NEED HELP>> Spectra Tac Voting System

[FMROB]

Hello all, quick scenario, need help.

Have 3 RX sites. One site Astro Tac RX'r through RTNA to the voter. The Second site is co located at the voter. The third site is the Quantar base station working as a RX'r / TX. This is going to the voter (which is located at HQ's) via micorwave (carlson wireless lease line emulator).


The system seems to vote nicely, however repeated audio is kinda low. I have status tone 2175 at almost -13 db throughout the system and the spoken audio is about from -10 to -8. When I try to raise the audio out of the voter it distorts the 2175 out the TX'r and wont reliably key the quantar.

So, the RTNA circuit has only 9 db loss (not to bad) and the is only about 2 db loss on the micro system. What am I doing wrong. I cant squeeze any more audio out of this thing.

The astro tacs are set to 0 db RX wireline and about - 15 on the status tone to get the -13 db status tone on the other end and about - 10 db audio. I am missing something here.

Please help, Thanks, Rob

[Batwings21]

Okay, here it goes. Maybe I am misunderstanding you, but I think you are trying to get all the 2175 status tone levels to be -13 at the voter, am I correct? If so read on, if not disregard.

1st things 1st. The -13db offset means that status tone is -13db lower than full system audio peaks. That means take a monitor, inject a 1k tone into the tac rx @ full system deviation 5khz for wide and 2.5 for narrowband. Measure and record that level. Then set the status tone to be -13 from that. Note all measurements as far as line level need to be made at the voter!!!

That's a basic alignment, but I think that will help you out. You also need to set the transmitter output of the voter to be no more than 0db for voice peaks or about -6db for 3khz deviation, then generate a 1k tone into one of the tac rx @ 3khz deviation, then adjust the transmitters tx alignment so you get the same or a little more out of the transmitter.

[FMROB]

Thank you for the reply...

Let me see if I understand this.

With our phone lines already verified to be within loss spec (less than 10db) at the RX site inject into the RX'r port 1khz tone at 2.5 kc of audio. At the voter (other end of the system) record the DB on the line and wirte this number down. Adjust the 2175 tone for -13db less than this nuber.

My question is should all of the RX audio (measurment at 1khz tone at 2.5 kc audio) be the same. We have a phone line/microwave/co located RX'r mix, so our audio paths are slightly different. Is there a majic level to aim for in DB, like -10 audio with the tone injected at the RX'r end.

My reason for asking is to try to figure out how to make all RX'rs the same audio characterisitcs and how to adjust the RX wireline and status tone settings on astro tac rx'rs and quantar Rx'r.

Hope I made sense of this. I just dont understand what audio leverl to use as a system common point and how much system voice audio shopuld be coming out of the RXrs, going thru the phone lines, and ending up at the spectra tac??

Thank you in advance for any further help... Rob

[Batwings21]

Normally you do not want any level over 0db when a phone line is involved, so normally we keep all levels below 0db. Now at the voter it doesn't really matter what the level of the status tone is, provided they are above like -33db or so. The thing that matters is the -13db offset from voice peak (1k tone @ 2.5khz full system dev for narrowband) to status tone. There are AGC circuits in the sqm's that set the amount of gain by the level of the status tone. So if you have the wrong offset, you could end up with low audio.

So to wrap it up, you could have a telco connected line with status tone coming into the voter @ - 25db, a microwave site at - 18db and a local site @ -13db. So long as the offset is all -13 it will work fine.

[FMROB]

Thank you so much brother.. I will give this a try over the weekend. Thanks, Rob

[Bill_G]

Excellent explanation bw21.

FMRob - the only thing I'm unclear on is your tx line out to the Quantar. Where is it set to? It sets the overall repeated audio level. Either it is set too low, or the module output is failed in some way to distort when you raise it up, or the Quantar input is set too low. Ideally you should have -10db going to the Quantar for 2/3 system deviation. If you inject 1k tone at 3.3k dev into the Quantar receiver (or any of the rcvrs), you should see the station xmit at 3.3k dev though most of us give it a skoche boost to make up for users with low mics and/or bad mic technique - generally 3.5k in gets 4k dev out.

Quantars are pretty forgiving about their inputs. It's hard to overdrive them since they auto-equalize. I suspect the Carlson link might be the sensitive one that wants to see a lower input level so that it's output doesn't distort which limits how much you can drive the Quantar. The Quantar doesn't have an input setting per se. That is, there is no pot you can turn to adjust the gain to set the deviation. It's done automatically based on the alignment done in RSS. If you follow the instructions on the alignment screen of the RSS, it is very easy to set the level too low. You might have to fake it out. You inject a test tone on the input and tell the RSS it is X level. If you want deviation to go up, you type into RSS an input level value that is lower than it actually is. If you want deviation to go down, you tell RSS the input is higher that it actually is. You'll have to experiment with the values to achieve the deviation you want.

while you're at it, you may also want to check the Carlson link to see what levels it will accept before distorting in both directions. They are hardy units, but are a lower tier product meant for human phone service, not as a high quality, faithful leased line replacement. Unlike high end microwave systems with telco quality channel banks that let you build a lossless system, the Carlson and products like it, need lower inputs and drive lower level outputs. They are very linear if you stay within their operating dynamic range. They get the job done at a lower cost as long as you are aware of their limitations.

[FMROB]

Bill,

Thank you for the response. The quantar (repeater audio) is set and sounds great as a standalone repeater or when we test In cabinet fall back repeat. It is only when we put audio throught the voting system and the wire line in do we get low audio.

Regarding wire line audio in to the quantar, it is a 12.5 system, so I guess audio would need to ideally be between 2-2,.5 khz of audio. So I would assume the incoming audio from the spectra tac into the quantar would still be -10db or at about 2khz of audio??

Is there a way to boost the incoming wireline signal into the quantar to make it have higher audio. I see the RX wireline adjustment, but the TX adjustment wants you to inject RF, not tone on the wireline I would assume?

Regarding the Carlson, we have had great luck with the system. Very reliable, and great response characteristics. We have a .5 db loss on the wireline circuits. The system has 3 four wire circuits and offres a flat, uniform response throughout all tones we tested, which is great.

Please let me know what you think about the above. Thanks, Rob

[Bill_G]

Rob - I'll look at one of my Quantars and get back to you about adjusting the tx wireline. My explanation above was from memory. There is a method in the RSS for adjusting the tx wireline input sensitivity to give you full deviation that matches the dynamic range of the telco circuit driving it (or Carlson link in this case). It's good you checked the radio internal repeat. That at least verifies it is capable of full deviation. Your assumptions about 12.5kc operation are correct - ideally (in a lossless system like a microwave / channel bank) you want -10db input to the station to give you 2/3 system deviation (+/-2kc). It's also good you checked the Carlson link linearity. My experience with them is they would not drive more than -10db output without distorting. That is, I couldn't get 0 out of them. Conversely, I could not put more than a -10db into them without distortion.

[Batwings21]

I would do your voter alignment, if that fixes it great. I think it will. If you still have an issue go after the quantar. If the quantar is on AGC there is no adjustment, if not then I think it tells you to inject a tone down the wireline at the level you want to equal 2/3 system deviation . You need to do the voter first to have this level.

[Bill_G]

Agreed. He needs to check the voter to ensure it is capable of sending out normal levels without distortion. It is quite possible the line driver on the command module has failed. SpectraTACs are getting old enough that the electrolytic caps are drying out and need replacement.

[FMROB]

You guys are great.. I will be there monday to try all of the stuff we discussed. I have two spare command modules to play with. I will update as to what I find.. Thanks again.

[FMROB]

Gentlemen,

Adjusting the audio and status tone seemed to go very well. We were able to get the audio in line and it is voting very nicely. However I still have an issue with the audio going out to the transmitter.

So before I go uping the TX wireline, I discovered a problem with the tone priority module. The module I have isntalled recognizes the console, and turns on the green LED on the face of the card when the console is PTT, however it still allows the SQMs to operate and does not allow the console to override. I also tried to adjust the line in and out adjustment on the priority module to no avail, it does not seem to want to cooperate or provide any definitive audio adjustments. I think that I may have a bad priority module. I am going to have to get ahold of another card.

Has anyone ever seen the card not provide console priority???

Thanks, Rob

[Bill_G]

You need JU5 and JU7 installed on the backplane. JU5 provides command module mute, and JU7 provides comparator disable. See the manual functional block diagram and the interconnect board diagram for notes and explanation. You also need JU22 in and 23 out for proper routing of tone control from command board through the priority module line level control.

[FMROB]

Bill, Which board is JU 22 and 23 located on. I will look at the block diagram to see what I missed. I was not aware of jumpers on tha back plane. More to follow, You guys have been a hugh help thus far. Thanks, Rob

[Bill_G]

All those jumpers are on the backplane. If you don't have the manual, let me know and I'll see what I can do to scan them in.

[pforbes]

Okay, here it goes. Maybe I am misunderstanding you, but I think you are trying to get all the 2175 status tone levels to be -13 at the voter, am I correct? If so read on, if not disregard.

1st things 1st. The -13db offset means that status tone is -13db lower than full system audio peaks. That means take a monitor, inject a 1k tone into the tac rx @ full system deviation 5khz for wide and 2.5 for narrowband. Measure and record that level. Then set the status tone to be -13 from that. Note all measurements as far as line level need to be made at the voter!!!

That's a basic alignment, but I think that will help you out. You also need to set the transmitter output of the voter to be no more than 0db for voice peaks or about -6db for 3khz deviation, then generate a 1k tone into one of the tac rx @ 3khz deviation, then adjust the transmitters tx alignment so you get the same or a little more out of the transmitter.

Then you need to set two receivers line out to match the highest line loss site without disturbing the -13 dB delta. This is to achieve the EXACT same level at the Comparator from all the sites. When a carrier is switching between sites you won't hear a level change.

Patrick

[Batwings21]

Then you need to set two receivers line out to match the highest line loss site without disturbing the -13 dB delta. This is to achieve the EXACT same level at the Comparator from all the sites. When a carrier is switching between sites you won't hear a level change.

Patrick

That is not needed. Thats what the the AGC in the SQM does. The offset is what prevents you from hearing the system vote. Of course at a certain point there is not enough audio for the agc to compensate, but then you have a bad line.

[pforbes]

Well, not exactly. At the moment the Status Tone is lost on ANY site, the Comparator locks the AGC of all Sites at the level of the 'Selected' Site. Should the Selected Site be switched to another, the level of THAT Site is routed to the Output, be it lower or higher. Hence the reason to set the level of all sites to the lowest level. Proven way back in the 70's and still holds true even into DigiTAC.

[Bill_G]

Not that I'm aware of. There is an agc on each sqm that sets the 2175 to about .5vac on their gated audio bus. Each sqm has about 30db dynamic range. So, each sqm can have a different level input, but as long as the differential between rx audio and status tone from each rcvr is 13db, the agc will do its job, and each sqm output will be within a db of each other. The only way it doesn't work is if the input levels change wildly which is pretty rare. OTOH, from a maintenance stand point, it is nice to have all levels equal arriving at the comparator, finding out why they are different, and resolving it. A comparator with widely different input levels generally indicates a lack of system maintenance.

[pforbes]

If I recall, the AGC line is a buss that parallels all the SQMs hence the AGC lock down during a single and continuous 'UnSq'. When one goes does it drags the others down too.

[Bill_G]

There is a common audio bus that feeds the command module, but each sqm card has it's own audio line that is gated to that bus if it is voted. They are completely independent, and there is no interaction of the agc. Or at least there better not be. I'll revisit the schematic and see if there is are possible jumpers that could get the agc's to track each other.

[FMROB]

All of the the RX'r audio levels are very close. As close as we can get it with the 9db loss of the telco RTNA circuit and the almost zero loss of the micro. A closer look at the manual and schematic confirms Bills thoughts about the SQM's being independent with the AGC. From a logical standpoint all levels should be close. In a real world situation if the AGC circuits did not exists the voter would never work due to varying line conditions (within reason). I marked all of the initial test results with a p touch on the voter so that in the future service will not be to much of an issue.

On a good note, jumpers were in fact creating the lack of console priority working. The only thing we have left to try to adjust is the TX audio out of the voter to the repeater, which is lower than I would like. I have a spare line and tone priority module (known working pulls). If I cant adjust the exisiting cards I will try the newer cards. If there is stil no way to adjust I guess the only thing I can do is increase the TX wire line deviation on the quantar?

I will update further... Thanks again, Rob