Jump to content

Gps


wcfd661

Recommended Posts

Ilive in Massachusetts, most of the caches are in wooded areas.  I've decided that I get close enough without WAAS, and since WAAS uses more battery juice, I only enable it when I'm street navigating.  One feature I'd like Garmin to include are settings profiles where you can change settings based on various profiles, ie street navigation vs geocaching.

Do you have any evidance to back that claim up? The first thing I did when I got my GPSr was to turn on the WAAS, and I can't say as I've noticed it eating up the batteries with it on.

 

When the batteries I've got in it finally "die" I think that when I put the new set in that I'll turn the WAAS off and see how long they last. Then when they "die" and I replace them I'll turn it back on and see how long they last. If there isn't any signfficant differance I'll just leave it on.

Link to comment
To free up 2 extra channels that thw WAAS sats use. I know they are lower on the horizon and perhaps less likly to get in that situation than some other sats.

Wrong. The WAAS satellites are also used for obtaining 'normal' fix positional information regardless of their WAAS capability.

Edited by Team Cotati
Link to comment
Ilive in Massachusetts, most of the caches are in wooded areas.  I've decided that I get close enough without WAAS, and since WAAS uses more battery juice, I only enable it when I'm street navigating.  One feature I'd like Garmin to include are settings profiles where you can change settings based on various profiles, ie street navigation vs geocaching.

Do you have any evidance to back that claim up? The first thing I did when I got my GPSr was to turn on the WAAS, and I can't say as I've noticed it eating up the batteries with it on.

 

When the batteries I've got in it finally "die" I think that when I put the new set in that I'll turn the WAAS off and see how long they last. Then when they "die" and I replace them I'll turn it back on and see how long they last. If there isn't any signfficant differance I'll just leave it on.

Save your energy. Even though there is no proven meaningful additional battery drain added by the enabling of WAAS, there will always be a segment of the population that will never accept the fact.

 

Since they have managed to convince themselves that WAAS actually delivers positional data that is LESS accurate than that which is obtained by using the technology, any attempt at correcting these thought patterns will ultimately fail. However, if you get extreme pleasure out of banging your head against the proverbial brick and mortar, go for it. I just worry about the Nu Bees who see this crap. :D

Edited by Team Cotati
Link to comment
Ilive in Massachusetts, most of the caches are in wooded areas.  I've decided that I get close enough without WAAS, and since WAAS uses more battery juice, I only enable it when I'm street navigating.  One feature I'd like Garmin to include are settings profiles where you can change settings based on various profiles, ie street navigation vs geocaching.

Do you have any evidance to back that claim up? The first thing I did when I got my GPSr was to turn on the WAAS, and I can't say as I've noticed it eating up the batteries with it on.

 

When the batteries I've got in it finally "die" I think that when I put the new set in that I'll turn the WAAS off and see how long they last. Then when they "die" and I replace them I'll turn it back on and see how long they last. If there isn't any signfficant differance I'll just leave it on.

Save your energy. Even though there is no proven meaningful additional battery drain added by the enabling of WAAS, there will always be a segment of the population that will never accept the fact.

 

Since they have managed to convince themselves that WAAS actually delivers positional data that is LESS accurate that that which is obtained by using that portion of the technology. That is unless you get extreme pleasure out of banging your head against the proverbial brick and mortar. I just worry about the Nu Bees who see this crap. :D

Team Cotati,

 

Ok, thanks I'll just leave it on and not worry about it.

Link to comment
To free up 2 extra channels that thw WAAS sats use. I know they are lower on the horizon and perhaps less likly to get in that situation than some other sats.

Wrong. The WAAS satellites are also used for obtaining 'normal' fix positional information regardless of their WAAS capability.

Yes, I know that. You are not listening to what I said. Being lower on the horizon those sats are more likely to be blocked by the growth and hills in the forested areas around here. Freeing up those two channels, those dedicated to just those two sats, betters the chances of pulling in other sats for a better position.

Link to comment
To free up 2 extra channels that thw WAAS sats use. I know they are lower on the horizon and perhaps less likly to get in that situation than some other sats.

Wrong. The WAAS satellites are also used for obtaining 'normal' fix positional information regardless of their WAAS capability.

Yes, I know that. You are not listening to what I said. Being lower on the horizon those sats are more likely to be blocked by the growth and hills in the forested areas around here. Freeing up those two channels, those dedicated to just those two sats, betters the chances of pulling in other sats for a better position.

Wrong. Exactly how many channels out of the 12 or 14 possible do you think are required to obtain an accurate fix?

Edited by Team Cotati
Link to comment

Digital, on your question about waas enabled position and find. No they will never be exactly the same, because you will always have some difference in the satellite geometry between the fix and the find. But yes they should be quite close.

 

On the battery usage issue. Yes, definately, there will be some extra usage of power due to the extra processing power involvled. This has been confirmed to me in the past by a Garmin tech. However, how much extra power usage? That is the open question. My feeling is it is not much. One poster in the past did some sort of test and said not much extra, but some.

 

No, I wouldn't worry about any of this. All of this is really splitting hairs. Sometimes it is fun to do that. Seems as though there are a lot of touchiness in this thread however. Including myself at times. I wish everyone was more open to learning and ferreting out facts instead of defending positions.

 

Lets learn folks!

Link to comment
To free up 2 extra channels that thw WAAS sats use. I know they are lower on the horizon and perhaps less likly to get in that situation than some other sats.

Wrong. The WAAS satellites are also used for obtaining 'normal' fix positional information regardless of their WAAS capability.

Yes, I know that. You are not listening to what I said. Being lower on the horizon those sats are more likely to be blocked by the growth and hills in the forested areas around here. Freeing up those two channels, those dedicated to just those two sats, betters the chances of pulling in other sats for a better position.

Wrong. Exactly how many channels out of the 12 or 14 possible do you think are required to obtain an accurate fix?

3 for a 2d fix, 4 for a 3d fix. The others are there and available for better position fixes, or if one in use drops out.

 

What's your point.

Link to comment
To free up 2 extra channels that thw WAAS sats use. I know they are lower on the horizon and perhaps less likly to get in that situation than some other sats.

Wrong. The WAAS satellites are also used for obtaining 'normal' fix positional information regardless of their WAAS capability.

Yes, I know that. You are not listening to what I said. Being lower on the horizon those sats are more likely to be blocked by the growth and hills in the forested areas around here. Freeing up those two channels, those dedicated to just those two sats, betters the chances of pulling in other sats for a better position.

Wrong. Exactly how many channels out of the 12 or 14 possible do you think are required to obtain an accurate fix?

3 for a 2d fix, 4 for a 3d fix. The others are there and available for better position fixes, or if one in use drops out.

 

What's your point.

Point: There is essentially zero benefit to be gained whether there are 10 or 12 channels available for satellite lock vs potentially 12 or 14 as was the implied point of the one to whom's posting I was supposedly not listening. Exactly how much greater accuracy do you believe is obtained from a 10 or 12 satellite lock vs an 8 satellite lock? :D

Edited by Team Cotati
Link to comment
Point: There is essentially zero benefit to be gained whether there are 10 or 12 channels available for satellite lock vs potentially 12 or 14 as was the implied point of the one to whom's posting I was supposedly not listening. Exactly how much greater accuracy do you believe is obtained from a 10 or 12 satellite lock vs an 8 satellite lock? :D

Following your logic manufacturers should only be making 8 channel (or less) receivers as opposed to 12 channel receivers. Numbskulls! Don't they know all those extra channels are a waste? Do't they know you only need 4 sats?

 

Availability. That's why they have 12 channels. I use to have one of those early single multiplexing revievers. What a pain. Don't make me go back!

 

Avialibilty. That's why they are planning 1 or 2 more WAAS sats over the US. Then we won't have to argue.

Link to comment
Digital, on your question about waas enabled position and find. No they will never be exactly the same, because you will always have some difference in the satellite geometry between the fix and the find. But yes they should be quite close.

 

On the battery usage issue. Yes, definately, there will be some extra usage of power due to the extra processing power involvled. This has been confirmed to me in the past by a Garmin tech. However, how much extra power usage? That is the open question. My feeling is it is not much. One poster in the past did some sort of test and said not much extra, but some.

 

No, I wouldn't worry about any of this. All of this is really splitting hairs. Sometimes it is fun to do that. Seems as though there are a lot of touchiness in this thread however. Including myself at times. I wish everyone was more open to learning and ferreting out facts instead of defending positions.

 

Lets learn folks!

EraSeek,

 

Thank you, I was wondering about that. I presume that they would be closer then say my eTrex Legend and someone else's Magelin. Or even say my eTrex Legend and say someone else's eTrex Ventura, or other model. . .

 

Ok, so once again I guess I'll just take the stand that I'll leave the WAAS, turned on and not worry about the "drain" on the batteries.

 

I can see where it can lead to that.

Link to comment

Yes, you are probably right. A legend would probably be closer than a legend and some Magellan. There is certainly hardware and firmware differences. But once again, a matter of splitting hairs. I have seen studies (plots) in the past and there are accuracy differences between most models of GPS's. Again, not huge.

 

I should correct one thing I stated earlier. It only takes 3, not 4 WAAS corrected sats before Garmin units kick in a differential position, but this would be a 2D not 3D differential. 3D usually follows quickly after.

Link to comment
Yes, you are probably right. A legend would probably be closer than a legend and some Magellan. There is certainly hardware and firmware differences. But once again, a matter of splitting hairs. I have seen studies (plots) in the past and there are accuracy differences between most models of GPS's. Again, not huge.

 

I should correct one thing I stated earlier. It only takes 3, not 4 WAAS corrected sats before Garmin units kick in a differential position, but this would be a 2D not 3D differential. 3D usually follows quickly after.

So then I guess that it's safe to presume then that if I was to go back to the site that I planted a cache in with my eTrex Legend on a different day. That it could give me a different set of coords? But still be in the "ballpark?"

Link to comment
Ilive in Massachusetts, most of the caches are in wooded areas.  I've decided that I get close enough without WAAS, and since WAAS uses more battery juice, I only enable it when I'm street navigating.  One feature I'd like Garmin to include are settings profiles where you can change settings based on various profiles, ie street navigation vs geocaching.

Do you have any evidance to back that claim up? The first thing I did when I got my GPSr was to turn on the WAAS, and I can't say as I've noticed it eating up the batteries with it on.

 

When the batteries I've got in it finally "die" I think that when I put the new set in that I'll turn the WAAS off and see how long they last. Then when they "die" and I replace them I'll turn it back on and see how long they last. If there isn't any signfficant differance I'll just leave it on.

According to Dale your Garmin GPSr can not be used in battery save mode when WAAS is enabled. So if you normaly use your Garmin with the battery save option turned on, then enabling WAAS will reduce your battery life. However, battery life with the battery save option turned off and with or without WAAS enabled is the same.

 

It isn't WAAS itself that shortens battery life on Garmins but rather the fact that the battery save option isn't available when WAAS is turned on.

Link to comment
WAAS *may* put you closer to the coordinates, but you don't know if the hider used it when they placed the cache, so it is no guarantee. They may have had an error of 100' when they marked the coordinates.

Ok, that brings up a question that I have.

 

Let's say that I go out and I hide a cache with my Garmin eTrex Legend. WITH WAAS enabled I have say a 10' accuracy reported by the unit. Does that mean that IF another cacher goes out and THEY have a Garmin eTrex Legend also with WAAS enabled and they're given a 10' accuracy. That they'll find my cache at the exact location that I posted?

The satellite configuration won't make much, if any, difference, since the estimated positional error (EPE) takes that into account (we're assuming the same EPE here, therefore an equally valid satellite configuration). There may be some difference in how accurate the EPE is between brands, but this is really splitting hairs. However, there is another, more basic reason why the same EPE does not necessarily lead to exactly the same position.

 

Think of your 10' EPE as a circle with a 10' radius. You and the cache are at the center of that circle, but the coordinates you're marking could actually be 10' in any direction from where you're standing. Let's say the coordinates you mark are really 10' to the east of where you're standing.

 

Along comes the seeker, who also has a 10' EPE. He's at the center of a circle with a 10' radius and your given coordinates (not the cache) are somewhere within that circle. Let's say he's 10' east of the real position of the coordinates. In this worst case scenario, when your two EPE circles are just touching at the coordinates, the seeker will be 20' feet from the cache (10' from the cache to the coordinates plus 10' from the cordinates to the seeker).

 

In all other scenarios the two EPE circles will overlap and the seeker will be less than 20' from the cache. Only if the two EPE circles overlap completely will he be at the exact location you posted.

 

EPE is, if course, only estimated, so the error could be greater 10'. No wonder it's so hard to find those caches!! :D

 

Edit: fix a typo and clarify first sentence.

Edited by Kai Team
Link to comment
...

Think of your 10' EPE as a circle with a 10' radius. You and the cache are at the center of that circle, but the coordinates you're marking could actually be 10' in any direction from where you're standing. Let's say the coordinates you mark are really 10' to the east of where you're standing.

 

Along comes the seeker, who also has a 10' EPE. He's at the center of a circle with a 10' radius and your given coordinates (not the cache) are somewhere within that circle. Let's say he's 10' east of the real position of the coordinates. In this worst case scenario, when your two EPE circles are just touching at the coordinates, the seeker will be 20' feet from the cache (10' from the cache to the coordinates plus 10' from the cordinates to the seeker).

 

In all other scenarios the two EPE circles will overlap and the seeker will be less than 20' from the cache. Only if the two EPE circles overlap completely will he be at the exact location you posted....

I think you are making an error here, but I'm not sure. Where's fizzymagic when I need him?

Link to comment
So then I guess that it's safe to presume then that if I was to go back to the site that I planted a cache in with my eTrex Legend on a different day. That it could give me a different set of coords? But still be in the "ballpark?"

Yes.

 

By the way, all this talk about EPE or Accruacy reading, understand that these are just programmed guesstimates. They are a good judge of how your GPS is doing but to think they are a part of the system solution is incorrect. Magellans method and Garmins method of the guesstimate will be quite different and even between models of the same brand it will vary.

 

The best way to get a good coordinate when planting a cache is to watch this reading and wait for it to get as low as possible before marking the position. Do this several times and pick what seems the most consistant.

Link to comment
The best way to get a good coordinate when planting a cache is to watch this reading and wait for it to get as low as possible before marking the position. Do this several times and pick what seems the most consistant.

Something we all agree on!!! (I think). :D

Link to comment
...

Think of your 10' EPE as a circle with a 10' radius. You and the cache are at the center of that circle, but the coordinates you're marking could actually be 10' in any direction from where you're standing.  Let's say the coordinates you mark are really 10' to the east of where you're standing.

 

Along comes the seeker, who also has a 10' EPE. He's at the center of a circle with a 10' radius and your given coordinates (not the cache) are somewhere within that circle.  Let's say he's 10' east of the real position of the coordinates.  In this worst case scenario, when your two EPE circles are just touching at the coordinates, the seeker will be 20' feet from the cache (10' from the cache to the coordinates plus 10' from the cordinates to the seeker).

 

In all other scenarios the two EPE circles will overlap and the seeker will be less than 20' from the cache. Only if the two EPE circles overlap completely will he be at the exact location you posted....

I think you are making an error here, but I'm not sure. Where's fizzymagic when I need him?

I've found EPE difficult to explain it terms most people will understand but here's my article explaining EPE: http://factsfacts.com/epe.htm

Link to comment
According to Dale your Garmin GPSr can not be used in battery save mode when WAAS is enabled. So if you normaly use your Garmin with the battery save option turned on, then enabling WAAS will reduce your battery life. However, battery life with the battery save option turned off and with or without WAAS enabled is the same.

If all that's so, it explains why the battery life dispute is baffling to me, a Magellan user. Apparently it's a Garmin thing.

Link to comment

WAAS will not work with Battery saver mode turned on, because with saver mode your unit's processing has been reduced to save power. Your unit updates much less frequently. WAAS requires full processing power and the ability to update at the normal rate. Yes, this takes more power! But someone must know how much more than normal it requires. I don't. I know it is much less than running with your backlight on. I know it is more than running with out WAAS on. Question is how much. If you use rechargable batteries who cares? I'd just like to know anyway.

Link to comment

ok so Ironwoman broke down and bought me a magellan eXplorist 200 for an early birthday present and 4 rechargeablebatteries. she really wanted to get me a garmin etrax legend but best buy was sold out. this one works just fine for what I need it for...thanks any way SBELL! I really do appreciate your help. :D

Edited by wreckelite
Link to comment
...

Think of your 10' EPE as a circle with a 10' radius. You and the cache are at the center of that circle, but the coordinates you're marking could actually be 10' in any direction from where you're standing.  Let's say the coordinates you mark are really 10' to the east of where you're standing.

 

Along comes the seeker, who also has a 10' EPE. He's at the center of a circle with a 10' radius and your given coordinates (not the cache) are somewhere within that circle.  Let's say he's 10' east of the real position of the coordinates.  In this worst case scenario, when your two EPE circles are just touching at the coordinates, the seeker will be 20' feet from the cache (10' from the cache to the coordinates plus 10' from the cordinates to the seeker).

 

In all other scenarios the two EPE circles will overlap and the seeker will be less than 20' from the cache. Only if the two EPE circles overlap completely will he be at the exact location you posted....

I think you are making an error here, but I'm not sure. Where's fizzymagic when I need him?

sbell111 - if you see an error, please say what you think the error is so we can discuss it. If you don't understand my post or aren't sure, please ask a question.

 

Thot - are you agreeing or disagreeing? Does the example I gave make sense to you?

Link to comment
Point: There is essentially zero benefit to be gained whether there are 10 or 12 channels available for satellite lock vs potentially 12 or 14 as was the implied point of the one to whom's posting I was supposedly not listening. Exactly how much greater accuracy do you believe is obtained from a 10 or 12 satellite lock vs an 8  satellite lock?  :D

Following your logic manufacturers should only be making 8 channel (or less) receivers as opposed to 12 channel receivers. Numbskulls! Don't they know all those extra channels are a waste? Do't they know you only need 4 sats?

 

Availability. That's why they have 12 channels. I use to have one of those early single multiplexing revievers. What a pain. Don't make me go back!

 

Avialibilty. That's why they are planning 1 or 2 more WAAS sats over the US. Then we won't have to argue.

What I have read so far seems to indicate that even though modern GPSr units can track up to 12 satellites, no more than 4 are used in the solution equation.

Link to comment
I think you are making an error here, but I'm not sure. Where's fizzymagic when I need him?

Here I am!

 

EPE circles are not hard circles of uniform uncertainty; the uncertainty in the position is a probability distribution that is peaked at the center and decreases with distance. But remember that the available area goes up as the distance from the center gets larger, too.

 

EPE is an estimate of the positional accuracy of the receiver. It generally is set at the 90% confidence level; supposedly, if your EPE is 10 feet, there is a 90% chance that you are within 10 feet of the displayed coordinates. However, EPE is tricky to calculate reliably and includes systematic errors, so take it with a large grain of salt.

 

My experience has been that if I average for a while (my Magellan units average automatically when they are at rest) and move away from and back to the position of a geocache, taking waypoints each time and averaging those waypoints, I can get coordinates that are reproducible to better than 10 feet most of the time. I think that is about the best you can expect from current GPS technology.

Link to comment
What I have read so far seems to indicate that even though modern GPSr units can track up to 12 satellites, no more than 4 are used in the solution equation.

Your take on what you have read is incorrect. GPS receivers are generally capable of using all 12 channels in a position solution.

Link to comment
What I have read so far seems to indicate that even though modern GPSr units can track up to 12 satellites, no more than 4 are used in the solution equation.

Your take on what you have read is incorrect. GPS receivers are generally capable of using all 12 channels in a position solution.

Where did you find this information?

Link to comment
What I have read so far seems to indicate that even though modern GPSr units can track up to 12 satellites, no more than 4 are used in the solution equation.

Your take on what you have read is incorrect. GPS receivers are generally capable of using all 12 channels in a position solution.

Where did you find this information?

Huh? I know how GPS units caculate their positions. I didn't "find" the information anywhere; I learned it from many sources over a long time.

 

One very simplified explanation of how more than 4 satellites are used to improve accuracy is given here. A fairly straightforward Web search can find you many more.

Link to comment

I think the misunderstanding is based on the fact that a GPSr needs a lock on at least four sats to obtain a 3D solution. However, the more sats you have a lock on, the better your solution is going to be. As I understand it, this is because, while you may have a lock on four or five sats, their 'constellation' may be such that they do not give you a very accurate solution.

 

Think of it this way... You are searching for an urban micro on New York City. The only sats you can get a lock on are in the small sliver of sky running in a straight line. Obviously, your solution is going to be poorer than if you had a lock on the same number of sats arranged over a larger part of the sky. Now, consider locking onto more and more sats, thereby giving your GPSr more info to compute your location. Certainly, this would yield greater accuracy.

 

BTW, heres something I've learned from roaming around the forums. If you have a question regarding math, ask fizzymagic. Also, if fizzymagic makes a statement involving mathmatics, take his word for it. I'll bet you that he's right.

Link to comment

Here's the question I emailed Garmin tech support on Friday:

Are there any circumstances where using WAAS degrades position accuracy?  If so, what would those circumstances be? Thanks!

and here's the reply I received today:

Thank you for Contacting Garmin Product Support.

 

WAAS will either increase the accuracy or will not use WAAS at all.

 

If you have further questions, please feel free to contact me!

Best Regards,

Chris Bishop

Product Support Specialist

Garmin International

 

I stand by my original statement:

WAAS will *always* put you closer, assuming you can pick up the WAAS signal.

It's clear from this discussion that added battery drain from WAAS is also not significant.

 

Abandon the urban myths: Turn WAAS on and leave it on. It may not make things better, but it won't make things worse! :P

Link to comment
I think you are making an error here, but I'm not sure.  Where's fizzymagic when I need him?

Here I am!

 

EPE circles are not hard circles of uniform uncertainty; the uncertainty in the position is a probability distribution that is peaked at the center and decreases with distance. But remember that the available area goes up as the distance from the center gets larger, too.

Fizzy - I was providing a simplified explanation of how to interpret EPE - which is why I said "Think of your 10' EPE as a circle with a 10' radius." Describing EPE as a circle is easier to understand and visualize, and I don't think it misleads anyone for all practical purposes.

 

EPE is an estimate of the positional accuracy of the receiver.  It generally is set at the 90% confidence level; supposedly, if your EPE is 10 feet, there is a 90% chance that you are within 10 feet of the displayed coordinates.  However, EPE is tricky to calculate reliably and includes systematic errors, so take it with a large grain of salt.

Yup, that's why I said, "EPE is, if course, only estimated, so the error could be greater than 10'"

 

I don't see anything in your post that contradicts my explanation from a practical point of view. Correct me if I'm wrong - I don't want to use that explanation again if it's misleading, and it's obvious that you know a lot about this stuff! Thanks.

Link to comment
What I have read so far seems to indicate that even though modern GPSr units can track up to 12 satellites, no more than 4 are used in the solution equation.

Your take on what you have read is incorrect. GPS receivers are generally capable of using all 12 channels in a position solution.

Interesting. Perhaps I have to unlearn something here if you are right. I do seem to remember reading that those (beyond the best 4) not used in a position fix stand ready to be used and the unit will pop back and forth across the sats using the best four. Perhaps I got this from the old days and my 45xl manual, or perhaps just the knowledge that you need the 4 and a presumsion on my part. If all are actually being used in a position fix, than that again bring up the question in my mind "do they mix WAAS and NON-WAAS sats?"

 

Thanks. More to think about and research. :P

Link to comment
If all are actually being used in a position fix, than that again bring up the question in my mind "do they mix WAAS and NON-WAAS sats?"

In this context, there are two ways to interpret the clarification from Garmin:

 

1) WAAS and non-WAAS satellites are both used (i.e. "mixed"), or

2) If using WAAS would result in a larger EPE (due to fewer satellites or a poorer configuration), the GPSr will not use WAAS.

 

Bottom line: with WAAS turned on, your GPSr will give you the best available fix, whether that entails using WAAS or not. The mixing of satellites seems to be a purely intellectual question, with no practical effect on your EPE.

Link to comment
EPE is an estimate of the positional accuracy of the receiver.  It generally is set at the 90% confidence level; supposedly, if your EPE is 10 feet, there is a 90% chance that you are within 10 feet of the displayed coordinates. 

I'm not saying you're wrong. All I know is what I’ve read on various websites, but they suggest the confidence level for EPE is much less than 90%.

 

Take this article for example. It says:

Garmin's EPE readout is generally accepted to indicate that there is an EQUAL probability that the error is GREATER or LESS THAN the indicated EPE.  This is the 50% CEP value given above.  [. . .] to be 95% confident that your measurement is within a circle of a fixed radius,  you would have to multiply Garmin's EPE value by two.  [. . .] Magellan's EPE numbers appear to be even more optimistic (maybe the 1 sigma value or even lower) while Lowrance seems to be someplace between the RMS and 2 sigma values.

This is but one of several web articles I’ve seen that suggest EPE confidence levels are more like one sigma than 90%.

 

Edited to add:

 

This article says:

 

On the new Garmins (GPSMap 76, 76s's and V's, etc.) the EPE is approximately at . . . a 68% (or 1 sigma) confidence level when you are using [WAAS or NDGPS] corrections.

And this one says:

 

t is often not known whether the EPE is based on a 50 % confidence interval, the route mean square error (68 % interval) or any other confidence interval. Given the above, the non-expert GPS user has no effective means to obtain insight in the accuracy of the handheld GPS equipment using it at a particular location and time.
Edited by Thot
Link to comment

Looks like perhaps Fizzymagic and I are both right:

 

"DOP is an indicator of the quality of the geometry of the satellite constellation. Your computed position can vary depending on which satellites you use for the measurement. Different satellite geometries can magnify or lessen the errors in the error budget described above. A greater angle between the satellites lowers the DOP, and provides a better measurement. A higher DOP indicates poor satellite geometry, and an inferior measurement cofiguration.

 

 

 

Some GPS receivers can analyze the positions of the satellites available, based upon the almanac, and choose those satellites with the best geometry in order to make the DOP as low as possible. Another important GPS receiver feature is to be able to ignore or eliminate GPS readings with DOP values that exceed user-defined limits. Other GPS receivers may have the ability to use all of the satellites in view, thus minimizing the DOP as much as possible. "

 

Here is the link: http://www.cmtinc.com/gpsbook/index.htm#chap2

Edited by EraSeek
Link to comment
Thot - are you agreeing or disagreeing?  Does the example I gave make sense to you?

If it's taken as an oversimplified explanation that may lead to a misconception that's one thing. It's always difficult to decide when to risk misleading in order to simplify and clarify an explanation.

 

The difficulty with your explanation is that it sounds like the 10 foot circle is a maximum limit to how far off the reading could be. In fact, it’s some kind of average distance – maybe less, maybe more.

 

When you say:

 

He's at the center of a circle with a 10' radius and your given coordinates (not the cache) are somewhere within that circle. 

It sounds like it's certain that your coordinates are in the circle and it isn’t. There's some likelyhood/probability that it's within 10'. The probability that it’s in the 10 foot circle depends on whether my 67% probability is correct or fizzymagic’s 90% is correct. All this is what makes it so hard to explain EPE and why I keep rewriting my article attempting to make it more understandable.

 

Also, a smaller problem -- even if your 10 foot circle were absolute the combination of your error circle and his would be less than 20 feet. But, I think your 20 feet is an acceptable simplifying error in this case.

 

I think all the issues in this thread (WAAS help or harm, battery drain or not, number of birds used, meaning of EPE) are all very healthy discussions I think have been a long time in coming, I somehow fear it may terrify newbies, if they think they need to understand any or all of this.

Edited by Thot
Link to comment
Interesting. Perhaps I have to unlearn something here if you are right. I do seem to remember reading that those (beyond the best 4) not used in a position fix stand ready to be used and the unit will pop back and forth across the sats using the best four. Perhaps I got this from the old days and my 45xl manual, or perhaps just the knowledge that you need the 4 and a presumsion on my part. If all are actually being used in a position fix, than that again bring up the question in my mind "do they mix WAAS and NON-WAAS sats?"

I realize I was somewhat curt in my previous response. In my defense, I am visiting the in-laws this weekend. :P

 

I suppose that for a scanning receiver, one might want to minimize the number of channels used in the solution to speed up the updates; in that case, I can perhaps imagine a receiver that would use the 4 "best" satellites.

 

But for a 12-channel receiver, there is no reason not to use as many satellites as you can see. The N satellites give N pseudoranges, which form an overdetermined system with N-3 independent measurements of the position. Those measurements can then be combined in a pretty straightforward maximum-likelihood manner to optimize the reported position error using all the satellites. The Neyman-Pearson lemma can be used to prove that the resulting solution is guaranteed to be statistically better than any of the 4-satellite solutions.

 

The maximum-likelihood method used is almost certainly some variant of least squares, which is the maximum likelihood estimator for Gaussian uncertainties. The trick to getting the optimum solution is in determining the weights used for each of the independent solutions; I recall having seen several papers proposing different weighting schemes. I would guess that most would use the DOP to weight the multiple solutions, so that those solutions with the minimum DOP have the highest weights. But the weighting might also include signal strength, WAAS corrections, and other factors.

 

In this formulation, it is easy to see how WAAS and non-WAAS satellites can be combined to give an optimal solution. Since the math is (relatively) straighforward, IMNSHO there is no reason that they shouldn't all be used in every receiver.

 

Does that make sense to anybody?

Link to comment
I don't see anything in your post that contradicts my explanation from a practical point of view.

I wouldn't say that I contradict you; I just think your explanation leads to an inaccurate mental model.

 

Maybe this would be more helpful:

 

If the uncertainties of the two receivers in question are ( a ) independent and ( b ) roughly Gaussian, then the uncertainty in the difference is simply the two added in quadrature. In addition, the Gaussian distribution has the unique property that Gaussians in linear dimensions can be combined to form a Gaussian in the radial dimension.

 

Translated into English, that means that if the person who hid the cache had an EPE of 10 feet and the searcher also has an EPE of 10 feet, then the expected error in finding the cache would be sqrt(10*10 + 10*10) = 14 feet.

 

If the person hiding the cache had an EPE of 7 feet and the finder has an EPE of 20 feet, then the expected error is sqrt(7*7 + 20*20) = 21 feet.

 

That's why it is so important to get good coords when hiding a cache.

Edited by fizzymagic
Link to comment

 

Interesting. Perhaps I have to unlearn something here if you are right. I do seem to remember reading that those (beyond the best 4) not used in a position fix stand ready to be used and the unit will pop back and forth across the sats using the best four. Perhaps I got this from the old days and my 45xl manual, or perhaps just the knowledge that you need the 4 and a presumsion on my part.

<snip>

<snip>

I suppose that for a scanning receiver, one might want to minimize the number of channels used in the solution to speed up the updates; in that case, I can perhaps imagine a receiver that would use the 4 "best" satellites.

 

But for a 12-channel receiver, there is no reason not to use as many satellites as you can see. The N satellites give N pseudoranges, which form an overdetermined system with N-3 independent measurements of the position. Those measurements can then be combined in a pretty straightforward maximum-likelihood manner to optimize the reported position error using all the satellites. The Neyman-Pearson lemma can be used to prove that the resulting solution is guaranteed to be statistically better than any of the 4-satellite solutions.

Just because it is better doesn't mean that a manufacturer is going to build something that way. Particularly the base models.

The maximum-likelihood method used is almost certainly some variant of least squares, which is the maximum likelihood estimator for Gaussian uncertainties. The trick to getting the optimum solution is in determining the weights used for each of the independent solutions; I recall having seen several papers proposing different weighting schemes. I would guess that most would use the DOP to weight the multiple solutions, so that those solutions with the minimum DOP have the highest weights. But the weighting might also include signal strength, WAAS corrections, and other factors.

 

In this formulation, it is easy to see how WAAS and non-WAAS satellites can be combined to give an optimal solution. Since the math is (relatively) straighforward, IMNSHO there is no reason that they shouldn't all be used in every receiver.

True, until you factor in the marketing department. Maybe, so that they can offer better accuracy on the more expensive models.

<snip>

 

I did a little digging around the w3 today for the articles I remember reading and this is what I came up with.

 

About.com has this to say, "...the 12-channel receivers sold today use each channel to individually lock onto and hold a satellite. The best four signals are used to provide accurate, reliable information almost instantaneously.

 

I found that flightsimaviation.com had this to say, "The receiver uses data from the best four satellites about its horizon, adding signals from one as it drops signals from another... "

 

And on wlsar.org.nz I found this under Satellite selection: "Selects the best four satellites. Those that are below the horizon, or soon will be, are excluded and, if possible, any that form a poor geometrical solution are also excluded."

 

Then I happened upon this at gpsinformation.net: "like other Garmin 12 channel units) uses an over determined solution when more than 4 SVs are locked. This produces somewhat higher accuracy than units which use only the best four SVs."

Did I read that right? Than units which use only the best four. Which lead me to search some more and i found this web page about

three kinds of receivers. Multiplexing receivers, sequential receivers, and parallel receivers.

 

So, I guess, it all depends on how your GPSr is built if it is using the best four method or if it uses all available satellites.

 

<snip>

BTW, heres something I've learned from roaming around the forums. If you have a question regarding math, ask fizzymagic. Also, if fizzymagic makes a statement involving mathmatics, take his word for it. I'll bet you that he's right.

Thanks, I'll remember that when I have a mathematics question.

Link to comment
Thot - are you agreeing or disagreeing?  Does the example I gave make sense to you?

If it's taken as an oversimplified explanation that may lead to a misconception that's one thing. It's always difficult to decide when to risk misleading in order to simplify and clarify an explanation.

I wouldn't say that I contradict you; I just think your explanation leads to an inaccurate mental model.

 

Maybe this would be more helpful:

 

If the uncertainties of the two receivers in question are ( a ) independent and ( b ) roughly Gaussian, then the uncertainty in the difference is simply the two added in quadrature. In addition, the Gaussian distribution has the unique property that Gaussians in linear dimensions can be combined to form a Gaussian in the radial dimension.

 

Translated into English, that means that if the person who hid the cache had an EPE of 10 feet and the searcher also has an EPE of 10 feet, then the expected error in finding the cache would be sqrt(10*10 + 10*10) = 14 feet.

 

If the person hiding the cache had an EPE of 7 feet and the finder has an EPE of 20 feet, then the expected error is sqrt(7*7 + 20*20) = 21 feet.

 

That's why it is so important to get good coords when hiding a cache. 

 

Both points very well taken (thanks for the clarification of the statistical reasoning, FizzyMagic - it is more helpful and makes sense to me).

 

I think all the issues in this thread (WAAS help or harm, battery drain or not, number of birds used, meaning of EPE) are all very healthy discussions I think have been a long time in coming, I somehow fear it may terrify newbies, if they think they need to understand any or all of this.

That's why I was trying to (over)simplify!! :D It is difficult to decide when simplification for clarity sake does more harm than good. I would probably use my explanation again in a "getting started" forum, but with an appropriate caveat about oversimplification (and perhaps including FizzyMagic's sum of the squares piece). Clearly this thread has far exceeded the getting started discussion level (newbies - stop now before your head explodes! :lol: ), though I agree with Thot that it's a very good discussion!

 

Edit: Typo

Edited by Kai Team
Link to comment

I guess it depends on what type of "newbie" is reading this...

 

As a newbie on this particular topic, I found the discussion to be quite informative. My head was exploding only when the discussion was circling (like a non-WAAS GPSr) instead of coming to a point. :lol:

 

In the field, my brother consistenly finds caches hidden by a certain hider who posts "spot-on" coordinates before I do. He has a GPSmap 60c with WAAS-enabled, and I have a non-WAAS eTrex Summit.

Link to comment

Well personaly I've learned something here. Even though the thread was a bit rancorous for awhile. Always willing to learn.

 

I checked with an outside source of mine and they stated , that yes, manufacturers use both methods but are very secretive about how there units are put together. One problem with using an "all sats" solution is that if you are using too many in the solution it can slow the processing of the unit, and if not masking low sats on the horizon it will degrade the position.

 

So I guess there are various methods and much of it is trade secret stuff, just as how they come up with the EPE is. And yes, I'm sure whatever sales often plays a role.

Link to comment
So, I guess, it all depends on how your GPSr is built if it is using the best four method or if it uses all available satellites.

If you check the date on that article, you will note that it was written in 1998, when 12-channel parallel receivers were still relatively expensive and new.

 

Today, even your base GPS is a 12-channel parallel receiver, and so all the reasons for using the "best four satellites" algorithm are moot. In my statements about the overdetermined solution, I was deliberately ignoring obsolete hardware. IMO, both sequential and multiplexing GPS units are obsolete.

 

The accuracy improvement from using an overdetermined solution is greater now that selective availability has been turned off, so I think that you will find that all newer units will use that method.

 

It may be that Garmin's method of including only those satellites for which complete WAAS data was known in the solution dates from the days of scanning receivers; nonetheless, I stand by my original contention that the algorithm was defective.

Link to comment
So, I guess, it all depends on how your GPSr is built if it is using the best four method or if it uses all available satellites.

If you check the date on that article, you will note that it was written in 1998, when 12-channel parallel receivers were still relatively expensive and new.

<snip>

I checked all the webpages and did notice that the oldest page, writen in 1998, is the only one that mentions 12-channel parallel receivers. The page written in 2000 aludes to more than 4 satellites being used in the solution. The 2005 page only references the best 4 method. The other two pages I couldn't locate dates but About.com likes to be up to date and that page also only references the best 4 menthod. You would have me believe that this would means newer GPSrs use only the best 4 menthod. :D

 

I do find it interesting that the oldest page I found has more relevant information about current technology than the newer pages.

 

<snip>

Today, even your base GPS is a 12-channel parallel receiver, and so all the reasons for using the "best four satellites" algorithm are moot.  In my statements about the overdetermined solution, I was deliberately ignoring obsolete hardware.  IMO, both sequential and multiplexing GPS units are obsolete.

<snip>

 

Just thinking out loud. I wonder why the best four method is still being referenced in articles written even today. Sometimes it is the only method given.

Link to comment
...Just thinking out loud. I wonder why the best four method is still being referenced in articles written even today. Sometimes it is the only method given.

I bet it has something to do with EraSeek's quote above:

 

manufacturers ... are very secretive about how there units are put together.

 

edit: Obviously, I spindled his post a bit because I was only extracting one of his post's essential points.

Edited by sbell111
Link to comment
BTW, heres something I've learned from roaming around the forums.  If you have a question regarding math, ask fizzymagic.  Also, if fizzymagic makes a statement involving mathmatics, take his word for it.  I'll bet you that he's right.

fizzymagic is a physicist and a magician (hence his geocaching name), what does he know about math? :yikes: Amazingly, I think his brother may know more about math. That's scary.

 

--Marky

 

P.S. My 14 channel eXplorist 600 kicks butt under all the extreme conditions I've used it in (even though it uses a patch antenna). With all this talk about 12 channels, I thought I'd throw that in. :laughing:

Link to comment

A 14 channel should do better than others under cover. It has 12 regular gps channels and 2 dedicated WAAS channels (as opposed to 10 and 2 switchable). More oportunity to find those sats when walking thru the trees.

 

Here is something of interest I found on a Dale DePriest site:

" On 12 channel parallel units Garmin has developed a technique that permits them to calculate an overdetermined solution by using more than the standard 4 satellites. This can lead to more accuracy, reported to be on the order of 15%, and permits a smoother transition when a satellite drops out of view for any reason. The G-12 family indicates exactly which satellites are currently being used in the calculation by showing black bars. Satellites may be dropped from the solution for a number of reasons such as the detection of significant multipath reception on one satellite."

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...