Speedo vs. GPS

[balthazarr]

G'day all

Just wanted people's opinion on which is more accurate... the speedo or GPS sat-nav unit?

I'm currently renting a car while I wait for delivery of my new one, and notice that the speedo and my portable sat-nav unit (Navman S90i) differ significantly. I didn't really notice a significant difference on my old car.

eg. Speedo shows ~72km/h; GPS shows ~80km/h

The difference is even greater when speeds reach 100+ km/h.

I know speedos can be affected by things like tire pressure, etc., and the GPS accuracy depends on how many satellites are in view, but the GPS location seems spot on - eg. I will cross over a bridge just as the GPS unit shows I'm crossing a bridge etc.

Thoughts?

[Big Mike]

GPS triangulates your position from AT LEAST 4 satelites many many times per second. It is the most accurate method of measuring your speed, even more accurate than the "indication-only" speed detection devices our law enforcement deploys.

Your speedo is a mechanical device - it can't possibly be "accurate" and even the ADR states a tollerance of 10% on brand new cars. However, there's more than 10% difference between new and worn out tyres as the circumference reduces. Not to mention distortion of the cable (if mechanical), polar (think compass) deflections of the magnetic field driving the needle in your dash, etc. etc.

Tyres are the biggest difference - but you know how TV's and computer monitors are region-specific due to the magnetic alignment and the pull of the N pole etc. deflecting and distorting the picture? Yeah well that makes a difference to spinning magnets on your speedo needle too. Doesn't affect satelite-actuated global positioning systems though.

[balthazarr]

Originally Posted by Big Mike

GPS triangulates your position from AT LEAST 4 satelites many many times per second. It is the most accurate method of measuring your speed, even more accurate than the "indication-only" speed detection devices our law enforcement deploys.

True, but the GPS speed shown is also delayed ever so slightly, it's not instantaneous.

Your speedo is a mechanical device - it can't possibly be "accurate" and even the ADR states a tollerance of 10% on brand new cars. However, there's more than 10% difference between new and worn out tyres as the circumference reduces. Not to mention distortion of the cable (if mechanical), polar (think compass) deflections of the magnetic field driving the needle in your dash, etc. etc.

Yes, and this is what annoys me the most about Victoria's ridiculous 3km/h "leeway" on cameras.... if 10% tolerance is built into the speedo, then at 100km/h (on the speedo), you could actually be doing 110 (more if your tire pressure isn't correct, or you've fitted different tyres etc.), and bam... more revenue. :

Tyres are the biggest difference - but you know how TV's and computer monitors are region-specific due to the magnetic alignment and the pull of the N pole etc. deflecting and distorting the picture? Yeah well that makes a difference to spinning magnets on your speedo needle too. Doesn't affect satelite-actuated global positioning systems though.

Are speedo needles magnetic though? I could have sworn the needle in the speedo is plastic?

[baboon]

I'd believe your gps dude. Check your tyre sizes to see if the rent company hasnt put bigger tyres on there for some reason, check your manual perhaps for standard tyres sizes? I have a navman and a toyota echo and the digital speedo is spot on with the gps. Its also spot on with the speed indicators on the princess freeway from geelong to melb. You could drive down there one day and check them out. Good luck, this problem is not in your favour, would be nice if the gps was slower than your indicated speed.

[paule11]

I have noticed a lot of new cars we have had indicate about 5km more than they are doing says 100 but is actually doing 95 one was doing 93 and indicating 100 manufactures wont fix it as they say its within tolerance. xr6 is showing 100 and actually doing 98

[burnz]

i have often wondered myself.
when climbing a hill or decending a hill you are traveling vertical as well as horizontal,
so that a gps for a car doesn't have an altimeter and can't measure the true ground speed,
only in a horizontal distance.
PS my tom tom says 100k my car say's 101/102.

[Mr Brooksy]

WOW, Ive never heard of a Speedo, indicating such a huge difference at such a low speed. At 100kmh in my Falc Im doing 96kmh GPS! Thats huge!

Agree with Baboon, Check wheels!!!

[Lynch'd]

Nothing scientific here but from what i've been told (and it seems to make sense) the GPS will be more accurate only when you're speed is constant. E.g. You've got the cruise control set on 110 and you're cruising down the highway. When your speed is constantly varying (eg. suburban driving, 50-60km/h, accelerating and braking between lights) you're speedo will be more accurate. I believe this is both because of the slight delay and that the GPS can only locate you within around a 10m round area.

Interesting discussion though. Its intrigued me for a long time.

[Peuty]

Originally Posted by balthazarr

Yes, and this is what annoys me the most about Victoria's ridiculous 3km/h "leeway" on cameras.... if 10% tolerance is built into the speedo, then at 100km/h (on the speedo), you could actually be doing 110 (more if your tire pressure isn't correct, or you've fitted different tyres etc.), and bam... more revenue. :

You will find that speedos can only show a speed up to 10% faster than actual speed, not 10% slower.

[Spanrz]

Originally Posted by Peuty

You will find that speedos can only show a speed up to 10% faster than actual speed, not 10% slower.

Most probably due to the size of wheels that are spec'd for the car.
i.e. 16" or 17"

But due to a lot more tyre size's and wheel size demand in the last 10 years, it would be better for a vehicle manufacturer to err on the side of caution than to be recall every single vehicle for an "update" to correct kph readings (Showing under than the actual kph than reading spot on or over)

Can you imagine if it was spot on and how many people would get booked and challange it in court, only to loose, then they try to sue the manf, due to making it too easy to speed.

[Peuty]

Originally Posted by Peuty

You will find that speedos can only show a speed up to 10% faster than actual speed, not 10% slower.

I mean thats what the ADR states.

[Airmon]

I had a work Van a year or so back, and I always felt it was off because whenever I would be doing the speed limit traffic would always be overtaking. I used a GPS to get to a job on day and it was fine until 60km/h, then at 70km I was doing 66, 80 was really 73 and 90 was 80, etc.

[Donny]

My take on it is the speed reading is pretty accurate on the GPS units, but the actual refresh rate is pretty slow at every 2 or 3 seconds on my Tom Tom 910. You only need 3 points of refrence for triangulation, and as you move the refrence points track you constantly, so the speed will be right, just the refresh will be a little behind, so if your speed stays pretty constant, it will be spot on.

[Donny]

My take on it is the speed reading is pretty accurate on the GPS units, but the actual refresh rate is pretty slow at every 2 or 3 seconds on my Tom Tom 910. You only need 3 points of refrence for triangulation, and as you move the refrence points track you constantly, so the speed will be right, just the refresh will be a little behind, so if your speed stays pretty constant, it will be spot on.

[Donny]

My take on it is the speed reading is pretty accurate on the GPS units, but the actual refresh rate is pretty slow at every 2 or 3 seconds on my Tom Tom 910. You only need 3 points of refrence for triangulation, and as you move the refrence points track you constantly, so the speed will be right, just the refresh will be a little behind, so if your speed stays pretty constant, it will be spot on.

[balthazarr]

Originally Posted by balthazarr

...
eg. Speedo shows ~72km/h; GPS shows ~80km/h
...

Sorry guys, this should read: Speedo 80, GPS 72. :togo:

[Buzz Box]

I personally think its GREAT that my speedo shows about 2/3k's more than whats there because if I know its a camera area I set it exactly on 100 and I know im safe.

Better than it showing 100 and your actually doing 105 etc and get booked.

But yes I have to agree its PURE revenue nothing to do with safety!!

[SBR_1]

I had a 3month old kia rio while my car was being looked at and the speedo was at 110ks and my gps was only reading 97ks

[SBR_1]

double post

[data_mine]

Originally Posted by Peuty

I mean thats what the ADR states.

Not anymore, updated for 2006, and car built 2006 or newer, the ADR stipulates.

-0/+0 0-40km/h
-0/+4 KM/H above 40km/h

So it can't show slower than you're actually going, and no more than 4km/h over what you actually are doing. If you're doing 40 or less it's no leeway.

But to answer the OP, GPS better than speedo, as long as you've got a 3D fix (4+ sats), if only a 2D fix (3-sats) the speed doesn't take into account vertical movement (e.g. going up/down a hill).

[flappist]

Originally Posted by xagt_coupe

My take on it is the speed reading is pretty accurate on the GPS units, but the actual refresh rate is pretty slow at every 2 or 3 seconds on my Tom Tom 910. You only need 3 points of refrence for triangulation, and as you move the refrence points track you constantly, so the speed will be right, just the refresh will be a little behind, so if your speed stays pretty constant, it will be spot on.

You need a minimum of 4 satellites visible to aquire a reasonably reliable fix. 3 sats must use an altitude assumption which can be VERY wrong. There are roads between 0 (almost) and 4500 ft ASL in Australia, much higher in other countries.

This is why sometimes cheaper GPSs jump after they seem to have a lock.

If you have lots of sats even cheapo GPSs are quite accurate reading speed. I have checked one of mine at up to a bit over 400km/h and it was dead on.

[WPN XR6T]

Originally Posted by Lynch'd

Nothing scientific here but from what i've been told (and it seems to make sense) the GPS will be more accurate only when you're speed is constant. E.g. You've got the cruise control set on 110 and you're cruising down the highway. When your speed is constantly varying (eg. suburban driving, 50-60km/h, accelerating and braking between lights) you're speedo will be more accurate. I believe this is both because of the slight delay and that the GPS can only locate you within around a 10m round area.

Interesting discussion though. Its intrigued me for a long time.

The GPS Reciever take "bites" for want of a better word, taking a grid reference at regualar intervals, measuring the distance between these grids, and calculating time over distance traveled and calculate an average to give you speed. Depending on the quality of the receiver (accuratcy, refresh rate, signal, licencing etc) will determine how spot on the speed displayed is.

The one thing I tell all people when talking GPS is to remember, it is an aid to navigation, not a means, and should always be treated as such. If signal is lost, then 9 times out of 10 the receiver will continue to display incorrect information after the fact, and this information goes into it's calculations.

Just so your aware.

[GSH]

I have a TomTom and when im doing 100kph on my speedo the GPS sez im going 96-97kph by memory

[prydey]

just to clear up another thing. as your tyres wear it will make your speedo read even faster than your road speed so even that is not a valid excuse for an innaccurate speedo.

cops know this which is why they can set fairly fine tolerances for speeding. if they do you for doing 65 in a 60 chances are your speedo was showing closer to 70 esp if your tyres were worn.

[DBourne]

i was always sure that speedos showed a slightly lower reading than what you were actually doing. kinda saves you from blaming the speedo for any wrong doing.

[james22]

Well my speedo is spot on correct now. According to GPS and Radar when I'm doing 100, thats spot on what my speedo tells me. Of course it helps to be able to adjust your speedo calibration lol. Originally 100 indicated was about 97-98kph true speed

[aquahead2001]

Originally Posted by flappist

I have checked one of mine at up to a bit over 400km/h and it was dead on.

Wow! I assume that wasn't on a road around here! Aircraft?

I've been using hand held GPS systems for a while and have checked them against other units, kilometre posts (time and distance over 10km) and the Vicroads Speed Checks on Melbourne Road and Hume Hwy and they are always accurate to the kilometre. I never trust a speedo now that there's a 3km/h tollerance in my little Police State.

[Pinch]

I have found Hyundai and Subaru to over read the most.

Some ones I have tested:

Hyundai Lantra 91 at indicated 100
Subaru Forester 92 at indic 100
BA Falcon XR6 96 at indic 100
08 Aurion Presara 96 at indic 100
Mini Cooper 97 at indic 100
NC Fairlane 98 at indic 100
Opel Corsa 131 at indic 143 (before I get flamed it was on an autostrada)
Iveco Eurocargo MR Truck wins at 99 at an indicated 100

[vanman_75]

speedos are close with original tyres after that it is the owners responsibility to callibrate.
in saying that some cars are rediculous all my patrols have been 1-2 kms each way with original tyres
towing me boat the other day in tandem with mate in new cruiser ute he kept calling me saying im flying ,was doing 110...his speedo said 125
as he tows in fouth was concerned about the economy till i gave him my gps ,not happy with 10 + km variation

[Peuty]

Originally Posted by data_mine

Not anymore, updated for 2006, and car built 2006 or newer, the ADR stipulates.

-0/+0 0-40km/h
-0/+4 KM/H above 40km/h

So it can't show slower than you're actually going, and no more than 4km/h over what you actually are doing. If you're doing 40 or less it's no leeway.

But to answer the OP, GPS better than speedo, as long as you've got a 3D fix (4+ sats), if only a 2D fix (3-sats) the speed doesn't take into account vertical movement (e.g. going up/down a hill).

Heh, well there you go. I always thought 10% was a bit high.