Jump to content

Dakota 20 vs Vista HCX - Accuracy, Precision and Rundown


Tui Chub

Recommended Posts

I did a battery rundown test between my Dakota 20 and eTrex Vista HCX. I had each receiver set to record a trackpoint every 15 secs. They were set next together. The batteries were MAHA Imedion 2100 MAH low discharge batteries. I bought the batteries at the same time. The batteries were charged up overnight. The Dakota died after 14 hrs 7 mins; the Vista HCX lasted 25 hrs 52 mins. I compared the trackpoints against each other. I trimmed the first half hour off to remove drift while the receivers were searching for satellites and then trimmed the Vista off so the two gps's covered the same time period. I then plotted them on Google Earth. The Dakota had an overall horzontal standard deviation of 4.7 ft while the Vista was 5.6 ft. The mean location between the two was off by 7.4 ft. They were both off from the actual location from Goggle Earth. The Dakota was off 18.4 ft and the Vista was off 11.6 ft. Now I do not know how accurate Google Earth is for my area. Red is the Dakota, Green is the Vista, the markers are the mean position for each track, the Actual is the location in my house as displayed on Google Earth.

 

GPSComp.jpg

 

Pete

Link to comment

I was surprised by the difference between the tracklog and resulting averaged waypoints between the Vista HCX and Dakota 20 and was wondering which one was displaying the more accurate results. I went to a known benchmark that I had used before (KR1744 from the NGS datasheet). This time I had three gps's, the third unit was an Edge 705 cycling gps (this one has a SIRFstar III chip). I entered the benchmark's coordinate into all the units and navigated to the benchmark. All the units displayed 0 feet when held over the benchmark (there was some obvious filtering going on here).

 

Units next to benchmark.

all.jpg

 

I then plotted the waypoints and tracklogs on Goggle Earth.

KR1772.jpg

 

I set the units to collect trackpoints every second for a one hour period. In addition, I tested the waypoint averaging of the Dakota and Vista (the Edge dosen't have this function). Also both units had WAAS turned on (the Edge doesn't have WAAS). The averaged waypoint for the Dakota was only 3 inches off the benchmark. This is interesting since the Dakota only used 2 points for the averaging. Garmin evidently is using a new waypoint averaging algorithm with these new units.

 

Waypoint Averaging

 

The Vista used the old method and averaged every second. The Vista averaged waypoint was off by 6.3ft. Since I was lazy I visually estimated the average waypoint for the Dakota from the tracklog. That average was 1.5 ft off. For some reason the Edge did not produce a track log. I also took a non averaged waypoint at the end of the one hour period. The Dakota was off 2.9 ft and the Vista was off 8.6 ft, the Edge was off 19.8 ft. The distance from the trackpoint averages were 7.2 ft, similiar to the 7.4 ft I saw in the first test.

 

I looks to me that Garmin changed the way the newer units determine positions. The newer units appear to be more accurate than the older ones.

Link to comment

Nice analysis work! Perhaps you could draw up some sort of chart that shows what the satellite constellation arrangement was like during your testing, meaning were the available satellites pretty evenly scattered overhead, or what?

 

And you gotta LOVE the battery fuel economy of the Etrexes eh? :ph34r:

 

I get a TON of hours from my Vista Cx, easily enough to use for a whole weekend (not continuous of course) with no worries of carrying spare batteries :P

Link to comment

Thanks for the tests/comparisons. I have a couple of comments:

1) Obviously both units perform well within the probable error that Garmin describes for these units.

2) I wonder what effect different firmware/chipset configurations would have had on the results and, assuming that there is such a thing, that the pairs of units in these tests were using comparable configurations.

 

Obviously there is no way to run a simultaneous test on the same unit using different software configurations, but my sense is that, with the Vista HCx, the unit seems to perform more erratically with some configurations than with other. And, if the high sensitivity eTrex units use the same chip as that in the Dakota, presumably they both would be prone to similar potential for erratic performance associated with variable software configuration.

Link to comment

It seems to me that putting two or more gps units so close together you'd create electrical interference that might throw them off. I can take an AM radio tuned between stations and bring my garmin 60 cs up to it and you get a noticeable spike in the static. The gps is definitely capable of causing electrical interference. You might have more valid test results by keeping some distance between them. ?

Link to comment

Great comparisons. I'm surprised you got all three units to zero out on distance to destination. I rarely manage that with one unit, let alone 3 on the same day. I've taken all three of my units to benchmarks a couple times now, but haven't taken all three at the same time, nor done any proper documentation, or testing other than getting an idea, or impression of how they perform.

Link to comment

Thanks for the tests/comparisons. I have a couple of comments:

1) Obviously both units perform well within the probable error that Garmin describes for these units.

2) I wonder what effect different firmware/chipset configurations would have had on the results and, assuming that there is such a thing, that the pairs of units in these tests were using comparable configurations.

 

Obviously there is no way to run a simultaneous test on the same unit using different software configurations, but my sense is that, with the Vista HCx, the unit seems to perform more erratically with some configurations than with other. And, if the high sensitivity eTrex units use the same chip as that in the Dakota, presumably they both would be prone to similar potential for erratic performance associated with variable software configuration.

The Vista was running software version 3.00 and GPS SW version 2.80 while the Dakota was running software version 2.10 and GPS software version 3.40. My Vista was one of the early ones and had a Mediatek chip where I believe the Dakota has a SMT chip. Both units were set to Degrees Minutes Seconds, the same as on the NGS Datasheet and were using the WGS84 datum. The datasheet uses NAD83, I did not try to do any conversions between the two datums.

 

Though not at the same time, here is an earlier test that I did in 2007 with the Vista.

 

Early Vista HCX Test

 

I've never seen the erratic behavior that some people have had with their eTrex HCX units, maybe I've been lucky.

Link to comment

Great comparisons. I'm surprised you got all three units to zero out on distance to destination. I rarely manage that with one unit, let alone 3 on the same day. I've taken all three of my units to benchmarks a couple times now, but haven't taken all three at the same time, nor done any proper documentation, or testing other than getting an idea, or impression of how they perform.

It is curious that all three zeroed out. Looking at the tracklogs and waypoints I wouldn't expect the Vista and Edge to, they seemed to be off. Only thing I can think is there's some filtering going on i.e. if the unit is within so many feet then show the distance as zero.

Link to comment

I set the units to collect trackpoints every second for a one hour period. In addition, I tested the waypoint averaging of the Dakota and Vista (the Edge dosen't have this function). Also both units had WAAS turned on (the Edge doesn't have WAAS). The averaged waypoint for the Dakota was only 3 inches off the benchmark. This is interesting since the Dakota only used 2 points for the averaging. Garmin evidently is using a new waypoint averaging algorithm with these new units.

 

Waypoint Averaging

 

The Vista used the old method and averaged every second. The Vista averaged waypoint was off by 6.3ft. Since I was lazy I visually estimated the average waypoint for the Dakota from the tracklog. That average was 1.5 ft off. For some reason the Edge did not produce a track log. I also took a non averaged waypoint at the end of the one hour period. The Dakota was off 2.9 ft and the Vista was off 8.6 ft, the Edge was off 19.8 ft. The distance from the trackpoint averages were 7.2 ft, similiar to the 7.4 ft I saw in the first test.

 

I looks to me that Garmin changed the way the newer units determine positions. The newer units appear to be more accurate than the older ones.

 

According to the information behind the link you provided, the new methods main advantage is to repeat a standard averaging at several times when the satellite constellation has changed significantly. You could recreate the method with any "old" unit by averaging 4-8 waypoints with 90 minutes between the measurements and then average the results manually. The only true advantage of the new method is that it is suggesting the proper time to take the next measurement and that it autmatically averages the averaged waypoints.

 

Please don't get me wrong, your comparison is by far the best that I have seen so far. But there are some factors which, in case you would like to repeat your test in the future, could be improved:

 

- The units are placed very close to a steel post (the benchmark), which may be blocking the line of sight from a unit to one or more satellites. More importantly, since the units are arranged around (!) the post, each individual unit may have a blocked view to a different set of satellites.

 

- The units should all point to the same direction

 

- The units should be positioned with some distance to each other

 

I would suggest an arrangement resembling a "T" - draw a line from the benchmark, one meter in length, in a known direction (1m north of the benchmark, for example). At the end of this line, place one unit. Place the next unit 50 cm to the west of the first unit, and the third one 50 cm to the east.

 

This way the units would not show the same location, of course, but you still know the true position that each unit should come up with. Then you can compare each result to the true position of each unit.

 

This may not have anything to do with every day usage, but if you want to truly compare the units you have to eliminate as many sources for errors as possible :unsure:

 

best,

Pulvertoastmann

Edited by Pulvertoastmann
Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...