Linking a quanta to a digitac voter opinions

[911-EMT]

I would like to get opinions on linking quanta to a digitac comparator via dsl I was looking on the nxu2’s and on the audiomate 360r
About the system : one quantar at the digitac set up for in cabinet fall back serves as the transmitter, two quantars to the west and east setup as auxiliary receivers with 2 wire radio tie lines (vg1) to the digitac
radio tie lines are always down and $$$. ptp/microwave is out of budget

[Bill_G]

The internet is not stable enough to support reliable status tone operation. You would have to use hardware receive indicate (COR). The Digitac doesn't boogey COR as far as I know. Someone would have to consult a manual. The next problem is latency. The Digitac does not have any delay capabilities. So, your local receiver would always arrive first followed by the satellite rcvrs many, many milliseconds later causing out of sync audio weirdness (unit fo-t four t-se. I am te- am ten ei- t ten nineteen. En- ending mila- is two poi- r). Finally, the Digitac only has rudimentary equalization. The NXU has a grand algorithm to reduce Raliegh fade and impulse noise. Does a great job, but leaves the voter with little noise to work with meaning you have to pump the high freq content up as much as you can on the satellite rcvrs while knocking it down on the local rcvr. Good times. The Digitac is not up for the task.

On the other hand, if you put an NXU in all the audio paths, including the local one, and you turned the rx buffers to max, and your dsl service is top notch, you might be able to get away with using status tone rx indicate, and all the inputs would have the same low noise content leaving just the audio delay problem to deal with. I would expect poor voter operation, but it might work.

[d119]

I'm with Bill on this. In a proper voting arrangement, all of your links should come over the same transmission medium.

You're going to have a difficult time with this with regard to stability and latency, as Bill said.

We had a customer who had two different types of microwave up for a short while, and we couldn't get all the audio to arrive at the same time over that system.

[FMROB]

I have been sucessful at mixing micrwave and telco in voting systems without any latency issues. The microwave (carlson wireless lease line emulator) has a built in mux, and works well for the radio and voting application. I could see however, if different types of muxes and/or IP based micorwave systems could induce some latency issues. I have currently two 5 site voting systems with the mixed mediums, and I think both sound pretty darn good (pat on my shoulder). I do have to say that I like the tonal quality of the spectra tac over the JPS SNV 12, as I have one of each deployed.

[d119]

Any chance you can provide some sample audio (a few transmissions) of the SNV-12?

I've not heard one (that I know of) and would like to hear how it sounds.

I can provide an e-mail address for this if you dont want to post links.

Thanks!

[911-EMT]

All saying microwave is the way to go so my question is for 2175hz tone and audio from the quantar to the digitac how many MBPS do i need for the link at 99.999% (PTP LinkPlanner)

[Bill_G]

Depends on how you transport the audio. If you use an ethernet appliance like the NXU , you won't need more than 100kbps which is lower than the lowest rate most links will operate. OTOH, if you use a DSU to transport a T1, then you'll need more than 3Mbps.

[911-EMT]

In PTP LinkPlanne I used 1x T1 and it gave me 4 MBPS at 100% this should be more then I need for 1 quantar

[alex]

Interesting thread.

You are going to have issues with links that you do not have control over delaying your packets getting to their destinations on time. It is not something that the internet is particularly good at. Your data will get there but using things like multicast which has no buffering or error correction to say oops, I missed that chunk of audio, send again is going to be your biggest problem.

So how do you get around this?

Talk to your internet service provider and ask them what the best way to do this is. Explain that you have 4 sites. You want to have the lowest possible route between sites A, B, C, and D. Ask them how to best support making this happen. If you are in line of site to each receiver think about doing PTP links or other methods of network back haul.

This is something that is doable if the planets are aligned in your favor. However the fact that you are building a system which includes points which you do not have actual control over means that you are going to have problems at different points. I guess you could say it is no different than when Verizon shows up and decides that the circuit you are using for your receiver should be used for someone's home telephone service....

Alex

[911-EMT]

Thanks Alex we now have VG6's and spending a lot of time with Verizon to prove that levels are bad to the point ware i showed the tech on my tims bad levels and the tech asked me when is the last time the tims was calibrated and had to show him the certificate that it was calibrated 4 weeks ago then suddenly they fond voltage on the line and fixed the problem but till then they refused that the line has problems so this is why i am looking to get control over the links

[Bill_G]

Hahaha!

Sorry.

God bless the phone company, but sometimes they do make our work hard. Since switching many ckts over to dsl with ip appliances, I became the phone company. They just have to ensure I have connectivity to a specific address. I take care of the rest. Levels became my problem. Timing became my problem. Which site talks to which is my problem. It's a beautiful thing. Swapping out the phone company with some ptp product makes sense if you have LOS. It solves a lot of problems, and the roi is pretty fast depending on what you invest in hardware. We're putting in a simulcast system based on an ip infrastructure. So far so good. The bulk delays are huge (100's of mS), but it works. We will be seeing more of this.

Any one out in gorge in the Cascade Locks to The Dalles OR can listen to 158.805. Hood River Co is not the busiest fire dept in the world. They log about 1h 10m of traffic per week. But, it's a three site simulcast based on Canopy backbone with Harris muxs into Quantars and a SNV-12 voter. Sounds great. They love it. We're building a six site, two channel system just like it for K Falls right now.

[d119]

Bill - Any off-the-air audio samples??? I would like to hear a few minutes worth of transmissions via an SNV-12.

[Bill_G]

I'll see what I can do.

[Bill_G]

I have a few minutes of off the air recordings of a local railroad using a SNV-12 voting system. VHF MTR2000 repeater with two additional rcvrs connected via NXU's through the interwebs. Dispatch Telex console comes in to voter console input also via IP223 and interwebs. In one recording you can hear occasional inappropriate voting at the beginning of transmission by a field unit because the COR arrives before the udp audio packets. The SNV-12 figures it out in about 50ms, and passes the correct audio after a short burst of fried egg. Other than that, it sounds pretty good. Send me a PM with your addy. I have three wmv files about 3-5M each.

[d119]

I have a few minutes of off the air recordings of a local railroad using a SNV-12 voting system. VHF MTR2000 repeater with two additional rcvrs connected via NXU's through the interwebs. Dispatch Telex console comes in to voter console input also via IP223 and interwebs. In one recording you can hear occasional inappropriate voting at the beginning of transmission by a field unit because the COR arrives before the udp audio packets. The SNV-12 figures it out in about 50ms, and passes the correct audio after a short burst of fried egg. Other than that, it sounds pretty good. Send me a PM with your addy. I have three wmv files about 3-5M each.

HOW are you getting all the audio to line up at the right time with NXU's, let alone NXU's connected via the interwebstubes?

[Bill_G]

HOW are you getting all the audio to line up at the right time with NXU's, let alone NXU's connected via the interwebstubes?

Two things - It is earballed in using the course delay adjustments on each SVM card in the voter, and adjusting the rx buffers on each NXU. It took some experimentation in the beginnning, but has remained fairly consistent for a couple years.

[d119]

You and I need to do lunch someday, somehow Bill.