Loading ...
Sorry, an error occurred while loading the content.

6698Re: Serious issues found with the Harris "simulcast" P25 Cqpsk system

Expand Messages
  • chris tofer
    Jun 1, 2012
    • 0 Attachment
      I got a letter from a consultant for the county forwarded to me.
      They say there are 3 new standards added to the P25.
      These were added because Motorola changed their waveform to work better with simulcast and enable the radios to decode 2 or 3 towers at the same time and add the signals together and clean up the time delay and possibly the frequency offset. Harris pressed on them to open up what they were doing so all vendors could roam on each others systems. The processing is probably not present in the GRE and Uniden did an update on the digital scanners that "allows better reception of simulcast systems" which may mean they added the waveforms to their decoders. Not sure the Uniden will combine two signals together unless they are zero beat. The suggestion I got from them was they were going ahead with the system as is since the radios they are buying will work. I suggested they buy the gps time reference for all their towers if available. That it would greatly improve the way their radios work on the system.


      In the letter they admitted digital coverage was half of what the analog coverage was. (The selling point was that it would be better)

      So I plan to pick apart that letter and look at the refrences maybe call them and have a few more rounds with the county supervisor about it. The letter did not satisfy me that's the best they can do. They really need the time standards. The uniden scanner has a tough time decoding with a 2 hz frequency offset between the two towers. It sounds like Harris radios have processing that will lock in on that much. I'm not sure the freq offset tolerance is in the spec or if I can even get to it without paying the big bucks.





      --- In BCD396XT@yahoogroups.com, "dataman60435" <dataman60435@...> wrote:
      >
      > WOW, thats a excellent article you put together...I monitor the IL-Starcom 700/800 9600 Digital system, the stand alone towers we call them are perfect on receive, the simulcast 700mgz in the Chicago suburbs are a pain for me living 25 miles or more using out door "Channel Master Triband with LMR400-60'....I can go two blocks down the street with a rubber RS800 duck and they come in great plus, on any given day the weather conditions seem to change reception ....I been monitoring for over 40 years and the old VHF-UHF were a briese compared to these type's of setup....I use a 996XT, RS197, and my 396XT....I sold my GRE500 as it was so bad it would not decode much of anything in my location 30 miles SW of Chicago....
      >
      > --- In BCD396XT@yahoogroups.com, "chris tofer" <chris451@> wrote:
      > >
      > > I've been a Uniden digital scanner user since these first came out, have used each of the series of radios. 296, 396, 396 xt.
      > > I've also traveled on some vacations and gathered some interesting data to support my points.
      > >
      > > To start with, several states use Cqpsk and most new sytems are using this instead of c4fm.
      > >
      > > A local agency Cedar Rapids City of, (Iowa) has had a motorola p25 system, using a mix of analog and digital, and all motorola, and using a two site "simulcast".
      > >
      > > A nearby agency Johnson County Iowa and cities therein have installed in the last year a 6 site Cqpsk system (by Harris).
      > >
      > > During my travels I've gathered some data on some Colorado Cqpsk multisite simulcast systems.
      > >
      > > While on a train heading to denver I was able to pick up the Colorado springs system at 115 miles away, log that lat/long into the 396xt for later license analysis, so I have a freq and a location I can look up and verify the distance.
      > >
      > > I mapped out where the 6 sites are on the Johnson county system.
      > > I know where the two sites for Cedar Rapids are located.
      > >
      > > My research shows the following info about each system, and I'll compare the performance of each.
      > > ____________________________________________________________________
      > >
      > > Colorado: Daniels electronics base stations, unknown user equipment.
      > > Colorado Springs: Cqpsk with 4 simulcast sites
      > > JoCo: Harris base and mobile portable. Cqpsk with 6 sites.
      > > CR: Motorola base and mobile portables. C4FM with 2 sites.
      > >
      > > Linn County: Planned Harris base and mobile portable with multiple sites, planned 700/800mhz and plans interface with state of Iowa.
      > > Currently on VHF repeaters.
      > >
      > >
      > >
      > > Now let's talk about the performance for each.
      > > For each digital mode the receiver is a 396xt.
      > >
      > > Observation points: NE side of Cedar Rapids, or NE end of Benton county and I-380 between I-80 and US-20.
      > >
      > >
      > > My observations for each system:
      > >
      > > Colorado springs:
      > > Detected 115 miles out, 5600ft with 4 bases at 7200feet
      > > Simulcast is assumed but has no carrier hetrodyne effects noted.
      > > No periodic chopping or garbling, audio quality good, no echo effect.
      > > No analog use noted or possible.
      > >
      > > Cedar Rapids:
      > > Solid coverage for 10 miles, some small amount of echo and breakup, analog always has a DPL rate heatrodyne but at low levels.
      > > Trunking is difficult at 17 miles out, subject to propogation changes the control channel fades in and out, P25 broken to good, some "self interference" effects are noted at 0.5 to 1 hz rate, some echo and chop at maximum range, but pretty good in close. If you fish for a spot where the hetrodyne between two towers is maximized, its pretty broken up.
      > > Method used: Program the control channel to a conventional test bank in AM mode, and maximize AM beat effect.
      > >
      > >
      > > Johnson County:
      > > Initial impression was it is an incompatable format such as mototrbo as it would not decode the control channel, as such was ignored.
      > > Further news reports and lack of activity on their old channels led to further investigation.
      > > Taking a closer look in analog, it was found to be a cqpsk but with a 1.7hz AM component. Beams indicate a site to the SE of Cedar Rapids.
      > > So Radio Reference info was loaded into the 396xt and on a trip by johnson county some broken audio was detected, but seemed to be reading encrypted, with very occasional clear.
      > > I queried the JoCo supervisors who didn't know much about it but forwareded the query to someone in the comms division who said which channels are clear, most of them except 10-28's but not the main, fire, etc.
      > >
      > > OK so I gave this a try again for a week or two. I found that yes the groups are clear but very rarely decoded.
      > > I have noted fairly good reception next to Cedar Rapids Airport.
      > > It turns out they have a 6th site very near the airport.
      > >
      > > Upon further investigation it was found there were not only 2 sites in this system but 6!!! Astounding. And there's lots of signal but poorly decoded.
      > >
      > > For the next trouble shooting step I connected to a 15 element Yagi, and maximized the strongest site, isolated the other sites and connected the digital scanner. Alas finally it decoded and tracked the system. With analog AM I found I pick up at least 3 of the sites, which when mapped out shows 2 in north Johnson county and 1 central site thats on high ground.
      > >
      > > When beamed in on the best site, Range 20 garbling and echo minimal but occasional, system fades in and out.
      > >
      > > I have noted each site is on a slightly different carrier freq but within 2hz. Most places in Linn County have 2 sites about the same level so they mix together, they are within 30% of the same level. Omni reception is poor to fair depending on being closer to one site or the other.
      > >
      > > Now let us compare to some known single site systems.
      > > Montana has one site on VHF on a mountain. Had great coverage even to the bottom of the mountain valleys witin 3 - 5 miles and good coverage on any line of sight locations.
      > >
      > > Wisconsin has a number of single site systems such as Granite Peak (wasau wi). Great coverage from the one site throughout that and the next county.
      > > No chop echo or in analog any am modulation effects.
      > >
      > >
      > >
      > > So let us summarize:
      > >
      > > Single Site: More dead spots, but very clean otherwise.
      > > Dual Site C4FM : Few dead spots but slight artifacts such as echo and chop, especially at the 15+_ mile mark.
      > >
      > > Multi Site Cqpsk: Great signal level but depends on vendors.
      > >
      > > Daniels as done in colorado: 115 miles (as soon as line of sight)
      > > Harris as done in Johnson County: 20 miles if using directional antenna; 5 miles if using omni, more if in a place where one tower predominates.
      > >
      > >
      > > Thats pretty dramatic. there's a ton of signal but the system interferes with itself.
      > >
      > > I think what's going on here is a [serious] design flaw in Harris' implementation.
      > > Motorola is the first vendor I've heard of using "simulcast"
      > > C4FM simulcast works fairly well because the decoder looks at FSK Freqency, and all towers are on the same shift at the same time,
      > > So everyone developed "simulcast offerings" but some implementations work very well ( Colorado/Daniels) and some don't(Harris).
      > > So what is C4FM? Four frequency shift modulation.
      > > And CQPSK? 4 phase shifts with some AM to reduce spectrum width.
      > > To detect C4fm a different method is used that doenst depend on carrier phase.
      > > To detect cqpsk the phase of the incoming carrier is decoded.
      > >
      > > If the incoming carrier is two signals of nearly equal amplitude but different carrier frequency then the decoder will try to lock in one carrier then the other, or a mix between the two. Maybe Harris has receivers that will work with 2 carriers combined together that are off frequency.
      > >
      > > It's fairly easy to detect this condition, when tuning to the control channel in AM analog one can hear what appers to be a control channel with AM on it.
      > > Cqpsk has a rather strong 4800hz component to it, and one effect as a beam antenna is rotated is the fades become very high as the two towers become equal in amplitide. The fades occur at the hetrodyne rate between the towers. This is 1.7hz at 851.050 mhz in this case.
      > >
      > > So why does the Daniels system work at 115 miles and the Harris work at 5 miles?
      > >
      > > Because there was no perceived carrier frequency offset on 4 towers that were in use there in Colorado, or their system only transmits on one site at a time.
      > >
      > > In the Harris system, the carriers are within 2 hz of each other.
      > >
      > > At 851.05 there is no doubt a reference in all the base stations such as a TXCO, which is multiplied up. So the 2hz translates to an error of .001parts per million, pretty hard to set up in a crystal or measure, takes a long time to set it that close.
      > >
      > > In the Daniels system, either they don't transmit more than one place or they use a time base reference at each base station such as a rubidium standard or a gps synchronized time standard, for all the synthesizers.
      > >
      > > I know Harris is not doing that, becuase of the analog beat note.
      > >
      > > I find all this interesting because I use my trunking radios for my own news, health, property protection and welfare.
      > > I don't like to see radio systems I use that work poorly.
      > > My argument to the county here is they need to check this out before they allow Harris to put the same defective design into our county.
      > > I believe they were budgeted at 17.8 million for a system like Johnson County's.
      > >
      > > According to a newspaper in Johnson County, someone in the communications division resigned recently. Perhaps it has something to do with my observations and research.
      > >
      > > Some other observations:
      > > A scanner user in Johnson County wonders what is wrong with his scanner program.
      > > Another poster on the groups observed Indianapolis has a simulcast system and its users have a tough time with it.
      > > Harris responded initially and has not given an answer to my query's.
      > >
      >
    • Show all 8 messages in this topic