Speedometer Accuracy to 1/2% Accuracy - Here's How

Low reading speedometers are costly in terms of police citations and create
safety problems as well.

Speedometers are inaccurate inherently and made worse as new tires are
installed and worn down due to the rolling radius variation.

A speedometer can be designed using the same principle as the optical
mouse. It reads the road and calculates V = D/T. Such a speedometer need
not be digital. A calibrated analog readout is feasible for 0.5% accuracy
at all speeds. It does not have to calibrate continuously if light or road
conditions are not within its capability. It simply uses memory and
reverts to the most recent self-calibration.

Some speedos are 5 or 6 mph off at road speeds. This should not be
tolerated in this day and age of high tech.

Such a speedo is quite elementary and involves absolutely no technology; I
have actually constructed a working model, breadboarded testbed. Although
it is the size of a shoebox and far from practical, it is entirely possible
to miniaturize it via a single dedicated I.C. I suggest this type
speedometer be standard equipment in the 2007 models and be made mandatory
by law.

 

Most speedometers, from the factory, read high. If it says you are doing
75, you are probably
doing less. Hence there is a bit of builtin safety. Cops in our state
normally wont ticket
you anyway until you exceed the limit by 5 mph.

Speedometers can often be calibrated by appropriate professionals. There
may be some
that cant be calibrated, but normally they can.

If you are worried about getting a ticket, run a calibration. Then dont
speed.

I sincerely doubt your optical mouse speedometer would maintain a high
accuracy over
varied types of road surfaces. There is no reason why it should.

 

I don't think this is true on modern vehicles. It ceratinly used to be
true in the past.

I have checked my speedo in couple of ways:
1. Reading the speed indicated by a GPS unit and comparing it to the
speedo.
2. Reading the speed on roadside radar speed signs.

It's quite possible that the latter (roadside speed indicators put in
place by local police) may read high, to encourage people to slow down,
but I rather doubt that the GPS unit would be inaccurate. However, both
tests show my speedo to be as accurate as I can read it (ie. within 1
mph).

 

I have checked my speedo in couple of ways:

According to my GPS, my speedo is bang on. Also, as HLS mentioned, police in
Ontario usually don't touch you unless you are going 15km/h over except on
the 401 where you have to be going at least 40km/h faster than everyone else
before getting pulled over.

Here in NS, 10+km/h over will usually get you at least a warning, but I have
gotten away with more.

Steve

 

Yea, let's pass another useless law that costs us all money with no real
benefit. I've checked several vehicles against a GPS and found them all to
be pretty damn close. Close enough at least that if the owner gets a
speeding ticket it isn't the speedo's fault.
Bob

 

Does Nomen Nescio by chance meen "Not a Clue"?????

Haven't seen ANYTHING of any substance from this poster in over a year
- just a lot of psuedo science babble nonsense.

 

We heard this from you about a month ago. Here we go again. Yawn.

Ted

 

Exactly. (Or they are correct.)

The OP is just polemical, based on an incorrect premise.

DAS
--
For direct contact replace nospam with schmetterling
---

[...]

 

Good points. I cant say I own anything really 'modern' right now. My 97
Dodge van tracks pretty well with the radar signs, but those are normally
in 45-50 mph zones. I wouldnt expect too much offset there. My son
checked the van with his GPS unit, and said I was 5 mph slow at an indicated
70 mph. Could be. Or not..

Since speedometers actually measure a function of drive train rotation, the
effects of effective tire diameter, speedometer readability, accuracy and
reproducibility would be some of the factors that might cause you to be
off a bit.

Still, it is a miniscule problem in the grander scheme of things.

 

FWIW, the effects of tire wear would be less than 4% over the entire
tread life (based on 12mm wear depth and 26" effective tread OD).
That's about 2 mph at 60 mph. You'd like to think that the car and tire
manufacturers split the difference in their tolerance stack. IOW - they
should figure the nominal speedo reading at 1/2 tread life (and some
reasonable assumed tire pressure), so, everything else being dead nuts
on, the worst-case contribution to speedo/odo error from tread (OD)
differences would be 1 mph at 60 mph in either direction (1 mph low
reading with new tread, 1 mph high reading at end of tread life.

With a cog/pulse generated speed signal and exact (gear or chain and
sprocket) known final drive ratio, as you pointed out, other than driver
reading inaccuracy (paralax error, etc.), the tires (effective OD due to
tire size and pressure and vehicle loading), there should be no other
errors (OK - *very* insignificant parts-per-million error due to
computer timebase inaccuracy).

Bill Putney
(To reply by e-mail, replace the last letter of the alphabet in my
address with the letter 'x')

 

Modern speedometers appear to be very accurate - on Fords, at least. For the
last few years, Ford has offered only Hybrid Electronic Clusters in their
vehilce line. These speedometers are electronically driven.

It all starts with the VSS..... Up unitl about 05, Ford used a variable
reluctance sensor to generate the VSS signal. This is placed close to an
indicator ring that induces current flow through the sensor and wiring as
the teeth on the indicator ring pass the sensor. The module that gathers VSS
information (and shares it across the network) knows how many teeth the ring
has and uses the frequency of the signal to determine the speed of rotation.
After 05, the sensor is still a two wire sensor but generates a digital
rather than analog signal. With these new sensors, we should see fewer
concerns related to incoherent VSS signals.

After this, tire rollout is the only concern. Most new tires will have a
tread depth of about 12 to 15 32nds of an inch..... meaning that a new tire
is nearly an inch taller than the same tire when it is worn out. This
translates into about 3" per revolution.... What a horrible state of
affairs.... what? Thats' 3" in a mile? A mile is 63,360"?.... oh....I guess
3" isn't going to matter much....

One last consideration is what a tall driver sees. Theough the beauty of
parallax, a short driver will see the needle in a different position than a
tall driver will see. This doesn't mean much since a *smart* driver will be
travelling , in most instances, at the rate of traffic flow. There are times
I wouldn't consider this since most traffic pile ups occur when someone
suddnely tries to change the rate of flow. We read about this nearly every
year as fog banks and sudden snow storms catch folks unaware.

Living in western Canada, we use our speedometers to be sure we are going
slow enough to avoid a ticket (police give us a 10% fudge factor meaning I
can drive all day at 110 kph in a 100 zonewith few worries). We use our
odometers to determine our next service interval... no great degree of
accuracy required.

If I'm driving to Edmonton ffrom Slave Lake, I know that it is about 2 and a
half hours in most weather... it doesn't matter how far something is, it
only matters how much time it will take to get there.

Living in the Great Arboreal Forest, our country side is crosshatched with
logging roads and oil patch roads rangiung from winter travel only to
fantastic. Even here, there is no great need for inch for inch speedo
accuracy.... "Its the second left after the old Camp 8 road" works for most
guys.

To sum it all up... If I were to kvetch about an inaccurate guage..... I'd
likely pick either the oil pressure of fuel level guage as my pet peeve....

 

My Chrysler vans have very accurate speedometers. Both of them
consistently read within 1 MPH of the local radar speed signs that the
police like to set up in my area.

A tire makes a lot more than one revolution per mile! That is, of
course, unless you are talking about a really large tire.

Matt

 

I've had absolutely no problems with the stone-age cable driven speedometers
in my Trans Ams. The speedometer in my '71 T/A is still dead-on after 35+
years and 150,000 miles.

--
_________________________________________________________________
Dennis Smith

-1971 Trans Am - 455 H.O. - M21 4speed - Cameo white/blue stripe-
-1973 Trans Am - 455 - TH400 auto - Buccaneer red-
-1984 Trans Am - 5.0 L - TH700R4 auto - Royal blue/silver aero-
_________________________________________________________________

 

My 90 Reatta senses the rotation of the axle from a toothed wheel and a
sensor. Then
the digitial pulses are converted from pulses per unit time to a function of
miles per hour
and output to the digital dash. It's as accurate as it needs to be.

Tire wear wont make much inaccuracy...maybe 1-2 mph at 70 mph, or 2-3 %
Under or over inflation might make a little difference, too.

But, honestly, I dont really care if it is off a skoshi.

I dont really see how Noman Nesciu's mouse model is going to work. If you're
going to
bounce a beam off the ground and try to correlate it to movement, you're
screwed. You
can bounce a beam off a wheel, or axle, etc which has a reflective digitizer
on it, and you
still have no better than what my 15 year old car has. You could use a GPS
system if
you had a lot of money you needed to pee away, or you could even chart the
earths magnetic
field and develop a global map which you could computerize and sense and
find out
about how fast you are going.

Does anybody really give a sheisse?

 

Simple doppler logic, easy to do.

Chips for the speed discrimination - about 1.98

 

For crying in a bucket, it is a non-issue.

DAS

For direct contact replace nospam with schmetterling
---

 

I feel cheated now... that's 1800 miles of warrantee or in my case, nine
days of business mileage.


Bill

 

Well, Im certainly no expert on this, but it wouldnt seem to be easy
to do. And not needed anyway.

What are you going to Dopple...sound, radio waves, light?

 

You must go stand in the corner if you can't interact with the
group positively...

 

I'm a bit familiar with speedometers and their accuracy when used in
commerce. When I last looked into this, a long time ago, speedometers
could be make accurate or inaccurate. And to set them to dead-on
accuracy was also fairly easy. All one had to do was to change the
speedometer gear so that it corresponded to exactly one mile with
whatever tires were on the car and so forth. As I recall, there was
really only one special speedometer shop that had a one mile track
nearby, straight, linear highway mile, that could give you exact
dead-on accuracy if requested and be tested for it. Extremely accurate.
I was surprised at the accuracy that could be achieved with just the
right speedometer gear installed.

But keep working on your notion. I see other possibilties for your idea
other than speedometers. I just wanted to post because I had actual
experience with testing old, analog speedometers and was surprised that
they could be made extremely accurate.