Recently HJ's Telegraph column featured some poor soul who was prosecuted for doing 31mph in a 30.
I am aware the traditional 10% + 1 or 2mph allowance for speedo error has been recently overturned by the PC brigade, creating this situation to exist but was wondering which police forces now operate this policy. Surely not all...?
Perhaps anyone can share their experience.
|
A worrying development - I don't think that many speedos can be accurate to within 2-3 mph, (and in my case I'm using my rev. counter to judge speed whilst I source a secondhand speedo head!)
(Another Peter M! )
|
|
|
Peter
1mile/hr in 30 = 3.33%
Change in speedo accuracy between new and worn tyres typically 1.7%
Basic speedo accuracy, typically 5 -10 % ?
Camera accuracy. Doppler radar, combined with different angles so no better than a few percent?
Is this:
(a) Arguable in Court
(b) A money spinner
take your pick.
Regards
John S
|
A few years back when I used to read Autocar they measured speedo accuracy in road tests and 90% over read by around 6 to 7%.
I cant remember any that under read at all.
|
|
I have an Autocar somewhere with a review of a Yugo which produced an indicated top speed of 115mph measured at 87 real mph.
|
|
It probably felt like 115mph, though...
|
|
|
|
Allowable margin for error is -0/+10%
ie not allowed to show a lower indicated speed than true speed but can show up to 10% higher indicated than true.
I still maintain that 31 in a 30 is fiction. There must have been other circumstances. eg the actual conditions were such that 20 mph was pushing it in terms of appropriate behaviour.
|
Stuart
You stated, rather more succinctly than I, that the 'booked at 31 miles/hr story' is probably apocryphal, and I think you're right - I was trying to highlight the potential problems of trying to achieve a prosecution at that margin of error.
You are quite right to say that a new speedo can only have an error of between +0 and +10%. However, as Ian says the errors can be inconsistent. I also recall a test (Top Gear?) a few years back looking at a Sierra with a few miles on the clock - more representative of what Mr Average drives. I believe this showed that the speedo showed both +ve and -ve errors at different speeds; no doubt a function of wear. Probably less likely with modern electronic speedos, but a further cause of doubt that could be used by a defence counsel!
Like you I got out the anorak and stopwatch because the speedo and tacho of my car previous car did not give results consistent with the specification sheet. The result (using the measured mile indicators on the M4 near Hungerford) - speed was 66 at an indicated 70.
Regards
John
|
Hi John,
There is another issue here. I remember back in school days being taught about how to calculate experimental error and one thing was that you could only read an analogue gauge to a maximum accuracy of +/- half of the minor scale divisions. To make any reading at a more accurate level than that is an estimate and not an accurate reading.
Now looking at my speedo the minor divisions are 5mph, on SWMBO's they are actually at 10 mph intervals!
So in scientific theory the most accurate my speedo can be accurately read is +/-2.5mph. That is of course reading a dial under laboratory conditions where one can take a steady look at the dial, eliminate parallax errors and so on. Yet in the car one is expected to take an occasional glance to confirm speed you think you are at or about.
Be interested to hear opinions on this, plus possibility for this as a legal defence? DVD? pugugly?
regards
Stuart
|
|
|
|
|
Brunstrom tried this in 31 rubbish in North Wales but, I am told, decided to back off.
|
|
|
|
In reply to Peter M who started this thread, no it didn't.
HJ
|
|
|
Interesting thread indeed!
I have just thrown some numbers through the computer, and it comes out with some fascinating stuff re speedo error.
The figures are from a Subaru Impreza roadtest, and they measure indicated v actual roadspeeds.
Here we go:
Indicated: 60km/h 80 100 120
actual: 59 79 97 117
% error 1.67 1.25 3 2.5
So the % error isn't even linear! So that whole myth of x% over could get you into a lot of bother!
Hope this helps!
|
Ian you are right.
It comes back to a point I have made earlier, that the *only* way to be sure is to find a measured mile and do a series of timed runs and draw a calibration curve for the speedo.
Providing you do this when they tyres are new, and every so often do an update check, then any error due to wear is on the safe side. Thus I can go past cameras in a 50 @ an indicated 53 knowing that is exactly a true 50.
Appliance of science.
|
surely it is very easy for manufacturers to get a 100% acurate figure by using radar - afterall they have radars for parking!
|
Need a fixed point for that!
|
|
|
|
|
CM. I guess makers could calibrate speedos and give you a correction chart. What they can't do (for the price) is make them read accurately over a 20mph to 80mph range.
|
Pete,
refer above to the non-linear change in error!
Also, another (yes, anorak) calculation gives some differences twixt a new tyre and one which is down to legal limit - a big Tonka-Toy with mud diggers can be out by about 3%.
|
|
|
|
Presumably the same applies to Police speedometers. Does anyone know how they calibrate them (like they are supposed to)?
|
|
|
|
|
Speedometers have and always will be inaccurate usually reading higher rather than lower , since being flashed and pointed twice in two months i now simply drive to each speed limit that i am driving though, except when on the motorway when i will do around 80mph
|
Magistrate friend of mine threw a case out because he was advised that legally car manufacturers are only required to provide a speedo where the margin of error does not exceed 10%
He also threw out a case involving a skid because the police used standard coefficents of friction for tyres which ignored road conditions, tyre wear,type of tyre etc.As a highly qualified engineer he was mightily unimpressed with such an unrealistic approach by the police .
|
|
If only all magistrates were as well qualified...
|
|
|
|
A new speedo cannot have an error between +0 and +10.
It *must* over-read. so it *can* be between +0.00000000000000000000000000000000000000000000000000000000000000000000000000000001 and +10.
It cannot be perfectly accurate.
Sorry to be a pedant!
|
|
|
HJ, Assume you mean this wasn't in your column... sorry, my mistake if that was so. Could have been sure it was though... anyway I definately read this somewhere though and naturally assumed it to be the result of the meddling by the 'road safety' brigade, who threatened to bring the policy of error margins to court.
As has been pointed out, perhaps there were other circumstances involved in this case. But it would be interesting to hear from anyone who knows of a similar prosecution being brought for a 1mph transgression.
When I contacted my local force to enquire about their policy on this I was told they still allow 10%+1mph, but intend to lower this in the future.
|
|
|
>A new speedo cannot have an error between +0 and +10.
>
>It *must* over-read. so it *can* be between
>+0.0000000000000000000000000000000000000000000000000000000000000
>000000000000000001 and +10.
>
Look up the word "between". It means not actually +0 or +10, but something inside the two limiters.
Therefore +0.0000000000000000000000000000000000000000000000000000000000000
000000000000000001 is inaccurate, because it excludes the possibilty of being only +0.00000000000000000000000000000000000000000000000000000000000000
000000000000000001
inaccurate.
>Sorry to be a pedant!
Better to apologise for the inaccuracy, rather than the pedantry.
Mark.
p.s. or, from another standpoint..
>It cannot be perfectly accurate.
It is not likely, but it *can* be perfectly accurate.
|
HJ, I did indeed get the 31mph bit wrong (must have read that somewhere else). Just checked back and the following is a paste from your 20/03/01 column:
Club 35+
I thought an allowance of 10 per cent plus 2mph was made before any prosecution for speeding. I was flashed by a speed camera driving at 35mph in a 30mph limit and face a fixed penalty of £60, plus three penalty points.
A.L., Worsley
Join the biggest club in the country, which consists of citizens who this year will be fined for minor infringements which take no account of weather conditions, time of day or overall standard of driving. Unfortunately, after a misguided legal challenge last year, police officers are no longer permitted to use discretion over speeding offences. This legal ruling needs to be re-challenged on the grounds that the measuring equipment may not have been 100 per cent accurate at the time of the alleged offence. You have no choice but to pay the fine.
|
|
|
|
|
I have trouble believing 58 in a 50, never mind 1 MPH over. I drive through all cameras at 10 MPH over and have never been flashed.
|
Richard,
you have been lucky just seen a relative's NIP for 40 in a 30.
|
Blimey, 40 in a 30 is an error margin on the part of your relative's brain of 1/3. Let's hope your relative is more accurate with parking, carrying hot coffee, and using sharp knives. He/she is not a brain surgeon I hope. ;-)
Chris
|
Chris Chris Chris!
I am not defending the 40 in a 30, if you read my posts in this and in a previous alias you would know my views on this.
Richard was proposing the view that cameras are set at such a trigger limit that 10 mph over the limit will not get a flash. I know this is definitely not the case and I could not let this totally erroneous statement stand unchallenged on the site for the sake of the cleanliness of readers licences.
FiF
(the artist formerly known as Stuart B ;-)
|
|
|
FiF
Maybe my smiley wasn't smiley enough. Tricky blighter, e-mail humour.
Chris
|
|
|
|
|
"I have trouble believing 58 in a 50, never mind 1 MPH over. I drive through all cameras at 10 MPH over and have never been flashed."
That what my old man does and he recently got done at 11mph over. I reckon they're tightening up...
|
|
|
>Look up the word "between". It means not actually +0 or +10, but something inside the two limiters.
Hmmm. But IIRC it can be plus 10 so between is still wrong! Although not for the reasons I thought!
> Therefore +0.0000000000000000000000000000000000000000000000000000000000000
000000000000000001 is inaccurate, because it excludes the possibilty of being only +0.00000000000000000000000000000000000000000000000000000000000000
000000000000000001
inaccurate.
Curses. ;-(
>It cannot be perfectly accurate.
It is not likely, but it *can* be perfectly accurate.
> I don't think it legally can, Mark. According to a tame rozzer in UK law a speedo *has* to over read. It cannot over read by 0.
|
|
|
If anyone has been able to get their hands on a GPS system, it's very interesting to put it on the dash board, and compare the speed it thinks your doing to the speed the speedo says.
In both my old ZX and the Xantia, it said 70mph when the speedos were doing 79 & 76 respectively, which seems to be in keeping with the 10mph thingy.
Still think it would be a good idea that if they fitted dodgy speedos to boyracer cars, the roads would be safer as they think they're doing 90 when actually they're doing 50.
Lee.
|
|
You mean they can read? Someone in another thread suggested that IQ be included in the driving test - sounds good to me.
|
Rearward facing baseball baps should be an instant fail. In fact, baseball caps period.
And anyone turing up wearing a tracksuit top or football coat should be sent instantly to a young offenders institution immediately to save paperwork. Preferably one with a sprinkler system.
Lee.
|
|
|
|
Baseball caps period?
Period?
Full bloody stop, sir.
No damned colonials here, if you don't mind......we have standards to maintain, don't y'know?
|
|
|
