Coverage Maps of Colorado DTRS sites (image heavy)

[soundchaser]

Here is Adams (Civic) and the area of interest.

 

[ecanderson]

Yes, that really is strange. My location at the time was in and around the white dot:

 

[ecanderson]

Soundchaser was good enough to drop down the hill for a confirmation run today. None of his rigs and none of mine can decode a strong Thorodin signal if we are in close proximity to the Mead site. The fact that these two sites have control channels only 12.5KHz apart seemed to be the culprit. None of our radios have the necessary selectivity to pull Thorodin out of the 'thorough din'.

In the FWIW department, we originally made it a tough case - sitting pretty much directly under the tower. But as we departed S on County Line, after dropping considerably in elevation (certainly wouldn't have helped Thorodin reception), the tower was finally shadowed by the hill behind us, and the decode on Thorodin started looking really good again. It nicely confirmed the nature of the problem. Then it bounced back and forth between good and lousy in the general area of County Line and 66. Throughout much of the vicinity, Thorodin was useless due to the adjacent channel interference. Only after getting into more southerly parts of Longmont does Thorodin become entirely reliable.

I'd love to get a piece of commercial gear up near Mead to see if it fares any better.

 

[ecanderson]

Have grown exhausted with all of the "it's easy" and "coverage is as good as VHF was" comments that have shown up in various forum threads here and elsewhere. The process of programming to avoid dropouts was just becoming too tedious. Time for some "facts". An email from jimmnn, without his realizing it, has spurred me into action.

I've put the final touches on a small application that concurrently grabs the decode % off my PRO197 and the NMEA $GPGLL data off my Garmin Oregon 450. The result is a CSV file of latitude/longitude and decode % (along with some other site data) at 5 second intervals. Drop the notebook in the front seat, and I can do some hardcore mapping of just what the propagation of some of these DTRS sites really looks like in real-time. Overlay the data from the available sites, and voila - a dead spot map.

Now all I need to do is come up with a good mapping solution to dump the data into and we'll see just how many holes there really are in my local DTRS area. It would be nice to have each of the coordinate waypoints represented on a map by a color. Green = usable decode %, Yellow = iffy decode %, Red = unusable decode %. I'm looking around for something other than a solution like Garmin Mapquest or BaseCamp to do this job, but I'm open to suggestions - especially suggestions that allow display of the results on a public site without too much hassle. Suggestions?

 

[ecanderson]

As a side note, the data currently looks like this - in case it should get anyone's imagination going...

Latitude  ,Longitude   ,D%,Site   ,TimeStamp   4/7/2011  12:18:53PM
40 10.7058,-105 06.2512,99,003-054,12:19:00PM
40 10.7062,-105 06.2512,99,003-054,12:19:05PM
40 10.7062,-105 06.2512,99,003-054,12:19:10PM
40 10.7062,-105 06.2512,99,003-054,12:19:15PM

 

[soundchaser]

Have grown exhausted with all of the "it's easy" and "coverage is as good as VHF was" comments that have shown up in various forum threads here and elsewhere. The process of programming to avoid dropouts was just becoming too tedious.

When last we spoke in person, you had discovered some coverage holes in Longmont that were not covered by either Mead, Longmont, or Gunbarrel. I suggested finding a hole and seeing if Thorodin filled it in. In looking at the coverage maps, I'd also check Bald, and Ft. Lupton. I gotta think that with 6 sites potentially providing coverage to Longmont there wouldn't be that many holes.

Longmont radios affiliate with Thorodin and Bald. I've never looked for them at Ft Lupton. Anyone know/want to log Ft. Lupton for a day and see if Longmont shows up there?

I like your GPS mapping idea. I'm trying to think if there is any way to do it a little differently so that you wouldn't have to drive the same route over for each site. In other words monitor multiple sites at once.

 

[ecanderson]

Longmont radios affiliate with Thorodin and Bald. I've never looked for them at Ft Lupton. Anyone know/want to log Ft. Lupton for a day and see if Longmont shows up there?

Have done. And yes, you find Longmont on Ft. Lupton and Bald N. as well. For one little town, they show up all over the place. Of course, we both now know the issues with Thorodin in the north end of town, but that end is more than covered by Mead that creates the issues to begin with, so we're good there (I think!)

I like your GPS mapping idea. I'm trying to think if there is any way to do it a little differently so that you wouldn't have to drive the same route over for each site. In other words monitor multiple sites at once.

Wish I could imagine a way to do that that didn't entail using another GPS and receiver for each site. Well, I did come up with a thought for the GPS issue, providing the clocks on enough computers could be sync'd... just write the GPS data with time stamps with one system and collect radio information with similar time stamps on other gear. But as I'm not running a van with 19" racks in the back, I think I'll start out by just working it with what I have and seeing how the data starts to accumulate. Much of what I'll be recording will be from trips that are repeated weekly anyway. Longmont will just be the test area for the setup. What I expect to find more than anything is that it takes far more sites to cover this few square miles of town than anyone could imagine, and will also see whether the coverage from particular sites is as good as one would expect. As much as anything, I'm interested in how the Adams simulcast system can leave so many gaps on I-25 between Weld and Denver counties.

 

[soundchaser]

Wish I could imagine a way to do that that didn't entail using another GPS and receiver for each site.

I was thinking along the lines of having your app sending commands to the radio over the RS232 to change sites every few seconds to check the decode. But the devil is in the details of that.

 

[ecanderson]

I was thinking along the lines of having your app sending commands to the radio over the RS232 to change sites every few seconds to check the decode. But the devil is in the details of that.

That could be done, although it would take a fair number of commands to get there. Since the CCDUMP method provides "Q:" data (decode %) on a very infrequent basis (sometimes more than a minute between reports for some reason), it's necessary to use Analyze mode and grab the decode % from the "remote display" data.

But the real fly in the ointment is that monitoring six or more sites in sequence means waiting on the rig to come up with the decode % value after reselection of each DTRS CC. That can be fairly quick for a strong site with the receiver in a fixed position, but a weaker site measured while on the move can take a fair bit of time (5 seconds or more) to evaluate before a % is produced. I'm not sure how many packets the receiver incorporates into the initial displayed value, but it can take a while on more difficult sites.

Unless the signal can no longer be decoded, if you hang on a particular site on a continuing basis, the rolling average (or whatever it is) is displayed continuously.

So my concern in measuring multiple sites in sequence is the "dwell time" required at a given location (or some REALLY slow driving!). Holes for individual sites can be as small as a couple of square blocks or so, and it wouldn't take long to pass through that much territory at even 25mph. As it stands, I can produce a new reading once a second, although I'm using five seconds at the moment. If I could get an 'instant' decode % capture when switching to a new DTRS CC, that would be fine. I could just as easily do 5 in those 5 seconds as 1. But with the delay involved, by sequencing multiple sites, I would reduce the granularity of the result (geographically) substantially. Based upon what I've seen, it would probably require at least 30 seconds to pull up good data for six sites that include both boomers and weaker ones. I can do a lot of driving in 30 seconds (about 0.2mi at 25mph), and once I hit roads like CO119 and CO66 where pooping along at 25mph is going to get the tires shot out from under my car, I'll have to be moving a good bit faster.

Hey, it's amateur hour in the propagation business. What can I say? Just don't have enough gear to make this easy, but at least I've got the right gear to make it accurate.

 

[tbiggums]

Don't forget the effect 800 MHz cellular sites will have on your drive test results. Pretty much any scanner will be affected when you're within an 1/8 mile or so of an 800 MHz cell site. The desense my PRO-197 gets from these is pretty frustrating. Often if I'm within a few hundred feet of an 800 cell site, no control channel decode is possible, even with the attenuator on.

So if you see some small areas that seem to have no radio coverage from any site, and they're not blocked by terrain or anything, cellular desense is probably what's going on.

 

[jeffreyinberthoud]

I wonder if this issue between thorndin and mead might be part of the Longmont coverage situation they are facing ..FD mainly ............Strange that PD does not seem to upset Ive not heard complaints all that offten other than the occasional "Gone digital "

 

[ecanderson]

I wonder if this issue between thorndin and mead might be part of the Longmont coverage situation they are facing ..FD mainly ............Strange that PD does not seem to upset Ive not heard complaints all that offten other than the occasional "Gone digital "

Not a problem for them (or me). If you're close enough to Mead that it's blowing away Thorodin, you'll be associated with Mead anyway, not Thorodin.

What I'm working on now are answers to why we hear the dispatcher saying "Paul 214, can you repeat that? You were digital". With the site coverage every which way around here, I'm wondering why this is possible and where all it can occur. Sure takes a lot of 800MHz sites to replace a simple 150MHz site.

 

[ecanderson]

Don't forget the effect 800 MHz cellular sites will have on your drive test results. Pretty much any scanner will be affected when you're within an 1/8 mile or so of an 800 MHz cell site. The desense my PRO-197 gets from these is pretty frustrating. Often if I'm within a few hundred feet of an 800 cell site, no control channel decode is possible, even with the attenuator on.

So if you see some small areas that seem to have no radio coverage from any site, and they're not blocked by terrain or anything, cellular desense is probably what's going on.

Could be -- it all counts. If it's poor coverage, I'm somewhat less interested (initially) in knowing why as I am where. Once the bad spots are found, then it will be of some interest to see if explanations can be found. That said, I can't say I've had your experience with cellular interference here. How far away from the DTRS site(s) are you when you get swamped by the cell site? Like I say, I have to be pretty close to Mead for it to blow out Thorodin's adjacent control channel. The cell freqs are far enough away that those sites must be taking out the whole front end where you are.

 

[ecanderson]

Did a bit of a drive-around and covered a lot of the NE corner of Longmont. What you see here are the dead spots (decode < 80%) on the 3-56 site. Will take a lot of time to collect data to this level, but it will be fun to see how it all works out. Eventually, will get the other sites coordinated with this data and we'll see what's what in Longmont.

Anyway, the tool is working well, and I'll be able to use it (probably not at such tight intervals, mind you) to map other areas and sites.

 

[soncorn]

Did a bit of a drive-around and covered a lot of the NE corner of Longmont. What you see here are the dead spots (decode < 80%) on the 3-56 site. Will take a lot of time to collect data to this level, but it will be fun to see how it all works out. Eventually, will get the other sites coordinated with this data and we'll see what's what in Longmont.

Anyway, the tool is working well, and I'll be able to use it (probably not at such tight intervals, mind you) to map other areas and sites.

Can you also plot where you actually went as opposed to just those areas that had less than 80% decode?

 

[ecanderson]

Can you also plot where you actually went as opposed to just those areas that had less than 80% decode?

Absolutely. Actually, I'm currently plotting with as many as 3 separate sets of waypoints: 90%+ (green), 80~89% (yellow), and <80% (red). However, I'm capturing the raw data (decode% and coordinates) every 2 seconds. It can be sliced and dice in any number of ways, to any level preferred. I'll need to come with a more sophisticated way of plotting the data one of these days.

Here's what it looks like if I turn on all 3 colors. Wherever there's a dot, I was recording while driving the area:

 

[soundchaser]

Map from the Radio Mobile program: (the colors are reversed... red = good)

 

[ecanderson]

Slick. May try to do an overlay with them. Looks like the "yellow" area on your map in the center right lines up pretty well with my dead spot there, but the rest of it is a bit of a mystery.

That business down to the east of 3rd and Main really is odd. The NW to SE shape of the shadow would make you think that the transmitter is to the WNW of that location, but it's actually pretty much due west.

The intersection +/- a block from 17th and Pace is pretty awful, too, and there are no tall buildings in the area between those spots and the water tower site to explain that, either.

There's some funny stuff going on. After I've had reason or time to wander around the other quadrants of town to fill in the blanks with Longmont/Sunset, I'll start with Mead next, then Gunbarrel. Since I'm running around with the CC locked on all the time, I haven't been listening to audio, so haven't made any headway in the process of finding other locations that aren't somehow covered by the four sites that I've already got programmed in for listening to Longmont.

 

[ecanderson]

Ah - and FWIW, I'm going to be locked onto the Adams simulcast system on a run down to Denver later this week. I'm going to be very curious to see what the dead spots are on my route (I-25 to I-470 to Quebec). Not all of that is Adams CO, of course, but all of it should be covered -- and isn't, from what I've been able to hear on a lot of past runs. Not even the Adams CO part of I-25 has contiguous coverage off the combined 3 Adams sites.

 

[tbiggums]

I bet desense from 800 MHz cell sites is the reason you're seeing those dead spots...

I'm in EDACS territory most of the time, so it's even more obvious when this is happening as you can hear the effects of it before it gets real bad. Often you can hear a faint buzz or whine mixed with the analog audio, in addition to increasing white noise. Usually the signal strength meter will still be showing full scale even though the audio is fairly noisy. With digital signals, this is harder to detect, as the audio will sound fine until the noise gets too bad, and then the scanner will usually just go silent.

So it's definitely a good idea to use percent decode instead of signal strength alone for this kind of testing...

Even in Lakewood, I'll get desense from cell sites that makes reception difficult when I'm only like 5 miles away from their Green Mountain site.

In Denver, I usually run the PRO-197 with the attenuator on most of the time, as their simulcast system generally provides a strong enough signal that the attenuator doesn't weaken the signal too much when I'm not near a cell site.

Lakewood, Wheat Ridge, Arvada, and Westminster are different, though. Since they've got the relatively distant mountain top site (Green Mtn, Mt. Morrison, or El Dorado) serving them, I have to turn the attenuator on and off as I drive around. Fortunately with analog, you can kind of hear when the signals are starting to get noisy, and know when to turn it on when you're near a cell site.

I bet if you were to re-run the test with the attenuator on, those dead spots would shrink in size. Some new ones might appear where the signal from Sunset is a little weaker. But I think there's a good chance that running with the attenuator on all the time in Longmont when listening to the Sunset site would provide good results.

Could be -- it all counts. If it's poor coverage, I'm somewhat less interested (initially) in knowing why as I am where. Once the bad spots are found, then it will be of some interest to see if explanations can be found. That said, I can't say I've had your experience with cellular interference here. How far away from the DTRS site(s) are you when you get swamped by the cell site? Like I say, I have to be pretty close to Mead for it to blow out Thorodin's adjacent control channel. The cell freqs are far enough away that those sites must be taking out the whole front end where you are.