Speed accuracy

Andy_P · Posted: Tue Jan 01, 2008 3:26 pm Post subject:

not very.... Wink

Still less than 1/2% unless I'm even more hung-over than I thought! Laughing

M8TJT · Posted: Tue Jan 01, 2008 4:04 pm Post subject:

Jeez, I'm loosing sleep over this as well. What I want to know is how much slow my GPS will be if I am going up a 1 in 3 hill at a lattitude of 51.0000N 0.0000 W travelling on a heading of 257.5 degrees with a 4.478 degree drift (It's icy) Please use the WGS84 spheroid.
PS the hill is 213.5m long. (the road measurement hypotenuse)

Andy_P · Posted: Tue Jan 01, 2008 4:49 pm Post subject:

I'm off to the pub.... Guinness

M8TJT · Posted: Tue Jan 01, 2008 6:25 pm Post subject:

Yo, Dude. Good idea. I wish that I had thought of that before posting my last. But as I didn't, I have done the calculation and find that the difference is so small that I don't really give a d***.
Happy new year to everyone who visits this thread. Angel

. In fact, not only them, but everyone Very Happy

teamgers · Occasional Visitor Joined: Dec 20, 2007 Posts: 10

Phew, see what i've started................ Laughing

DennisN · Posted: Tue Jan 01, 2008 8:54 pm Post subject:

A far more accurate method was used when I were a lad. You counted the number of clicks as you went over the joins in the rails, then divided them by something and it told you how fast you were going. (You sort of said "Diddly dee, diddly dah, diddly dee, diddly dah and your pal counted them). Now that I've got demented, I can't remember what it was you divided them by, nor how long you had to count for. Anita could tell you.
_________________
Dennis

If it tastes good - it's fattening.

Two of them are obesiting!!

DennisN · Posted: Tue Jan 01, 2008 9:05 pm Post subject:

Anita · Posted: Tue Jan 01, 2008 9:38 pm Post subject:

Andy_P · Posted: Tue Jan 01, 2008 9:38 pm Post subject:

Hello1024 · Occasional Visitor Joined: Jan 01, 2008 Posts: 3

By my understanding, a GPS has theoretically 100% accuracy if you average it over enough distance in a straight line, because it measures time and distance from start and end points to calculate speed. Over enough distance the accuracy of the start and end points becomes irrelivant, and with atomic clocks the time is accurate to many more significant figures than is relevant.

The downsides are that there may well be rounding internally in the GPS, rounding for displaying the figure, and most units only calculate speed on a sliding 3 sec (or so) moving average, which is nowhere near as accurate as a minute or so in a straight line.

I believe GPS units will always under-estimate speed when going up and down hills or round bends, because they calculate the distance as the distance between points where they got a fix from the GPS satellites. When you draw straight lines between a finite number of points on a curve, the sum of the lengths of the lines is always less then the length of the curve.

Andy_P · Posted: Wed Jan 02, 2008 12:56 am Post subject: Re: back on topic...

Hello1024 · Occasional Visitor Joined: Jan 01, 2008 Posts: 3

Using doppler shift from multiple sources is great for an instant measure of velocity, but if there is no need for instant results (ie. you can average it over time) doing the old way of measuring time and distance has to be more accurate.

The report that doppler shift is used surprises me, because if that is the case, there is no significant reason to delay or average the displayed velocity, since at over a GHz it would be theoretically possible to measure the shift (by lots of possible methods) in at least a matter of miliseconds.

DennisN · Posted: Wed Jan 02, 2008 1:30 am Post subject:

	PocketGPS
	Web