I've just bought a new satnav for SWMBO and it's set me thinking about the difference between actual speed and distance and that indicated by the car's instruments.
From comparing the speed on the clock against my satnav I know that my speedo is about 5mph on the optimistic side at an indicated 70. This is an error of ~7%. As the speedo drive also drives the milometer am I right in thinking that the error is also reflected in the milometer? If this is the case I can knock something like 14k off the car's mileage:)
I'd be interested in other members' opinions and experience; I've got to drive over to Wales this afternoon so I'll compare the mileage from the satnav against the trip meter.
|
The speedo is "set" fast by the maker deliberately. The odometer is usually bang on so no, you cant knock anything off the cars mileage. The sat nav is not fully accurate round bends, It works on point to point vectors and there will always be a small element of "cutting corners"
Edited by Altea Ego on 20/11/2009 at 10:15
|
The speedo is "set" fast by the maker deliberately.
I don't agree. By law the speedometer must not read slow, but any high reading is purely due to the inevitable design/manufacturing tolerances.
|
That's not the case, in my experience. On a few of the cars I've had there is a "secret" diagnostic mode that gives a read-out of the cars true speed, always in Km/h. This speed I've found to be spot on in agreement with the sat nav, where the speedo reads a few MPH over.
The car knows full well how fast it is really going, but deliberately adds 5-10% to the speedo reading to the driver to avoid litigation by drivers caught speeding against the manufacturer.
Maybe 10-15 years ago the speedo reading was inaccurate at source, but the sensors in a modern car are way too precise to give more than a small margin of error.
|
|
Thanks for the input so far. My car's 13 years old, so the speedo was probably inaccurate when fitted - how much does wear affect a speedo?
|
|
|
but the sensors in a modern car are way too precise to give more than a small margin of error.
Presumably they don't allow for under/over-inflation of the tyres and tread wear, both of which affect the tyre radius and hence odometer/speedo reading?
|
Presumably they don't allow for under/over-inflation of the tyres and tread wear both of which affect the tyre radius and hence odometer/speedo reading?
thats amount is insignificant. How much does tread depth vary over life of tyre? 6mm at most.
|
>> Presumably they don't allow for under/over-inflation of the tyres and tread wear thats amount is insignificant. How much does tread depth vary over life of tyre? 6mm at most.
Agreed, but that's still about 1%? The point I was trying to make is that there's no point having sensors which are 'way too precise' because they don't allow for all factors.
Edited by Focus {P} on 20/11/2009 at 11:07
|
Agreed but that's still about 1%?
Ok so what's +/- 1% really mean? at 70mph 70.07 or 69.03
Round it up, 1MPH.
|
>> Agreed but that's still about 1%? Ok so what's +/- 1% really mean? at 70mph 70.07 or 69.03 Round it up 1MPH.
Either you're still missing the point or I'm missing something. I know this is a small difference BUT the post I was replying to was talking about 'way too precise' sensors. That high precision is wasted when you have 1% errors elsewhere in the system.
|
I think that in this case 1% is consdered to be precise enough - thats +/- 1mph at 100 mph.
|
|
|
|
|
You can disagree as much as you like. It does not alter the fact that speedos in modern cars are deliberately set fast. Its nothing to do with design tolerances, thats the way they are set.
And I can prove it to you.
On my Recent VWs there is a speed overspeed warning. It matches the speedo needle perfectly. IE you set it at a speedo needle postion (say 80mph) and the digital setting says 80mph. Hit that speed and the two always match up.
Now Travel at an indicated 80mph and the sat nav says 74. Reset the average speed in the trip computer, and 15 seconds later it comes back and says average speed 74mph - matching the sat nav.
Now this proves that all the systems supplying the information, are providing totally accurate info, yet the systems that display spot data are overeading. This is the way its set.
And its done this way so they never fall below any posibility of design/manufacturing tolerences. In trith there are very few so you can safely assume your speed will always read fast.
|
I've checked the CC3's speedo against the roadside displays - often in roadworks - that display your speed.
My conclusion is the speedo is only a one mph or so optimistic at 50mph, if at all.
The margin - if there is one - would probably increase at higher speeds.
|
|
|
|
>> The speedo is "set" fast by the maker deliberately. I don't agree. By law the speedometer must not read slow but any high reading is purely due to the inevitable design/manufacturing tolerances.
>>
L'escargot given manufacturing tolerances if you, as a manufacturer are going to be penalised if it reads slow you are surely not going to try to set it up at the correct level, you will set it high to prevent there being any chance of it reading low, won't you?
|
SQ
L'escargot given manufacturing tolerances if you as a manufacturer are going to be penalised if it reads slow you are surely not going to try to set it up at the correct level you will set it high to prevent there being any chance of it reading low won't you?
Upper and lower design limits of any component/asembly can be specified as required. In the case of a speedometer the lower limit will be specified so that no sample indicates low, and the upper limit (which will result in some samples indicating high to one degree or another) will be determined by the level of technology available at whatever cost is considered acceptable to give a result within the confines of the relevant law. Speedometers are not deliberately set to read high, they are deliberately designed to not read low. tinyurl.com/ydzx534
Edited by Dynamic Dave on 20/11/2009 at 19:11
|
'Speedometers
are not deliberately set to read high they are deliberately designed to not read low.'
Precisely. The speedo on my car reads the same as my sat nav speed read out, and the same applies to other cars run by mates; the claim about deliberate design falls down on the fact that there are car models where the speedo reads the same as sat nav speed indicators.
|
|
The speedo on todays cars is legally required to be at 0 to +10% accuracy and this is checked at three set speeds-remember that many cars are offered with a selection of tyres which may not have exactly the same rolling radius and the speedo setting will be adjusted to cope with all.A large change in this would require the ECU to be reset.As someone else has said tyre wear also must be taken into account.The legal test will be carried out on run-in tyres with a recorded degree of wear.
|
|
|
You and your mates cars are the first cars to do this then.
|
|
|
|
are not deliberately set to read high they are deliberately designed to not read low.
Its the same thing YOU SET THEM ALL HIGH, some it is tru wil be higher than other BUT THEY WILL ALL BE SET HIGH
|
Its the same thing YOU SET THEM ALL HIGH some it is tru wil be higher than other BUT THEY WILL ALL BE SET HIGH
I'm sure that in this day and age of automated production there is no setting procedure. The reading of any particular sample will be -0% low to +10% high as required by law. The range of accuracies of various samples will be determined by design parameters not setting procedures.
|
sorry me old snail, the parameters are set in software (well firmwear microcode call it what you like) becuase its all electronic. There is no batch that varies from 0 to +10% They are all the same and all set to read +xx%
|
They are all the same and all set to read +xx%
Even your figure of +xx% has to have tolerances on it. They can't all read exactly the same.
|
to within 1% why not?
1% of what? 1% of xx% or 1% of true speed? And is the 1% tolerance symmetrical or asymmetric?
|
within 1% of whatever overspeed they specify.
|
|
|
|
|
|
This topic was discussed here only a few weeks ago. In summary, car speedos intentionally over-read by 5-10%, but distance recorders (odometers) are pretty well accurate.
For 30+ years I have checked my cars' odos (usually Peugeots) against the 100-metre markers on the M'ways (about 50km on the M6 usually). These are pretty reliable (a few are missing). The results indicated over-reads of 2-3% with mechanical-drive meters, but since electronic ones came in the error is only 1% or less (about ½% on my present car). I would guess other makes are similar?
In any case, what would be the point of 'adjusting' your car's mileage? Any buyer will expect to compare like with like, so what you intend is mild deception ...
Edited by Andrew-T on 20/11/2009 at 11:11
|
|
No intention at all to make any adjustment Andrew - it was purely a tongue in cheek remark. Given mileage of nearly 200k, 14k won't make much difference anyway.
|
|
|
Alterego is wrong in his ardent claims for "known" accuracy.
I know this because I worked as a designer for over 20 years on instrumentation used for dynamic brake testing.
There are many variables, effective tyre diameters vary greatly, not just wear, but tolerances vary with different manufacturers and different designs. Tyres are not round when they are on a vehicle. The "flattish bit" in contact with the road has a varying contour depending on load and many other factors. There are difficult to quantify effects known generally as "slip" and "creep" which can give errors of 2% or more. These vary with load, acceleration, road surfaces and tyre design and materials. I could go on and on.
Ever noticed that the rpm in your car vs speedo say at 70mph sometimes is a little different? Thats slip and creep. Note, "slip" is happening at a microscopic level here - not to be confused with lack of adhesion. Its all to do with when a curved flexible surface meets a flat surface and the complex interaction thereof.
But alterego is right in that manufacturers deliberately set speedos to read speed high. Legal requirements require a 0% to +10% band. Manufacturers set the reading somewhere mid way. Its programmable in most modern ECU's.
Mileometers tend to be set at 0% level but their accuracy will vary also.
Its a similar problem to "how long is a piece of string?".
|
>Alterego is wrong in his ardent claims for "known" accuracy.
My tests would indicate otherwise Mr Brum. As above I proved (using sat nav and the average speed output of the trip computer) the data available to the speedo on two of my last two cars (both VW based) is within 1mph accurate. And this has been over 40k miles with two sets of tyres.
>Ever noticed that the rpm in your car vs speedo say at 70mph sometimes is a little different
No. Not enough to notice anyway.
Edited by Altea Ego on 20/11/2009 at 12:25
|
|
|
At an indicated 70, my Charade shows 64 on my satnav however my wifes Sirion shows 68.
Even at 30 indicated mph my car is only doing a sat nav 26mph, however my wifes car shows 29.
Personally, id far rather drive mine as even at town speeds, I have to really want to break the limit ( to get a ticket for 35 mph ive got to be doing an indicated 40 mph ). Seems to me the most sensible way of psychologically limiting how fast people drive.
|
|
The last two LHDs that I've driven have been absolutely bang on target with the sat nav at 110 kph, compared to mine which over-reads about 6 mph at 70 mph.
|
|
|
Stu,
I have no beef about how accurate speedos are. It matters not to me if they are 2mph or 5mph out and I have no preference either way. I am merely asserting that it is easily possible to make a speedo indicate to +/- 1mph given current technology, but that makers choose to have them over read by what ever margin they like.
Most of the time I drive at sat nav indicated speed limit + 10%,
So on the motorway I cruise at 77mph sat nav, 83 or so Speedo.
Edited by Altea Ego on 20/11/2009 at 13:03
|
|
Assume for the sake of argument a tyre with a radius of 30 cm, which then wears 5 mm so that it has a radius of 29.5 cm. That's a 1.7% smaller circumference, so for a given speedometer reading the car will actually be going more slowly.
|
Assume for the sake of argument a tyre with a radius of 30 cm which then wears 5 mm so that it has a radius of 29.5 cm. That's a 1.7% smaller circumference so for a given speedometer reading the car will actually be going more slowly.
your argument is not valid becuase you have chsoen a deliberaterly small tyre radius. What is it a dinky car? Work out a better proportion of tread to tyre.
|
Of course they are all high, or nearly all. Autocar road tests used to check the speedometer accuracy at 10mph increments up to the maximum using a third-wheel calibrated speedo, a sort of bicycle wheel thing attached to the rear bumper. Some cars would exaggerate by as much as 10 or 12 mph at the top end, although some were pretty accurate especially at low and medium speeds (Rolls-Royce I seem to remember).
I remember with affection a hired Mk 1 Cortina 1500 estate whose speedometer needle oscillated between 90 and 115 across the better bits of Salisbury Plain. Averaged out at a ton I thought.
|
|
I can't remember why I did it - there must have been a pressing reason - but I fitted a second-hand rear axle from a 1500 estate to our Lada 1200 estate, theoretically a slightly higher final drive ratio. Didn't seem to affect the running in any palpable way. It gave me quiet satisfaction to think that perhaps a speedometer 70 really meant 70 with that one, when one could stand the racket.
|
I don't think that the law has in fact changed since the days when Autocar/Motore used their 5th wheel attachment, the requirment was for the speedo not to under read and to over read by no more than 10%. In the days of cable driven analogue devices it was no doubt difficult to produce accurate instruments at a reasonable cost but these days with digitsl instruments no problem. The speedo could be acuurate apart from the inevitable error due to tyre wear or the owner decideing to use a tyer different rolling radius.
The satnav of course should give an accurate reading of the speed in a straight line between the last two points where it has taken a position fix. Given that the car will seldom be travelling in a perfectly straight line the satnav reading would generaly be slightly low. How low would of course depend on the section of read being travelled on, verly little on a level motorway but a lot on a series of hairpins or a roundabout. Satnav speed is of course not the current speed but the speed between the last two position fixes.
There was a school of thought that said that the speedo would overread quite substantialy so that the owner could boast to his mates about how his XYZ Bloggsmobile would do 90 or 100 or whatever MPH. There was the concurrent school of thought that said that the odometer would over read so that the owner would think that he was getting better MPG and that serviceing would be done at slightly shorter intervals.
|
|
|
|
SQyour argument is not valid becuase you have chsoen a deliberaterly small tyre radius. What is it a dinky car? Work out a better proportion of tread to tyre.
?? I chose 30 cm because that is the radius of my Volvo's tyres. They measure approximately 2 feet across - look on your school ruler, 12" = 30 cm.
Volvo estates are not dinky cars!
Edited by Dynamic Dave on 21/11/2009 at 03:55
|
|
|
|
|
When I started driving I made a classic mistake of thinking a road was 40 but it was 30. There was speed cameras there and I zoomed past an indicated 35. I was worried sick I would get a ticket but never did. I then got a sat nav and realised at 35 by car (then a Fiesta) was actually doing 30-31mph.
My Corsa is the same, however I treat at as even then if I go through a speed camera at 30 I know I was actually doing 27 and there is way way on earth I could get a ticket.
|
I can't remember the figure for mph per thousand revs in top for my A3 (33 or 34mph) but I do know at 2,000 revs both my A3s showed 70mph and the sat nav read 66 or 68mph, so the tachometer figures seemed more accurate than the speedo.
|
The manufacturers do set the speedos to over read. Modern car instruments are designed with various parameters which could potentially be tweaked to under read or over read. In practice, the only variable which is adjusted is the tyre/ wheel combination which is set automatically by commisioning software at the time of manufacture of the vehicle.
If you change your tyre/ wheel combination you can have this variable reset by the main dealer. There are companies in Germany, for those so inclined, who will tweak your electronic speedo to make it read (theoretically) 100% accurate.
|
OK, here's the result from my (unscientific) test run - 54 miles according to the sat nav, 58 miles recorded on the tripmeter. Even allowing for corner cutting by the sat nav - and most of the miles were in a straight line on the M5 / M4 - this seems a fair amount of error by the tripmeter.
|
well, hi, sur
they would need to copy tractor technology
which uses a radar unit to record millimetric accurate speed over the ground
since when a tractor is working there will always be some wheelslip
simples
|
www.my-gti.com/wp-content/uploads/unece_3900_revis...f
Section 5. Specification. This spec as you will see does not permit an underread
Most modern cars have a correction table that can be corrected for the speedo reading my Audi A6 is 4.5 mph fast on new tyres, but I leave it that way and the audi speeding beep is set at an indicated 77. Regards Peter
|
|
What you have shown us is a UN-ECE regulation-you should be referring to a EEC directive which is what you need to meet to sell in the EU.
|
|
well hi sur they would need to copy tractor technology which uses a radar unit to record millimetric accurate speed over the ground since when a tractor is working there will always be some wheelslip simples
Tell us more about these tractor systems, sounds interesting.
|
|
So post it up jc2. Regards Peter
|
|
As requested,you need European directive 75/443/EEC as modified by 97/39/EC. UN-ECE regulations are usually similar but do not always apply in the EU and,when they do,frequently have different dates of introduction.
Edited by jc2 on 23/11/2009 at 10:33
|
I blush
I cannot tell you any more about tractor speed radar systems
Except I know that they are used
I will see if I can find owt out about
|
|
|
|
|
|
|
