Dakota 20 vs eTrex HC "Accuracy"

[+ecanderson]

I've had my Dakota 20 for about 10 days now, and I'm beginning to wonder if there's a new approach to the "accuracy" reading. My Summit HC settles a lot faster, and in side-by-side tests, I'd swear the Summit HC is using a radius figure and the Dakota a diameter. Seems my Dakota 20 produces "accuracy" results of about 2X the Summit. New algorithm? Actually less accurate? Less optimistic?

 

I'm also in the process of doing some side-by-side tests of the reported current position between my two units. I'm concerned that I am seeing a difference of 0.003 in the N/S axis - that's 18 feet. Time to do some hard-core repeatability and accuracy measurements on these gadgets now that I have a pair to play with.

[+Renegade Knight]

Just to clarify.

 

You are talking the units calculated estimated position error? Or actual on the ground measured error?

[+StarBrand]

My experience is that there is some amount of variation between the "accuracy" (EPE) calculation even within units of the smae manufacturer.

 

Any long time Oregon owner knows that several of the older firmware updates "tweaked" the accuracy reading with each release.

 

I've had both my Garmin Legend HCx units sitting within 1 foot of each other each with different accurcy readings and coordinates up to .004 from one another. I've also seen the accuracy way different and the coordinates identical.

 

Either way - I believe the reading was well within the "average" accuracy I would expect with a consumer grade GPS unit.

[+ecanderson]

Just to clarify.

 

You are talking the units calculated estimated position error? Or actual on the ground measured error?

Yes, the EPE is quite different.

 

It is interesting that the Dakota, given eons to settle, may slowly begin to approach the sort of figures I'm accustomed to seeing on the Summit HC, but never actually gets there. The Summit seems to "bottom out" at an 8 foot reading, and arrives there quickly with an unobstructed sky. The Dakota hasn't dropped below 12 feet yet, and is often substantially above that, and takes a great deal longer to reach its minimum than the Summit. Just a SWAG at what I've been seeing -- the Summit reaches best readings in 20 seconds or less, and the Dakota takes more than 2 minutes.

 

It's as though one or the other is using a more or less linear approach to the computation. Either the Summit's slope is steeper, or the Dakota's slope isn't as linear, or .. wish to heck you could do graphing or mathcad here!

[+ecanderson]

I've had both my Garmin Legend HCx units sitting within 1 foot of each other each with different accurcy readings and coordinates up to .004 from one another. I've also seen the accuracy way different and the coordinates identical.

Have you ever experimented to see if the 0.004 difference tended to be consistent - i.e., that one had a N/S or E/W axis reading that was consistently above or below the other?

 

I'm running a semi-obstructed experiment now, and will run unobstructed sky measurements later. Looking to see if the absolute positioning differences between my two units is in any way consistent... which takes us to another thread about "lemons" nearby.

[+StarBrand]

I've had both my Garmin Legend HCx units sitting within 1 foot of each other each with different accurcy readings and coordinates up to .004 from one another. I've also seen the accuracy way different and the coordinates identical.

Have you ever experimented to see if the 0.004 difference tended to be consistent - i.e., that one had a N/S or E/W axis reading that was consistently above or below the other?

 

I'm running a semi-obstructed experiment now, and will run unobstructed sky measurements later. Looking to see if the absolute positioning differences between my two units is in any way consistent... which takes us to another thread about "lemons" nearby.

nope - never any consistency between the two.

[+Tui Chub]

From my limited testing it appears that the Dakota 20 is a little more accurate than my Vista HCX.

 

Dakota 20 test

 

Pete

[+ecanderson]

From my limited testing it appears that the Dakota 20 is a little more accurate than my Vista HCX.

 

Dakota 20 test

 

Pete

Nice comparison. If the weather EVER gets back to decent around here (perhaps early next week) I plan to start a series of experiments on my Summit HC/Dakota 20 pair. Over the two years that I've had it, I've been VERY impressed with the repeatability of the Summit HC. And while time will tell what the EPE readings on the Dakota 20 really mean - and why they're so very slow to drop after power on - I've learned over 2000+ caches to respect the information that my Summit HC is providing in that regard.

 

We have in our area a special benchmark that has supposedly been studied and qualified down to a gnat's behind that will be one of the targets.