speedometer inaccuracies

[richardg8jvm]

As to GPS accuracy, 

the sampling is only done at 1sec intervals, which leads to errors as you have to travel at a constant speed long enough 

to allow for settling, and the error to average out, and make sure you are receiving signals from at least 6 sats.

More accuracy requires the use of surveying GPS receivers which sample at 20 times per second.

 

The last camera usage guidance in the public domain is very old 1981, but that quotes errors of +10/ -8.5% for a 10 degree offset.

If you put the 90 degree angle in drawings posted you get the same answers as I do if the right angle is put at the point of offset, not as shown

 

But you cant just use one right angle triangle to demonstrate this as each as the speed can be measured within one revolution of the car wheel and is taken several times.

 

If use use the calculations for bi-static radar and null each in turn, placing one where it should be ,at the side of the road, and the other offset from the road by 8 mtrs 

I would recommend trying the analysis unless you are familiar with using MatLab as its a process similar to fast Fourier analysis.

There is further documentation on errors produce on the IEEE.org site. unfortunately I no longer have access to that since retiring.

 

I asked as simple question, turning in to a trolling match, whos going to throw in quantum theory to really make things complicated ?

 

One thing very important to note.

Don't rely on any thing posted on web sites where the author remains anonymous, only rely on professional papers where the author/s are identified and their qualifications are listed, and 

they list references to other documents.

[Defenderben]

Must say this is all rather scientific.

I have re-read the OP's original post and I can say, I do not read he is in any way implying or advocating people or he are trying to break the speed limit.

As I see it he was asking how accurate is the Speedo...simples.

Most people know manufacturers cannot produce a Speedo that under reads the actual speed but it can give a higher reading, thus avoiding breaking the speed limit.

[Ryeman]

Obviously a lot of theoreticians have avoided speeding fines, so a search through court records should be interesting.

[Ootohere]

The GPS is good enough to find out at your leisure the difference between the Speedo & the Sat Nav.

I know for the 23 miles of 70 mph Speed Limit average speed cameras between Perth & Stirling if i can set the cruise control at 72, 73, or 74 for what ever i am driving.

Then after that it is a case of remembering, not checking the GPS every time.

 

Then there are villages with Signs that are telling you if you are doing 20, or 30, or 40 mph, or what ever speed, these are very accurate and easy for comparing with what the speedo is showing.

Edited 24 November, 2016 by Offski

[JCP]

Obviously a lot of theoreticians have avoided speeding fines, so a search through court records should be interesting.

Been in the law business for more than 40 years =not so although some make a living trying and others are sent away muttering about their sincerely held but unacknowledged grievance.

[Ryeman]

Been in the law business for more than 40 years =not so although some make a living trying and others are sent away muttering about their sincerely held but unacknowledged grievance.

I think the Daily Mail type newspapers would have been all over it as in the crude old days of radar scattering which must have been interesting if you were an insomniac.

[richardg8jvm]

The Daily Mail made a Freedom of information Request in June 2005 if I remember what I've seen list on the gov.uk site

bit no content of the info supplied

[Ryeman]

The Daily Mail?.

Radar or laser?.

[richardg8jvm]

From the date, 10 years ago I would guess X band or 24 GHz radar; both systems use doppler shift measurement, but the laser only needs an area of 

1 millionth of a millimetre (squared ) where the incident angle is 0 degrees, so its much faster to acquire returned pulses.

 

Maybe faster, but its less useful in wet conditions as rain drops are substantially larger, so you get refraction as the light passes through the rain drop.

[Ryeman]

The guidelines for location and weather have been nailed down by now surely.

There are relatively few locations approved from my observation of them in my frequent travels.

I set max+5 which gives me +1 (kph) and minimise the frustration of the tailgaters without getting booked.

Those overtaking me take the risk.

[Gobmax]

Hi my Yeti speed o  read's 6 km/hr slower than actual speed. Checked against 2 Mobile navigation devices.

1 Police time that I called for Traffic Police to contact their District Inspector. 135 in 100 km zone.

 

Surprise they did, and Inspector came and I  took him for a drive and we came back through Mobile Police.

 

Inspector apologized and said he would look into the matter, Seems Constables were adding km to reading

which was on top a unit that was mis calibrated. Career  enhancement was apparent Reason????.

 

So at 106 km actual 100 km designated speed and is the same across the board.

 

Received my License in 1968,  '68 to '72 received many fines, since X fingers none.

Edited 24 November, 2016 by Gobmax

[KenONeill]

It was an attempt to show the cosine effect which the OP, probably needed a better explanation...

 

 

Radar can't derive the object speed from the time it takes a single pulse to return, only distance - at least not without multiple receivers.

 

If you also take into account the frequency of the return pulse, then it's possible to get both distance and the radial component of the objects velocity.

 

As it's missing both the axial and tangential components, it will be lower than the actual velocity when not travelling directly towards or away from the radar.

 

If you take multiple measurements then you can get an average speed but it'll only be the actual object velocity is the angles are measured. Without the angles, you will still be missing the axial and tangential components.

 

Back to the OP statement, would you agree it's not possible to get a higher speed measurement than the actual object speed? At least assuming the radar/lidar is correctly aimed at a consistent part of the car.

That's heading into using doppler pulse radar.

 

Oh and I think you know enough mathematics to know that taking delta R over, say, 1s using a 20hz  sampling frequency is effectively integrating the area under the distance curve to get speed.

[g6zru]

These posts seem to be getting out of hand. It seems to me that it has turned into a contest about who can quote the most complex math. Does any of this really relate to day to day motoring or is it just an ego trip for a few?

Perhaps the moderators could find somewhere else for these posts so that common sense can resume?

 

Fred 

[ejstubbs]

That's heading into using doppler pulse radar.

 

I thought radar speed guns, like the old X band ones, did use doppler shift to measure the speed of the 'target'?  My understanding is that it is LIDAR that uses the return pulse time.  Or have I misunderstood what you were saying?

 

AFAIAA oppler systems are still subject to the equivalent of cosine error since the doppler shift will only be created by the component of the target's velocity along the line between the radar emitter and the target.  (I'm sure that's not 100% correct, but near enough.)

 

All of the above is "as I understand it", and I'd be more than happy to be corrected by those in the know.

[Brijo]

These posts seem to be getting out of hand. It seems to me that it has turned into a contest about who can quote the most complex math. ...

Fred 

I think the OP raised an interesting question, in essence, although he/she strayed into contentious territory with the final comment.  Yes, a serious driver would want to have a good idea of the difference between indicated and actual speed but splitting nanoseconds bears no relationship to real life. 

 

Hey ho... We can't be too critical, chaps - this is the Yeti forum, after all! 

[ejstubbs]

I keep a small aide-memoire printed out on an index card and tucked under the strap on the driver's sun visor.  Any time I need reminding - especially when I've just changed tyres - I can just flip the sun visor down and there they are.  The kph speeds are as indicated on the MFD.  These figures have been checked vs GPS.  The red percentages show the error between indicated and actual:

 

225/50R17

70 mph = 119 kph 5.6%

60 mph = 102 kph 5.6%

50 mph = 86 kph 6.8%

40 mph = 69 kph 7.2%

30 mph = 53 kph 9.8%

 

205/55R16

70 mph = 125 kph 10.9%

60 mph = 108 kph 11.8%

50 mph = 90 kph 11.8%

40 mph = 72 kph 11.8%

30 mph = 55 kph 13.9%

20 mph = 38 kph 18.0%

 

This suggests that my car's speedo is technically illegal when I have my winter tyres fitted.  Hmm...

 

(I have recently installed a NextBase dashcam which displays speed measured by GPS, so the card is now basically redundant.)

[brianja138]

This suggests that my car's speedo is technically illegal when I have my winter tyres fitted.  Hmm...

 

Hi ejstubbs, I had the same problem with 205/55 R16 winter tyres. When they were worn-out I switched to 215/60 R16's.[ They were recommended on the Goodyear web site for my car.] They not only eliminated the speedo problem, but also fill the wheel arch much better. Hope this helps.

[ejstubbs]

Can't use snow chains with 215/60R16s unfortunately.  Not on the front, anyway.

 

A pal of mine did point out recently that the usual advice re new tyres is to put them on the rear if only buying a pair.  So maybe putting chains on the rear only wouldn't be so bad?  I know that the owner's manual does mention it but it doesn't actually recommend it.  I still don't fancy the idea, at least in the absence of any authoritative information in favour.

[fatzy]

215/60 R16 is 0.9% bigger than 225/50 R17.

205/60 R16 is 0.9% smaller and listed for Yeti. Both are more suited than 205/55 R16.

[Ootohere]

Theory and converters are good, but then measure tyres with the matching numbers but from different makers and different types of tread and you will on occasion find that the tape measure shows they can be rather different physically.

Pirelli as an example might have a tyre that is 205/40 R 17 which next to some other brand of 215/40 R 17 is the same or even bigger. ie Total Diameter.

(this after noon i just fitted Maxxis Winter Tyres in 215's that measured the same as Pirelli 205's Both were 40 R 17's)

Checked the Speedo & Mileometer in my usual way 'accurately' to see the difference on the car, but no change between the different tyres when fitted.

http://kouki.co.uk/utilities/visual-tyre-size-calculator.

[TruckbusUK]

 

I keep a small aide-memoire printed out on an index card and tucked under the strap on the driver's sun visor.  Any time I need reminding - especially when I've just changed tyres - I can just flip the sun visor down and there they are.  The kph speeds are as indicated on the MFD.  These figures have been checked vs GPS.  The red percentages show the error between indicated and actual:

 

225/50R17

70 mph = 119 kph 5.6%

60 mph = 102 kph 5.6%

50 mph = 86 kph 6.8%

40 mph = 69 kph 7.2%

30 mph = 53 kph 9.8%

 

205/55R16

70 mph = 125 kph 10.9%

60 mph = 108 kph 11.8%

50 mph = 90 kph 11.8%

40 mph = 72 kph 11.8%

30 mph = 55 kph 13.9%

20 mph = 38 kph 18.0%

 

This suggests that my car's speedo is technically illegal when I have my winter tyres fitted.  Hmm...

 

(I have recently installed a NextBase dashcam which displays speed measured by GPS, so the card is now basically redundant.)

 

 

Don't forget any GPS speed reading is only accurate when the car is on a flat level road, and any gradient will make the reading inaccurate, as the "GPS" only ever sees the road your on as flat and level, so maybe not as accurate a method of measurement as you may think, best used when calculating a speed as an average over many driven miles.

[Ootohere]

When you drive a car daily and on flat roads you get the chance to see day after day, or weekly or even yearly.

It is pretty accurate really as you get to know just what your GPS is showing compared to the speedo over those many driven miles.

50 mph, 60 mph, 70 mph and you have it sussed for the Average Speed Cameras pretty much.

[fatzy]

Theory and converters are good, but then measure tyres with the matching numbers but from different makers and different types of tread and you will on occasion find that the tape measure shows they can be rather different physically.

Yes, you are right. Manufacturers do have production margins. In practice there are many parameters which matters. Of course, the tyre width is the starting point. That is dependent of how the manufacturer measure the effective tread width. Then there is the rim width, stiffnes of the tyre sides, air pressure. All of these will have influence of the actual rolling circumference.

The point is, 5-10% over-read on the speedo is mandatory for car manufacturers. As long as you are aware of it, you may compensate by equivalent "speeding" or fitting tyres with a little more circumference. Too small circumference will worsen the speedo reading related to the real speed.

[ejstubbs]

Don't forget any GPS speed reading is only accurate when the car is on a flat level road

 

Fair point.  The measurements I listed were deliberately made with the car being driven on cruise control on a flat, straight, level road (we do have a few like that in Scotland!)  Now that I have a continuous GPS speed readout in the car I can confirm that they still match very well in those circumstances.  I did notice that the MFD indicated a higher speed on a stretch of steep road this afternoon.  So it looks like the aide memoire may still be the best way to stay fully within the law.  I wonder how many people who drive to the GPS speed readout on their satnav are aware of this?

[Brijo]

Don't forget any GPS speed reading is only accurate when the car is on a flat level road, and any gradient will make the reading inaccurate, as the "GPS" only ever sees the road your on as flat and level, so maybe not as accurate a method of measurement as you may think, best used when calculating a speed as an average over many driven miles.

With all due respect, this isn't completely true.  The story's a bit more like this: 

 

A fairly modest modern GPS device will have the capacity to read signals from up to 12 satellites and these will be randomly scattered over the sky in accordance with their individual orbits.  Those at a high elevation (nearer the zenith, straight up) will indeed offer the most useful data regarding movement on the horizontal plane.  Those nearer the horizon will offer better data regarding the vertical plane than the high satellites do.  The GPS device simply (simply!!) comes up with information that is an average of all the available signals, however many and wherever the source.  (However, it certainly is acknowledged that vertical data isn't as usefully accurate as horizontal, that's why aircraft don't use pure GPS information for landing!)  So even here, it's a question of compromise between various inputs of information.  Splitting hairs.

 

Even with CC, my Yeti's indicated speed can and does vary up and down by 2-3mph (by speedometer, standalone satnav and, just to check things, my Garmin handheld).  What counts most is, of course, the official speed camera or calibrated speedometer.  As I said earlier - do you feel lucky? 

 

(Is this the Silly Season? )