Gps Speed

[Parthiban]

I know we often talk of the speed shown on our GPS devices and use that as being more accurate than the speedo, but I was wondering how accurate they actually are?

Reason I ask was that the other day I was driving down a steep hill at an indicated 50mph, but my GPS was saying 35mph, obviously because of the steepness of the hill, my lateral speed was a lot lower than my ground speed.

Therefore, as very few roads are perfectly flat, wouldn't GPS devices always be underestimating our speed to varying degrees?

[DJ Wozza]

That's a fair point matey, but on a flat level road I'd suggest they were prety accurate

[Geoffers]

It depends if they take altitude into the equation. Most don't.

You are 100% right - GPS speeds are only accurate on a level with respect of the previous calculation.

Although the distance covered over time equals your speed, the elevation differences have not been calculated.

[Parthiban]

Although the distance covered over time equals your speed, the elevation differences have not been calculated.

That is actually a fair point, and in those terms it will be spot on. Although I meant less in terms of "real" speed, and more in terms of which readout to rely on when going past speed cameras! :winky:

[Rikos]

Well if you go by your speedo then you know you will always be ok.

[Vyker]

My view on this is...

It is accurate regardless of altitude, becuase you have to remeber that the sattillites up there are beacons, and it is the car that picks up the signal,

so if you're going down a hill from point A to point B, and that distance is 100m, on the road, but the straight line latitude distance is 90m, the car picking up the signal still traveled with you doing that same 100m, so it will work it out doing the actual distance.

I hope that makes sense.. correct me if im wrong

Vyker

[Rob]

if u recieve 4 or more sats then the altitude should be calculated but as the altitude ( on a hill) is changing the displayed speed wont always be accurate or "real time"

[The Editor]

3 Sats will give you triangulation and the fourth (or more) sat fix will give you a 3 dimensional fix

GPS is VERY accurate - it's the processing in Sat Nav systems that doesn't make full use of the data or refresh rates.

The Ed

[Geoffers]

My view on this is...

It is accurate regardless of altitude, becuase you have to remeber that the sattillites up there are beacons, and it is the car that picks up the signal,

so if you're going down a hill from point A to point B, and that distance is 100m, on the road, but the straight line latitude distance is 90m, the car picking up the signal still travelled with you doing that same 100m, so it will work it out doing the actual distance.

I hope that makes sense.. correct me if im wrong

Vyker

I can see where you are coming from but they don't calculate it that way I'm afraid.

Your latitude and longitude are calculated based on zero inclination by the difference between the last and present Almanac/Ephemeris data figures received. As an example (and an extreme one here I know), if you went vertically at 100mph, your gps speed would still stay at 0.

More satellites do indeed allow you to calculate the altitude, but a lot of car GPS systems do not make use of this data.

It is very accurate now, but positioning will get almost pinpoint with a lot less errors soon, as the new chipsets pick up the US Navstar, the Russian Glonass and the Euro Galileo.

[Sagitar]

I know we often talk of the speed shown on our GPS devices and use that as being more accurate than the speedo, but I was wondering how accurate they actually are?

Reason I ask was that the other day I was driving down a steep hill at an indicated 50mph, but my GPS was saying 35mph, obviously because of the steepness of the hill, my lateral speed was a lot lower than my ground speed.

Therefore, as very few roads are perfectly flat, wouldn't GPS devices always be underestimating our speed to varying degrees?

Sorry, but I think you have to look for a different reason for an error of that magnitude.

To get the indicated speed down to 35mph you would have to be on a slope that was greater than one in one. I.e. the hill would have to be so steep that the horizontal component was only 35 yards for every 50 yards actually travelled. If you can remember your pythagoras, that implies a vertical component of nearly 36 yards and an included angle for the road of about 45 degrees.

A ten percent gradient changes the horizontal component by about a half of one percent, so your 50mph should show as 49.75mph on such a gradient. i.e. within the reading tolerance on normal instruments.

A one in three gradient (e.g. Hardnott Pass . . . ) reduces the horizontal component by about 5% so the GPS would read 47.50mph. Not that I can imagine anyone doing 50mph down Hardnott.

[Geoffers]

While I agree with your calculations, I don't dismiss it so quickly. Its all a question of timing.

It does throw the GPS out dramatically sometimes, even with multiple sats. Especially as you go over the brow of a hill, etc. Sudden changes really do confuse some systems for a short period.

[Parthiban]

Yep it was a sudden steep drop from a flat road so there was a change, but while it was definitely 35 on the GPS, there's a good chance it was about 45ish on the speedo which i rounded to 50.

Either way my question was more to do with the drastic difference between the two speed readings - thanks for reminding me of pythagoras' theorem though, haven't used that stuff in a long while!

[Wallace]

I have an old TomTom 3 on a Dell Axim in my van and its always 10% slower than the speedo. Guess its all to do with how fast your processor is and how many rubbish programmes are running, same as any PC.

[ColinBarber]

I have an old TomTom 3 on a Dell Axim in my van and its always 10% slower than the speedo. Guess its all to do with how fast your processor is and how many rubbish programmes are running, same as any PC.

It's 10% slower because your speedo is set to read 10% faster than the actual speed. All speedos in UK cars read above the true speed.

[ColinBarber]

Yep it was a sudden steep drop from a flat road so there was a change, but while it was definitely 35 on the GPS, there's a good chance it was about 45ish on the speedo which i rounded to 50.

Either way my question was more to do with the drastic difference between the two speed readings - thanks for reminding me of pythagoras' theorem though, haven't used that stuff in a long while!

and if the speedo was reading 45ish then you were really only doing around 40mph so the difference between the true speed and the GPS could be caused by the altitude.