GPS Tech...accuracy

[keithbradley]

Ok, I'm having a little trouble making sense of this. I have been using my auto GPS (Garmin Nuvi) for a long time now. The only problem is it's large, so it's difficult to use in smaller boats.
I also have a Garmin Forerunner 305, which is a watch/heart rate monitor/GPS unit. I tried this unit in my cars and found it was terribly inaccurate. It seemed like the refresh rate was too slow for RC use. When I used it in my car and compared it to my other GPS, it was accurate, but I had to drive a pretty lengthy distance before it got up to speed. Using it in RC cars netted a low speed reading, relative to the distance traveled under speed.

So here's where Im confused. While considering a new GPS unit for RCs, I checked out some specs. The forerunner 201, which is used succesfully by quite a few members here, says:
Accuracy
Position - 49ft
Velocity - 0.16ft/sec

My forerunner 305, that didnt work worth a crap, says:
Accuracy
Position - 33ft
Velocity - 0.16ft/sec

Both had refresh rates of 1hz.

Anyone make any sense of this? If I am understanding those specs correctly, I would assume my watch would be MORE accurate...so why doesn't it work worth a crap?
What should I get? I want something $150 or under...

[6sHyper]

I have no clue what those rates mean i'm not gonna lie about that haha But what i can tell you is that my forerunner 201 has been VERY consistant for me. I've literally only had one reading that i knew was trash, other than that they always come back in run after run within 2-3mph up or down from each other and as i have propped my boats up it seems to pick up the etxra mph or 2 with what i believe to be very good accuracy.

When i mean consistant an example would be in my DF33 i started with an x445 and did 2 runs, one at 50.3 and the other at 51.8, switched to m447 and did 2 runs with 53.1 and 54.8. Lastly i put on an x645 and thats what i run now and i've never had a reading under 55 and never over 58.1. pretty damn consistant if you ask me.

[m4a1usr]

Its a bit more technical then price or product statistics. I too bought into GPS accuracy for many years. Its like selling a bridge in Florida. Then I did a bit of research, spoke to some very knowledgable folks who do software programming for GPS systems (government types) then looked at the very simple technology under which GPS operates under and the answer so darn simple. Speed accuracy can only be determined using differential receivers. Typical WAAS accuracy, using 4 satellites locked on to the Rx, has horrible accuracy providing all 4 sat's stay locked on during the data measured! The variation is typicaly greater then 10% over the measured distance when just one sat is lost! And thats the accepted norm. Lose 2 sat's and the data becomes worse.

Sounds like you have already discovered some of the pit falls. Claiming a certain speed based on the particular GPS unit you purchased only makes the user feel good. Its not accurate enough to be used in a documented format. Has no significant value to be offered in a technical document using empirical data. Its totaly a joke in the scientific community. If I were to present a vehicle performance data package based upon WAAS data as my basis to my colleagues' for hardware performance and could not demonstrate the sat lock during the entire value measured I would be laughed at or worse. No longer would they trust my judgment. Worse. The technical guys who do know GPS would not want me on their technical teams.

Folks, if you have to believe something dont use GPS as your tool for performance. It shows how little you know. Use the very simple tool of speed Vs time over a measured distance. Its proven. Its accurate. It does not lie. More importantly its a standard. GPS is not.

John

[keithbradley]

Its a bit more technical then price or product statistics. I too bought into GPS accuracy for many years. Its like selling a bridge in Florida. Then I did a bit of research, spoke to some very knowledgable folks who do software programming for GPS systems (government types) then looked at the very simple technology under which GPS operates under and the answer so darn simple. Speed accuracy can only be determined using differential receivers. Typical WAAS accuracy, using 4 satellites locked on to the Rx, has horrible accuracy providing all 4 sat's stay locked on during the data measured! The variation is typicaly greater then 10% over the measured distance when just one sat is lost! And thats the accepted norm. Lose 2 sat's and the data becomes worse.

Sounds like you have already discovered some of the pit falls. Claiming a certain speed based on the particular GPS unit you purchased only makes the user feel good. Its not accurate enough to be used in a documented format. Has no significant value to be offered in a technical document using empirical data. Its totaly a joke in the scientific community. If I were to present a vehicle performance data package based upon WAAS data as my basis to my colleagues' for hardware performance and could not demonstrate the sat lock during the entire value measured I would be laughed at or worse. No longer would they trust my judgment. Worse. The technical guys who do know GPS would not want me on their technical teams.

Folks, if you have to believe something dont use GPS as your tool for performance. It shows how little you know. Use the very simple tool of speed Vs time over a measured distance. Its proven. Its accurate. It does not lie. More importantly its a standard. GPS is not.

John

I was hoping you would chime in on this John. I realize GPS data cannot be relied on as any sort of scientific "proof". I can however, use my auto GPS to compare my RCs quite reliably. For instance, I could put my GPS in my brushless truggy with a fully charge 5s lipo and run it down the road in fron tof my house. I could absolutely guarantee that it comes back and says 53mph top speed. Ive done it a number of times and its absolutely consistant. I understand loosing satellite signal, or certain sudden movements (like if I flipped it) could cause an innaccurate reading, but in a somewhat controlled environment (same place, smooth run, low cloud cover, ect.) it will consistantly read the same.

Now with my forerunner 305, I could strap it in that same truggy, run it, and it might say 23mph. I could run it further, staying at speed for an extended period, and it might read a bit higher, but still not up to speed. The same can be seen in a my real car. I can drive 40 mph steady down the road, and the watch will not get up to 40mph on the readout for maybe 1/2 mile. Its just SLUGGISH.

Is there a way I can compare GPS units and get an idea how well they will work in this respect?

[m4a1usr]

What it really boils down to is a Rx sensitivity issue Vs software calculations. In looking at the commercial units currently produced most manufacturers are going to only offer accuracy in the feet per sampling time frame. Some are in meters! So its a distance Vs time measurement calculation. And since most GPS unts are designed or directed towards hikers, caching types, navigation (fishing locations), global positioning, distances in a given time frame, top speed is not one of their priorities for reliable accuracy. Doesnt mean you cant get repeatable measurements with an indication of verifiable results. But thats the crux or weakness of the performance offered. Enough to convince one that its pretty darn close, maybe even dead on in the right conditions, however just enough variable to not be trusted 100% of the time in a sampling period.

Heck if it was that easy we could use $100 GPS units tied into a cruise missile and not have the tax payer lay out $10 to 25k for weapon performance accuracy. But it doesnt work that way in reality. When it has to be right, you use the tool capable of performing the job. Trust me when I say. If we used WAAS technology to deliver weapon systems. We would be trying to bombing Libya but laying down explosives in Egypt. That would be fun to explain.


John

[G1ST]

Does Eagle Tree use the same technology as Garmin in their GPS unit?
Does it sample at a higher rate? Is it likely to be more accurate?
Greg

[Brushless55]

Its a bit more technical then price or product statistics. I too bought into GPS accuracy for many years. Its like selling a bridge in Florida. Then I did a bit of research, spoke to some very knowledgable folks who do software programming for GPS systems (government types) then looked at the very simple technology under which GPS operates under and the answer so darn simple. Speed accuracy can only be determined using differential receivers. Typical WAAS accuracy, using 4 satellites locked on to the Rx, has horrible accuracy providing all 4 sat's stay locked on during the data measured! The variation is typicaly greater then 10% over the measured distance when just one sat is lost! And thats the accepted norm. Lose 2 sat's and the data becomes worse.

Sounds like you have already discovered some of the pit falls. Claiming a certain speed based on the particular GPS unit you purchased only makes the user feel good. Its not accurate enough to be used in a documented format. Has no significant value to be offered in a technical document using empirical data. Its totaly a joke in the scientific community. If I were to present a vehicle performance data package based upon WAAS data as my basis to my colleagues' for hardware performance and could not demonstrate the sat lock during the entire value measured I would be laughed at or worse. No longer would they trust my judgment. Worse. The technical guys who do know GPS would not want me on their technical teams.

Folks, if you have to believe something dont use GPS as your tool for performance. It shows how little you know. Use the very simple tool of speed Vs time over a measured distance. Its proven. Its accurate. It does not lie. More importantly its a standard. GPS is not.

John

good stuff bro :thumbup1:

[m4a1usr]

Does Eagle Tree use the same technology as Garmin in their GPS unit?
Does it sample at a higher rate? Is it likely to be more accurate?
Greg

Thats beyond me to answer. One of the basic rules to remember is that when software algorithms use linear values to determine a calculation its typically more reliable for the end value. But since things like acceleration are not linear and not a constant the algorithm adjusts in the short measurement time frame. So the end calculation has an error rate during it measurement. That value is determined by the initial accuracy determined by the Rx. Sort of a inherent weakness to the operating system. So if you get an lock on with 4 sat's and the reading says accuracy of 24 feet using one particular GPS model, but have another GPS model that says 4 sat's with an accuracy of 9 feet. The difference is not 1/3rd (approx) of error. Its worse. How much? I dont know. Its just a fact that its not linear. Due to the variables.

I know it sounds like mad scientist hoohaa but its truely not. The sad fact is I'm not the guy to offer an accurate scientific answer. But that doesnt change the fact that there indeed exists an error in the measurement. Believe what you want. We are all adults. Just my experience and of course 2 cents.

John

[keithbradley]

Thats beyond me to answer. One of the basic rules to remember is that when software algorithms use linear values to determine a calculation its typically more reliable for the end value. But since things like acceleration are not linear and not a constant the algorithm adjusts in the short measurement time frame. So the end calculation has an error rate during it measurement. That value is determined by the initial accuracy determined by the Rx. Sort of a inherent weakness to the operating system. So if you get an lock on with 4 sat's and the reading says accuracy of 24 feet using one particular GPS model, but have another GPS model that says 4 sat's with an accuracy of 9 feet. The difference is not 1/3rd (approx) of error. Its worse. How much? I dont know. Its just a fact that its not linear. Due to the variables.

I know it sounds like mad scientist hoohaa but its truely not. The sad fact is I'm not the guy to offer an accurate scientific answer. But that doesnt change the fact that there indeed exists an error in the measurement. Believe what you want. We are all adults. Just my experience and of course 2 cents.

John
There are definitely inaccuracies to GPS speed readings. Anyone (marginally intelligent) who uses them for a decent period of time will figure that out.
Some models are obviously better than others though.
I have heard the military uses units that refresh at 200hz...a bit different from the standard 1hz RR.

[scoota]

A lot of people bag me for trusting my GPS readings , we have a lazer set up for saw runs at the club & my gps has always been with in 1-2 mph of the lazer setup , one of the members has a $2000 buck speed gun and again with in 1-2 mph difference .

So all this leaves me to belive my Garmin e-trax is a good tool for working out my speeds . The only time i have had bad readings was after a high speed crash , which has to be expected...

[m4a1usr]

I have heard the military uses units that refresh at 200hz...a bit different from the standard 1hz RR.

Thats pretty much what we use. 200mhz but interpolated at a lower frequency. I cant state what that value is. Sensitive information, not that I'm some sort of a spook. However most commercial units, and I mean 2011 GPS units with new hardware and software use 4 or 5 hz sampling but using an actual 1 hz frequency. Thats an accepted norm for today. And its pretty accurate for runners, bikers, boaters. The list is long. GPS wasnt invented to measure high speeds. It was designed for determining locations. They have just modified it for those later purposes to suit a need. True to nature, we take it where we feel its applicable. Not its intent. My own feelings are the long odds gamblers in Vegas would love to be on shore with a stop watch, measuring the time between bouys, whilst the gambler staked his bet on the GPS measured value.

John

[keithbradley]

A lot of people bag me for trusting my GPS readings , we have a lazer set up for saw runs at the club & my gps has always been with in 1-2 mph of the lazer setup , one of the members has a $2000 buck speed gun and again with in 1-2 mph difference .

So all this leaves me to belive my Garmin e-trax is a good tool for working out my speeds . The only time i have had bad readings was after a high speed crash , which has to be expected...

Yes in my experience the two things that throw them off are crashes, and losing signal. In the event there was no crash, and I get a wonky reading, I assume I lost signal for a brief moment. My nuvi is pretty consistant as long as there is a good signal and no crash though.

I would consider the etrex, but I dont really think its much smaller than my nuvi, is it?

[keithbradley]

Thats pretty much what we use. 200mhz but interpolated at a lower frequency. I cant state what that value is. Sensitive information, not that I'm some sort of a spook. However most commercial units, and I mean 2011 GPS units with new hardware and software use 4 or 5 hz sampling but using an actual 1 hz frequency. Thats an accepted norm for today. And its pretty accurate for runners, bikers, boaters. The list is long. GPS wasnt invented to measure high speeds. It was designed for determining locations. They have just modified it for those later purposes to suit a need. True to nature, we take it where we feel its applicable. Not its intent. My own feelings are the long odds gamblers in Vegas would love to be on shore with a stop watch, measuring the time between bouys, whilst the gambler staked his bet on the GPS measured value.

John

I dont think there's any question which method is more likely to glitch. There is an aspect of convenience to consider though. If I have to plant buoys an exact distance apart and time myself while driving every time I go to a pond, I just wont bother. Its a great method for events where a course can be set up for multiple members, but a bit impracticle for someone who just runs down to the nearest water to make some runs. Also, in instances with vehicles like my road cars, I usually use all the room I have to get up to speed. I dont normally have the room to run 120mph between two points for someone to time it.

[scoota]

Dont know the Nuvi , but the Forerunners are not up to the job, me & a few others had signal strenght problems , always lossing signal when it was in the boat...

[Mel279]

keith

get this one http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=220656183633&ssPageName=STRK:MEWAX:IT
This is the GPS used by the guy who run HPR 135 @ topspeed 205km/h

[scoota]

keith

get this one http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=220656183633&ssPageName=STRK:MEWAX:IT
This is the GPS used by the guy who run HPR 135 @ topspeed 205km/h

Cool , nice find :thumbup:

[drwayne]

Eagle tree V4 GPS 'reads' 5Hz
can log motor rpm with cheap AD module.
useful because rpm Vs speed is a great tool to evaluate prop efficiency.
The waterproof Forerunner201 are not... store became tired me replacing under warranty.. refunded money, told me to go away !
haha

Scoota,.. you going to SAW2011 ?

W

[m4a1usr]

I dont think there's any question which method is more likely to glitch. There is an aspect of convenience to consider though. If I have to plant buoys an exact distance apart and time myself while driving every time I go to a pond, I just wont bother. Its a great method for events where a course can be set up for multiple members, but a bit impracticle for someone who just runs down to the nearest water to make some runs. Also, in instances with vehicles like my road cars, I usually use all the room I have to get up to speed. I dont normally have the room to run 120mph between two points for someone to time it.

Yep. I'm with you Keith. I dont know what I would do without my GPS units. Its the only reliable method I know of measuring improvements to hardware adjustements, changing props, adding a new motor, etc. I hope no one has taken any of my comments as anti-GPS. I use mine all the time. Its just a tool.

I have a couple of the Geko's (101 and 201). Cheap and work fine. The 101 has no data interface capability. The 201 does.

I also have a forerunner 305 which I think is the better of the three I own. Whats funny is if I put all three in a hull and run it. I will get at least 2 different top speed values. Thats why you hear me being so skeptical. And the fact that these units cannot be calibrated. Since I work in the scientific community, where everything literaly gets calibrated to some standard, Well,..........you see why I am polarized. Good luck with your choice.

John