that the siren is causing the voting selector to sense noise where it should be sensing full-quieting. I had thought that the sampling done by most voters for noise was closer to 4 kHz, but offhand I can't recall where I got that value. If this is the problem, then replicate the failure sequence and pay close attention to the voting selector annunciator; what you should see is the "good" receivers suddenly voting bad.
If the problem is that the siren is flooding the portion of the audio spectrum that the voter is looking to for quiet, perhaps the solution is some form of low-pass audio filter between the receiver and the phone line.
As I said (or should have said) at the outset, from this distance all one can do is put forth some guesses, but these are my guesses.
-
spareparts
- Batboard $upporter
- Posts: 502
- Joined: Fri Dec 19, 2003 5:33 pm
Joe,
There are a bunch of different types of leased telco circuts. You can order them dry (will pass DC), with or without battery, with several different response curves.
The telco can & will change the physical charastics of a circut whenever they have a need to. As long as it meets the minimum level of service stated by the tarrif, ther are legally OK.
Check the system documentation for the circut order code & ask the telco what is provisioned.
There are a bunch of different types of leased telco circuts. You can order them dry (will pass DC), with or without battery, with several different response curves.
The telco can & will change the physical charastics of a circut whenever they have a need to. As long as it meets the minimum level of service stated by the tarrif, ther are legally OK.
Check the system documentation for the circut order code & ask the telco what is provisioned.
-
/\/\y 2 cents
- On Moderation
- Posts: 851
- Joined: Sun Nov 11, 2001 4:00 pm
- What radios do you own?: iPhone, Blackberry, HT220
I was thinking something as simple as over deviation, or a problem we had with a Mastr II repeater once - certain folks' voices would consistently make the PL transmitted by the repeater drop out, of course thus sounding choppy - like being over deviated - to the users' receivers.
That problem was never figured out and the repeater is no longer in service - for other reasons.
That problem was never figured out and the repeater is no longer in service - for other reasons.
Chris,
Hamming 31 years
http://www.wa2zdy.com
Wesley Chapel, Pasco County, Florida
Snow? What's that?!
The human race is proof that Darwin was wrong.
Hamming 31 years
http://www.wa2zdy.com
Wesley Chapel, Pasco County, Florida
Snow? What's that?!
The human race is proof that Darwin was wrong.
-
/\/\y 2 cents
- On Moderation
- Posts: 851
- Joined: Sun Nov 11, 2001 4:00 pm
- What radios do you own?: iPhone, Blackberry, HT220
Results Of Tests On This Project
Hello to the group again. This has been an ongoing project that just gets more complicated. Still looking for thoughts on what is happening and if anyone else has experienced this same problem of the fire alarm horns causing a no transmit condition in a repeater system with satellite receivers.
What my tests have revealed, in a nut shell, is that the portable radio (ht1000) TX deviation with the alarm horn on and voice added is 4.8K. Listening to the transmit freq reveales some distortion but not enought to be unreadable.
Once the alarm horn is turned on all the receivers in the system loose the signal at the same time and status tone returns. Monitoring the status tone reveals that it is a true clean 2175 tone and not a component of the alarm horns high frequency that is "fooling" the receivers.
Here is what is really happening. Looking at the transmit freq. of the portable with a spectrum analyzer showed the signal balloned out to about 18K wide at the half-power point with the alarm horn on. Shut the alarm off and the signal narrowed down to less than half that. Using a JT1000 we got the same results. The JT was then re-programmed to Mid band with the same results. It was then re-programmed to narrow band and this time signal was getting through and being repeated but with reduced audio and the distortion.
We also noted that the Astro receivers we infact "hearing" the incoming signal, from the wideband portable, because a remote monitor speaker plugged into the receiver was producing audio. However, the receiver would not allow it onto the phone line to get to the comparator. The audio heard out of the remote speaker included a buzzing type noise but it was thought that the DSP in the receiver might be causing that.
The last thought we had was that the high audio frequency of the alarm horn, upwards of 3600 Hz, was causing the portable radios to transmit a signal so wide that it exceded the band width of the receivers and that was why no signal was getting through. The only question we have is why was the receiver remote monitor speaker producing audio but it was not getting through to make it to the comparator?
I am still looking to see if anyone else has experienced this type of problem. I did speak to several local radio shops that work on big department systems and one tech reported to me this his FF's routinely complain about not being able to talk out of some buildings. When he goes to investigate the complaints he can't find anything wrong and he transmitts just fine out of the same buildings. I suggested the alarm horn problem to him and that it could be mimicking a dead spot syndrom inside of a building. He felt it was worth investigating but we have been unable to touch base since then.
Any thoughts, comments, or suggestions?
What my tests have revealed, in a nut shell, is that the portable radio (ht1000) TX deviation with the alarm horn on and voice added is 4.8K. Listening to the transmit freq reveales some distortion but not enought to be unreadable.
Once the alarm horn is turned on all the receivers in the system loose the signal at the same time and status tone returns. Monitoring the status tone reveals that it is a true clean 2175 tone and not a component of the alarm horns high frequency that is "fooling" the receivers.
Here is what is really happening. Looking at the transmit freq. of the portable with a spectrum analyzer showed the signal balloned out to about 18K wide at the half-power point with the alarm horn on. Shut the alarm off and the signal narrowed down to less than half that. Using a JT1000 we got the same results. The JT was then re-programmed to Mid band with the same results. It was then re-programmed to narrow band and this time signal was getting through and being repeated but with reduced audio and the distortion.
We also noted that the Astro receivers we infact "hearing" the incoming signal, from the wideband portable, because a remote monitor speaker plugged into the receiver was producing audio. However, the receiver would not allow it onto the phone line to get to the comparator. The audio heard out of the remote speaker included a buzzing type noise but it was thought that the DSP in the receiver might be causing that.
The last thought we had was that the high audio frequency of the alarm horn, upwards of 3600 Hz, was causing the portable radios to transmit a signal so wide that it exceded the band width of the receivers and that was why no signal was getting through. The only question we have is why was the receiver remote monitor speaker producing audio but it was not getting through to make it to the comparator?
I am still looking to see if anyone else has experienced this type of problem. I did speak to several local radio shops that work on big department systems and one tech reported to me this his FF's routinely complain about not being able to talk out of some buildings. When he goes to investigate the complaints he can't find anything wrong and he transmitts just fine out of the same buildings. I suggested the alarm horn problem to him and that it could be mimicking a dead spot syndrom inside of a building. He felt it was worth investigating but we have been unable to touch base since then.
Any thoughts, comments, or suggestions?
Over-deviation - plain and simple. I've seen this in some racing situations as well. The cure: throat mics or in-ear mic/speakers.
Reason: if you turn deviation down so low that the 100+dB siren doesn't cause the radio to overdeviate, you'll wind up with whispers on the repeater when people speak in a normal voice.
With the in-ear or throat options, you don't get the background noise and thus, all you're going to get is voice.
John
Reason: if you turn deviation down so low that the 100+dB siren doesn't cause the radio to overdeviate, you'll wind up with whispers on the repeater when people speak in a normal voice.
With the in-ear or throat options, you don't get the background noise and thus, all you're going to get is voice.
John
-
ricciticcitembo
- Posts: 730
- Joined: Tue Feb 26, 2002 4:00 pm
What John Said. Overdeviation because the Alarms are so damn loud.
I have a similar problem with alarm vibration causing oscillation.
I have an Alpine Alarm in My Van with a Real Mechanical 45Amp Siren.
Sometimes when the alarm is tripped, when it goes to reset itself,
it Starts Screaming AGAIN- due to the vibration from the spinning
down Siren Motor setting off the motion detector. It took me a while to figure this out.
Same culprit of a super loud alarm. Different solutions though.
I have a similar problem with alarm vibration causing oscillation.
I have an Alpine Alarm in My Van with a Real Mechanical 45Amp Siren.
Sometimes when the alarm is tripped, when it goes to reset itself,
it Starts Screaming AGAIN- due to the vibration from the spinning
down Siren Motor setting off the motion detector. It took me a while to figure this out.
Same culprit of a super loud alarm. Different solutions though.
Another possibility. Sometimes the telco crams multiple voice channels onto one wire, heterodyning channels to higher frequencies at one end, then demod-ing them at the other. This usually means some pretty steep filtering at the sending end. If the circuit you have is limited to something like 2800 Hz at the high end, anything higher than that could be inverted and actually appear BELOW 2800 Hz by the same amount that it's above 2800 Hz. In other words, a 3600 Hz tone might appear at 2000 Hz (2800 - (3600 - 2800)). Also, as others have said, the high frequency tone could be detected as noise on the receiving end, and squelching the receiver. But this should cause the same effect as losing carrier. Of course the voter will go nuts as well, but the receivers should be notching the idle tone very narrowly, so you shouldn't even notice their presence.
When a user talks with the alarm sounding, some of the alarm frequencies will still get through, so if the alarm by itself causes no problem, the addition of the voice shouldn't either. But obviously the voice IS causing different things to happen. Most of the voice energy should be below 2800 Hz and the radio system should take care of that.
This is a good problem that may require more testing and observation of the voter or comparator to see what is causing loss of audio.
Bob M.
When a user talks with the alarm sounding, some of the alarm frequencies will still get through, so if the alarm by itself causes no problem, the addition of the voice shouldn't either. But obviously the voice IS causing different things to happen. Most of the voice energy should be below 2800 Hz and the radio system should take care of that.
This is a good problem that may require more testing and observation of the voter or comparator to see what is causing loss of audio.
Bob M.
There seems to be many layers to the onion here which are getting picked off slowly.
You are smart in heading down the audio path of the portables. Now with the results from what you have posted it seems that there might be the audio problem with the high pitched noise entering the speaker/mics, or just the mic on the radio itself.
Now you also said you did try a JT1000, but my question is this: Have you tried different portables or are you using the same portable for testing over and over? It could be a portable issue with a couple of them. Yes chances are slim that it is that but you just never know.
Once you can rule out the portables 100% I would move on to the system and the rx'vers. From what you have said I ran into this same problem but not with the high pitched noise like the alarm siren.
Our problem quickly was as follows: Hwne ever an officer would talk it would be "loud and clear" for X period of time and then boom no audio at all until they unkeyed and keyed up again. Now to make it even worse this would happen with different officers in the same PD. After looking at the whole picure I found that when lower tone of voice officers would talk it would either never or very rarely ever happen to them. But when a higher pitched voice would come on I might only get 3 to 5 seconds of audio if I was lucky.
I checked my comparator and it appeared to be voting normally. Officer keys up and the site would follow them, etc... But then I noticed the pattern which was showing up when audio would go bye bye.
The system would vote site 4 and function fine until it hit that audio level and boom site 1 would get voted with site 4 and audio would disappear. This was the same case everytime audio went bye bye. So it must be a card so we swapped out cards and sent that one back to /\/\ for repair.
We tested again and the problem was not happening so it appeared ok. But no sooner did we leave and 30 mins later it was back.
So we tested levels on the backplane to see what we were getting. Now we run 9er, 9er for level settings. So avg. audio is 9 dB less then status tone. So we set the status tone level to -18 dB (as our system is a mix of telco and microwave and this seems to work) so avg audio would be -9 dB.
Now these status tone levels matched what was on the PM records from about 3 to 4 months ago. Now we locked out each site and left only one on and sent a 1004 Hz tone over the air with a distacne of no more then 1 mile from each rx site.
We found each site was in the -9 dB range except site 1 which was in the -7.5 dB range. Well we did some checking and found that the telco company put in a new copper pair from the ped out by the street to the D-Marc in the building which raised it a few dB's.
We reset the level of site 1 to -9 dB again and the problems went away.
So in conclusion our problem was since the audio level was higher the comparator would vote that site when higher levels of audio would come in and since the system was unbalanced it would loose audio due to the fact that it was trying to pull off of a site that had balanced levels for the system and one that did not.
I would watch the comparator and see if when the audio cuts out does it do it on one site only. Then I would shut off that site and test again to see if the problem moves or goes away. If it seems to always be there regardless of site that is voted I would check the levels coming in and make sure the system is balanced.
I see alot of radio companies over hear who only balance a system off of there ear and not by line level setting. These are the same people who if you ask what a TIMS unit is that say "that is gross do not talk about a guy that way."
Good luck and keep us updated.
You are smart in heading down the audio path of the portables. Now with the results from what you have posted it seems that there might be the audio problem with the high pitched noise entering the speaker/mics, or just the mic on the radio itself.
Now you also said you did try a JT1000, but my question is this: Have you tried different portables or are you using the same portable for testing over and over? It could be a portable issue with a couple of them. Yes chances are slim that it is that but you just never know.
Once you can rule out the portables 100% I would move on to the system and the rx'vers. From what you have said I ran into this same problem but not with the high pitched noise like the alarm siren.
Our problem quickly was as follows: Hwne ever an officer would talk it would be "loud and clear" for X period of time and then boom no audio at all until they unkeyed and keyed up again. Now to make it even worse this would happen with different officers in the same PD. After looking at the whole picure I found that when lower tone of voice officers would talk it would either never or very rarely ever happen to them. But when a higher pitched voice would come on I might only get 3 to 5 seconds of audio if I was lucky.
I checked my comparator and it appeared to be voting normally. Officer keys up and the site would follow them, etc... But then I noticed the pattern which was showing up when audio would go bye bye.
The system would vote site 4 and function fine until it hit that audio level and boom site 1 would get voted with site 4 and audio would disappear. This was the same case everytime audio went bye bye. So it must be a card so we swapped out cards and sent that one back to /\/\ for repair.
We tested again and the problem was not happening so it appeared ok. But no sooner did we leave and 30 mins later it was back.
So we tested levels on the backplane to see what we were getting. Now we run 9er, 9er for level settings. So avg. audio is 9 dB less then status tone. So we set the status tone level to -18 dB (as our system is a mix of telco and microwave and this seems to work) so avg audio would be -9 dB.
Now these status tone levels matched what was on the PM records from about 3 to 4 months ago. Now we locked out each site and left only one on and sent a 1004 Hz tone over the air with a distacne of no more then 1 mile from each rx site.
We found each site was in the -9 dB range except site 1 which was in the -7.5 dB range. Well we did some checking and found that the telco company put in a new copper pair from the ped out by the street to the D-Marc in the building which raised it a few dB's.
We reset the level of site 1 to -9 dB again and the problems went away.
So in conclusion our problem was since the audio level was higher the comparator would vote that site when higher levels of audio would come in and since the system was unbalanced it would loose audio due to the fact that it was trying to pull off of a site that had balanced levels for the system and one that did not.
I would watch the comparator and see if when the audio cuts out does it do it on one site only. Then I would shut off that site and test again to see if the problem moves or goes away. If it seems to always be there regardless of site that is voted I would check the levels coming in and make sure the system is balanced.
I see alot of radio companies over hear who only balance a system off of there ear and not by line level setting. These are the same people who if you ask what a TIMS unit is that say "that is gross do not talk about a guy that way."
Good luck and keep us updated.
" ah the fatman made a funny!" - Stewie from the family guy.
I went to the doctor and all he did was just suck blood. Never go to Dr Acula - M. Hedberg
I went to the doctor and all he did was just suck blood. Never go to Dr Acula - M. Hedberg
One small suggestion:
1) Motorola specifies that the steady-state 2175 Hz receiver status tone level is to be set at 13 dB below peak receiver audio (1 kHz tone level set for 4.5 kHz deviation), not 9 dB. This number comes from the AT&T telephone network standards and exists because the telcos do have separate values for the maximum peak audio level and maximum continuous steady-state audio level that their digital equipment can handle.
If you must rely on an extra 4 dB of audio compression in the outbound direction for intelligibility, do it ONLY on the comparator output line going to the transmitter, never on the receiver input lines.
1) Motorola specifies that the steady-state 2175 Hz receiver status tone level is to be set at 13 dB below peak receiver audio (1 kHz tone level set for 4.5 kHz deviation), not 9 dB. This number comes from the AT&T telephone network standards and exists because the telcos do have separate values for the maximum peak audio level and maximum continuous steady-state audio level that their digital equipment can handle.
If you must rely on an extra 4 dB of audio compression in the outbound direction for intelligibility, do it ONLY on the comparator output line going to the transmitter, never on the receiver input lines.
Karl - WA8NVW AFA5VB
SHARES + NCS
SHARES + NCS
Telco standards.
Hello.
The limits on tone within the 1700 Hz to 3000 Hz range have nothing to do with digital equipment.
The old in-band signalling system used 2600 Hz as the idle tone on RX and diconnect on TX.
1600 Hz was busy on TX and ring on RX.
There are several other tone groups, such as modem carriers, touch tone, MF and DTMF to name a few, but in-band signalling has long been replaced by SS7.
The limits on tone within the 1700 Hz to 3000 Hz range have nothing to do with digital equipment.
The old in-band signalling system used 2600 Hz as the idle tone on RX and diconnect on TX.
1600 Hz was busy on TX and ring on RX.
There are several other tone groups, such as modem carriers, touch tone, MF and DTMF to name a few, but in-band signalling has long been replaced by SS7.