Ellis's GPS Support Page
CDPD Modem RTK

On January 15th , 2003, I went around the Fort Collins, CO area with a CDPD modem hooked up to my trusty System 530. The base station was HIX1 at the Hixon Manufacturing Inc. headquarters in Fort Collins. It is an RS 500 pumping phase measurements to the internet via Taltech's TCP/Com software.

I was using an Airlink Redwing CDPD modem hooked to port 3 of my receiver. This is the way to RTK (as long as you have CDPD coverage, ie. near urban areas again).

The position we used for the base station was the mean of 3 different OPUS solutions from 3 different days. (It's not an offical base station yet). I gathered control within 20 miles around the station and entered the coordinates into DeLorme's Topo USA. Then, with my laptop as my guide (assisted by the NGS to-reach descriptions), I went in search of monuments. What started out as a nice day soon turned windy and then included some snow squalls, so we cut the experience short...no deadlines here.

This time I put the antenna on a tripod to avoid pole (the bottom section of my pole is bent a little) and bubble errors. I visited the first three stations (Greer was unsuccessful because I didn't have CDPD coverage behind the hogback), but was joined by Shannon Hixon at C 391, for the last three stations.

Once again, this was not a rigorous test for accuracy, just seeing how well the CDPD modems perform around Fort Collins. I will expand the CDPD testing when Hixon Manufacturing puts up it's reference station here at their new store in Denver.

Here is a map of the area:

This is what the geoid looks like in the area:

The first station I went to was I25 MP281.96. Of course I forgot to take a photo and record anything other than the fact that I did 2 OTF, 2 Static, and 2 Known Point intializations. Is there a party chief in the house? I did manage to write down the antenna height.

The next point was R 308. I did remember to take a photo of the monument here. And again I got the antenna height, the fact that I did 2 OTF, 2 Static, and 2 Known initializations, but nothing else. Hey, it was windy.

I then tried to get Greer (in the northwest corner of the map above), but couldn't connect to CDPD because it is behind the hogback. I did connect about 1/2 mile southeast.

Next I drove east (well easterly, no straight connection on the township roads) to C 391 and met Shannon Hixon. Here we had some problems. It didn't want to do an OTF, so I entered in the published coordinates and did a known point initialization. Then I did two static initializations, one took 23 measurements, and the second took 15 measurements. Then I tried OTF again without success. Finally we tried another static, but we hadn't fixed at 100 measurements so we bagged it. We had some problems with losing reference data, so I think this area might be a little weak in CDPD coverage.

Our next station was K 262. Here again I forgot to take a photo. We did two OTF solutions that I didn't time. Next we did two static intializations, the first in 19 positions and the second in 17 positions. Finally we did two known point intializations, the first in 9 positions and the second in 9 positions.

The final station was HORA. The first OTF took 10 seconds, the second 9 seconds. The first static intialization took 11 positions, the second 10 positions. The first known station initialization took 6 positions, and the second one 5 positions. Just as I was getting into the swing of keeping records it was time to go.

Here is example of the individual shots at R 308.

Here is example of the individual shots at C 391. The others are virtually the same.

Here are the results comparing the mean values calculated by SKI-Pro from the repeat measurements recorded in the receiver with the NGS published values:

Station     Northing      Easting      Ell. Ht.    Orth. Ht.  Geo. Sep.  Distance    SD N   SD E   SD H
                                                                                    
hix1       442791.6686   956383.4665  1492.5205   1508.8785   -16.3580   0          0.0000 0.0000 0.0000
                                                                        
MP281.96   463630.364    957279.190   1630.249    1646.218    -15.969    20858 m    0.0018 0.0018 0.0013
Published  463630.331    957279.183   1630.22     1646.2      -15.97     12.96 miles    
                 .033          .007       .029        .0                            
                                                                                    
R308       473998.741    949730.352   1712.772    1727.992    -15.220    31909 m    0.0020 0.0010 0.0044
Published  473998.724    949730.343   1712.77     1728.042    -15.22     19.83 miles
                 .017          .009       .002       -.050                          

GREER (no CDPD connection)
                                                                                    
C391       469243.027    974254.716   1652.367    1669.174    -16.807    31923 m    0.0029 0.0014 0.0033
Published  469243.025    974254.709   1652.36     1669.240    -16.81     19.84 miles           
                 .002          .007       .007       -.066                          
                                                                                    
K262       457456.663    975248.588   1565.211    1582.242    -17.031    23895 m    0.0012 0.0038 0.0034
Published  457456.643    975248.583   1565.18     1582.269    -17.03     14.85 miles
                 .020          .005       .031       -.027                          
                                                                                    
HORA       438548.814    981047.848   1458.849    1476.592    -17.743    25027 m    0.0016 0.0016 0.0027
Published  438548.773    981047.837   1458.80     1476.6      -17.74     15.55 miles           
                 .041          .011       .049        .0                            

All in all I was pretty impressed. 32 kilometers (19.8 miles)...not bad for RTK. The System 500 and SKI-Pro database is really nice for this kind of work, as it records each measurement of a point and automatically means it. I will continue exploring the CDPD performance in Denver as soon as there is a reference station up. One question I have is why can I match ellipsoid height and be off in ortho height when the geoid separation is the same? Stay tuned.

Some photos of the monuments:

Monument R 308:

Monument C 391:

Site of C 391:

Monument HORA: