Accuracy of speedo

esskay

Hi folks,
I recently bought an gps unit. The unit also displays your speed. Now, my father had it in his car and said that on the dual-carriageway his speedo and the gps unit were both reading 100km/h. Why then does my van read 110km/h on the speedo when the gps unit displays 100km/h?? I also noticed that when passing a sign that measures your speed it registers 110km/h (approx) on my van as 98km/h. Now, my father's car is a lot more expensive than my van, and since noticing this I have also noticed a lot of the lower priced cars cruise along at what the gps claims is 90km/h while the more expensive cars cruise at 100km/h.

Is there any requirement for vehicle manufacturers to make accurate speedo's?
Are lower priced cars clocked slower to make more expensive cars seem faster?
How accurate is the gps at calculating speed? The unit I have is a garmin c510.

Atomic Pineapple

the gps is accurate, your speedo is not, most speedos are out of calibration by a percentage, in fact i think most are set to read a little bit higher than what your actually doing.

your gps is more accurate anyway

colly10

esskay wrote:

Is there any requirement for vehicle manufacturers to make accurate speedo's?
Are lower priced cars clocked slower to make more expensive cars seem faster?

Speedo's only have to be accurate to a certain percentage. Manufacturers make the car look like it's going slightly faster than it actually is rather than slightly slower.

The reason that they choose slightly faster is because they would have problems if the speedo was saying your not breaking the speed limit even do you are (you'd be pretty pissed off getting a ticket if you believed you were within the limit.)

10% does seem excessive though

When I turned on the test mode on my old ford focus the digital speedo read 8mph slower than the clock.

Edit: I messed that up it was the wrong way round!

colly10

ronoc wrote:

When I turned on the test mode on my old ford focus the digital speedo read 8mph faster than the clock.

Thats possible as well, did you have the standard wheels on the car at the time?

Spitfire666

its also a covering their own arse thing. imagine the costs of being sued because the speedo said you were going one speed yet you were breaking the limit.

esskay

Thanks for the replies :-)

colly10

colly10 wrote:

Thats possible as well, did you have the standard wheels on the car at the time?

Not standard but same rolling radius as stock.

Garrr

I just relalised I had that the wrong way round. The clock is reading 8mph faster than the digital speedo.
Doh!

colly10

Spitfire666 wrote:

its also a covering their own arse thing. imagine the costs of being sued because the speedo said you were going one speed yet you were breaking the limit.

Ye thats the reason they overestimate rather than underestimate.

unkel

esskay wrote:

Is there any requirement for vehicle manufacturers to make accurate speedo's?

Yes there is. Speedos can read anything from 0% to 10% higher than the actual speed

esskay wrote:

Are lower priced cars clocked slower to make more expensive cars seem faster?

Nah, but more expensive cars could possibly be fitted with higher quality (read: more accurate) components such as speedos.

esskay wrote:

How accurate is the gps at calculating speed?

Depending on the software, the computing power (hardware) and the number of satellites in view, but generally quite accurate

There is a dedicated GPS forum here on boards.ie too for any GPS related threads

Spitfire666

i would class this as a speedo topic brought about by the use of a gps rather then a gps topic

unkel

Spitfire666 wrote:

i would class this as a speedo topic brought about by the use of a gps rather then a gps topic

I agree. Otherwise I would have moved it to said GPS forum like I do very regularly these days

Alun

colly10 wrote:

Thats possible as well, did you have the standard wheels on the car at the time?

The digital speedo (part of the diagnostics) takes it's signal from exactly the same sensor as the dial does, so the size of the wheels has nothing to do with it. It's therefore clearly a deliberate decision made by the manufacturers to make the dial over-read by a certain percentage.

Wossack

Alun wrote:

The digital speedo (part of the diagnostics) takes it's signal from exactly the same sensor as the dial does, so the size of the wheels has nothing to do with it. It's therefore clearly a deliberate decision made by the manufacturers to make the dial over-read by a certain percentage.

I dont think its a deliberate decision to over-read, I just think its more to do with the effective range of the gauges. Arent gauges more accurate when measuring in the centre of their range? As in when you start going out to the maximum and minimum extremes, you start to pick up these inaccuracies. The 0-20 is pretty damn inaccurate also on most cars, but I'd say the 40-60 range is the most accurate. Course Ive no figures though, and only vague recollections of my engineering days, so I'd take all this with a grain of salt

Alun

Wossack wrote:

I dont think its a deliberate decision to over-read, I just think its more to do with the effective range of the gauges. Arent gauges more accurate when measuring in the centre of their range?

Maybe, if it was a purely analogue device such as a simple 'voltmeter' type affair, or an old fashioned 'induction motor' style mechanical speedo, but the dials (all of them , not just the speedo) in modern instrument clusters are driven by stepper motors, which in turn are driven by the digital electronics on the instrument cluster. So, if there's any discrepancy, it's deliberate, and is probably there to cover their asses against them under-reading which is prohibited by some EU law, that has been quoted here before.

It BeeMee

esskay wrote:

Now, my father had it in his car and said that on the dual-carriageway his speedo and the gps unit were both reading 100km/h.

Based on other comments here, the car's speedo should be higher?

I did a rough experiment with my own car: Was on the motorway. Nice sunny day, no winds, not much traffic around. Set the cruise to 105km/h, and timed how long it took to travek 10km. It took 6 minutes and 5 seconds.

Given that the odometer is (usually) more accurate than the speedo, the time taken should be 6 minutes exactly.

Not the most scientific approach, but I can now be comfortable with the fact that when my speedo says 105, I'd travellin as close as dammit to the speed limit.


Interestingly enough, I did the same thing at 125km/h, but over a shorter distance (ran out of road ) and it would appear the speedo is constantly 5km/h (as opposed to 5%) off.

Really need to do the same for lower speeds......

gucci

For what its worth, going by the speed signs you see erected on roadsides it usually reads 3-4 kmph less when im doing 100kmph according to my clock. its a 97 VW polo im driving......weapon of a machine

AndrewMc

esskay wrote:

Is there any requirement for vehicle manufacturers to make accurate speedo's?

All the requirements are in the European Council Directive 75/443/EEC. Scroll down to "consolidated versions 1997-07-25", and check Annex 2. It defines the accuracy requirements, including that the speedometer can never underestimate your speed, but may overestimate within defined limits.

The website structure is a it confusing, but I'm pretty sure that's the latest and greatest version of the directive.

Linford

It BeeMee wrote:

Based on other comments here, the car's speedo should be higher?

I did a rough experiment with my own car: Was on the motorway. Nice sunny day, no winds, not much traffic around. Set the cruise to 105km/h, and timed how long it took to travek 10km. It took 6 minutes and 5 seconds.

Given that the odometer is (usually) more accurate than the speedo, the time taken should be 6 minutes exactly.

Not the most scientific approach, but I can now be comfortable with the fact that when my speedo says 105, I'd travellin as close as dammit to the speed limit.


Interestingly enough, I did the same thing at 125km/h, but over a shorter distance (ran out of road ) and it would appear the speedo is constantly 5km/h (as opposed to 5%) off.

Really need to do the same for lower speeds......

I've done similar tests with my car at different speeds and its seems my speedo is always overestimating my speed by about 10km/h - from 50km/h to 120km/h

I have a 325 coupe

unkel

Here's the definite answer:

2.3.7. the speed displayed must never be lower than the actual speed. Between speed V1, read on the speedometer and actual speed V2: there must be the following relationship with the test values specified in item 2.3.5 and between those values:

0 <= (V1 - V2) <= 0,1 · V2 + 4 km/h.

Linky

This means that an actual speed of 100km/h can mean a speed read on the speedometer of between 100km/h and 114km/h

And 180km/h indicated could be as little as 160km/h...