Re: [piclist] Re: Speedometer Calibrator GPS cal


Jun 25, 2004

 


----------------------------

#6274 Jun 25, 2004

You should do this at the speeds you commonly drive.

I was rather surprised to find that my '93 Caravan gave accurate

speeds (GPS) for values greater than 70km/hr, but read high

by 5km or so for values lower than this. Bizarre, given that

the dial seems linear, and the pointer is driven by a stepper motor.



I do remember coming across a web page which mentioned that

the electronic speedometer can be calibrated by the dealer

'service and maintenance module', presumably by changing some

scale factor in EEPROM (they use HC11's for body controller).



I also remember reading (decades ago) that car makers deliberately

made mechanical speedos read high at high speeds to fool people in

driving lower. This seems to be the reverse of that. Makes you drive

slower in town.



Robert



Don Hackler wrote: >

> I've found the easiest speedometer check is using a GPS. The Magellan

> units (and I assume the others) have a display mode that will show

> speed. Toss the GPS on the dashboard, set the cruise, and see how far

> off your speedo is. In my car, when the actual speed (by the GPS) is

> 65, my speedometer shows 70. This works best on reasonably straight and

> level roads..

>

> - Don.

>

> Dave Mucha wrote:

>

> >--- In piclist@yahoogroups.com, "rtstofer" rstofer@p...> wrote:

> >

> >

> >>I haven't seen the device you are looking for but I can tell you

> >>that it is common for the mechanical cable to rotate at either 800

> >>or 1000 revolutions per mile. Which it is depends on the

> >>manufacturer.

> >>

> >>

> >

> >

> >Unfortunatly, it also depends on tire inflation and wear as well as

> >the car being fitted with proper tires.

> >

> >The last time I worked on a car was an '83 cutlass that kept breaking

> >cruse control cables.

> >

> >Almost everywhere you go, there are calibrated mile markers.

> >Typically a 2 mile stretch where you get your cruse control up to

> >speed, pass the first marker, then a mile later, a second marker.

> >And, often those have reflectors that you can use for feedback.

> >

> >Dave

> >

> >>I don't know what car you have and even if I did I wouldn't know

> >>anything about it. The last car I messed with was a '73 Porsche

> >>

> >>

> >914

> >

> >

> >>that I used in sports car rallies. I had just graduated with a

> >>

> >>

> >BSEE

> >

> >

> >>and I built a rally computer with dimmable incadescent displays and

> >>all that stuff. In rallies there is a calibration leg where you

> >>have the opportunity to match your odometer to the rally master's.

> >>Hence, I know a little about the old ways of doing things. This

> >>thing had a barn full of 74xx logic and could easily drain the

> >>battery in an hour or so.

> >>

> >>But it was fun!

> >>

> >>Anyway, if you have a strictly digital system it should be easy to

> >>grab the pulses coming from the transmission, clean them up and

> >>stuff them into a PIC. Then you have to do some type of

> >>

> >>

> >computation

> >

> >

> >>and output a similar pulse to the odometer.

> >>

> >>If you don't get more help here, look into taxi meters. They must

> >>do the same kind of thing.

> >>

> >>--- In piclist@yahoogroups.com, "Roy Wheaton" roy_wheaton@y...>

> >>wrote:

> >>

> >>

> >>>I've noticed several products for autos and trucks that provide

> >>>

> >>>

> >>speed

> >>

> >>

> >>>corrections for electronic speedometers. Is this something that

> >>>

> >>>

> >>can

> >>

> >>

> >>>be done with a PIC? How does this type of device work? I'm

> >>>

> >>>

> >already

> >

> >

> >>>using a PIC in my Toyota for various other functions, so it would

> >>>

> >>>

> >>be

> >>

> >>

> >>>a great fit.

> >>>

> >>>Thanks in advance for your help.

> >>>Roy



----------------------------

#6277 Jun 26, 2004

There is a mathematical fallacy in trying to get speed from a GPS

and it has to do with the uncertainty of the distance covered during

a sample period. Remember that EVERY point given by the GPS has an

error - less now that selective availability has been removed but

still real.



Take 60 MPH for example - that is 88 feet per second. Now, if the

GPS measured 95 feet (only a 7 foot difference) that would be 65

MPH. If the measurement was 81 feet the speed would be 55 MPH. And

again, EVERY point was an error. The more distance you cover during

the sample period the more accurate the reading will be because the

sample error is a smaller percentage of the distance.



I suppose you could use the CEP (circular error probable) and kind

of draw the circles and geometrically figure the range of error but

it is sufficient to know that it exists.



--- In piclist@yahoogroups.com, Robert Rolf robert.rolf@u...> wrote:

> You should do this at the speeds you commonly drive.

> I was rather surprised to find that my '93 Caravan gave accurate

> speeds (GPS) for values greater than 70km/hr, but read high

> by 5km or so for values lower than this. Bizarre, given that

> the dial seems linear, and the pointer is driven by a stepper

motor.

>

> I do remember coming across a web page which mentioned that

> the electronic speedometer can be calibrated by the dealer

> 'service and maintenance module', presumably by changing some

> scale factor in EEPROM (they use HC11's for body controller).

>

> I also remember reading (decades ago) that car makers deliberately

> made mechanical speedos read high at high speeds to fool people in

> driving lower. This seems to be the reverse of that. Makes you

drive

> slower in town.

>

> Robert

>

> Don Hackler wrote:

> >

> > I've found the easiest speedometer check is using a GPS. The

Magellan

> > units (and I assume the others) have a display mode that will

show

> > speed. Toss the GPS on the dashboard, set the cruise, and see

how far

> > off your speedo is. In my car, when the actual speed (by the

GPS) is

> > 65, my speedometer shows 70. This works best on reasonably

straight and

> > level roads..

> >

> > - Don.

> >

> > Dave Mucha wrote:

> >

> > >--- In piclist@yahoogroups.com, "rtstofer" rstofer@p...> wrote:

> > >

> > >

> > >>I haven't seen the device you are looking for but I can tell

you

> > >>that it is common for the mechanical cable to rotate at either

800

> > >>or 1000 revolutions per mile. Which it is depends on the

> > >>manufacturer.

> > >>

> > >>

> > >

> > >

> > >Unfortunatly, it also depends on tire inflation and wear as

well as

> > >the car being fitted with proper tires.

> > >

> > >The last time I worked on a car was an '83 cutlass that kept

breaking

> > >cruse control cables.

> > >

> > >Almost everywhere you go, there are calibrated mile markers.

> > >Typically a 2 mile stretch where you get your cruse control up

to

> > >speed, pass the first marker, then a mile later, a second

marker.

> > >And, often those have reflectors that you can use for feedback.

> > >

> > >Dave

> > >

> > >>I don't know what car you have and even if I did I wouldn't

know

> > >>anything about it. The last car I messed with was a '73

Porsche

> > >>

> > >>

> > >914

> > >

> > >

> > >>that I used in sports car rallies. I had just graduated with a

> > >>

> > >>

> > >BSEE

> > >

> > >

> > >>and I built a rally computer with dimmable incadescent

displays and

> > >>all that stuff. In rallies there is a calibration leg where

you

> > >>have the opportunity to match your odometer to the rally

master's.

> > >>Hence, I know a little about the old ways of doing things.

This

> > >>thing had a barn full of 74xx logic and could easily drain the

> > >>battery in an hour or so.

> > >>

> > >>But it was fun!

> > >>

> > >>Anyway, if you have a strictly digital system it should be

easy to

> > >>grab the pulses coming from the transmission, clean them up and

> > >>stuff them into a PIC. Then you have to do some type of

> > >>

> > >>

> > >computation

> > >

> > >

> > >>and output a similar pulse to the odometer.

> > >>

> > >>If you don't get more help here, look into taxi meters. They

must

> > >>do the same kind of thing.

> > >>

> > >>--- In piclist@yahoogroups.com, "Roy Wheaton"

roy_wheaton@y...>

> > >>wrote:

> > >>

> > >>

> > >>>I've noticed several products for autos and trucks that

provide

> > >>>

> > >>>

> > >>speed

> > >>

> > >>

> > >>>corrections for electronic speedometers. Is this something

that

> > >>>

> > >>>

> > >>can

> > >>

> > >>

> > >>>be done with a PIC? How does this type of device work? I'm

> > >>>

> > >>>

> > >already

> > >

> > >

> > >>>using a PIC in my Toyota for various other functions, so it

would

> > >>>

> > >>>

> > >>be

> > >>

> > >>

> > >>>a great fit.

> > >>>

> > >>>Thanks in advance for your help.

> > >>>Roy







----------------------------

#6278 Jun 26, 2004

Hi,



With current GPS without SA, the velocity accuracy that can be achieved on

flat ground is quite high, much higher than by other means. It is true that

GPS has errors, but I find it WAY more accurate than any vehicle tachometer

I have tried. In general, at least in the vehicles I've driven, the tacho

shows a figure some 10% higher than the true speed. Doing distance tests

with a GPS and the car's tacho, the GPS achieved a much more accurate result

in terms of velocity.



Another factor is that GPS doesn't rely just on position changes over time

to measure velocity, the calculation is actually done using the

pseudo-ranges and pseudp-range-rates (PRs and PRRs), which include the

apparent doppler shift in the signals.



I know of a Danish-made device which reports distance travelled using SMS

messages, which all it has is a tachometer sensor which measures turns. To

calibrate it, you drive around for a while, and send it the distance

readings on the car's tacho, from there onwards it can tell you distance

travelled.



Best regards,



Mike



----- Original Message -----

From: "rtstofer" rstofer@...>

To: piclist@yahoogroups.com>

Sent: Saturday, June 26, 2004 4:34 PM

Subject: [piclist] Re: Speedometer Calibrator GPS cal





>

> There is a mathematical fallacy in trying to get speed from a GPS

> and it has to do with the uncertainty of the distance covered during

> a sample period. Remember that EVERY point given by the GPS has an

> error - less now that selective availability has been removed but

> still real.

>

> Take 60 MPH for example - that is 88 feet per second. Now, if the

> GPS measured 95 feet (only a 7 foot difference) that would be 65

> MPH. If the measurement was 81 feet the speed would be 55 MPH. And

> again, EVERY point was an error. The more distance you cover during

> the sample period the more accurate the reading will be because the

> sample error is a smaller percentage of the distance.

>

> I suppose you could use the CEP (circular error probable) and kind

> of draw the circles and geometrically figure the range of error but

> it is sufficient to know that it exists.

>

>

> --- In piclist@yahoogroups.com, Robert Rolf robert.rolf@u...> wrote:

> > You should do this at the speeds you commonly drive.

> > I was rather surprised to find that my '93 Caravan gave accurate

> > speeds (GPS) for values greater than 70km/hr, but read high

> > by 5km or so for values lower than this. Bizarre, given that

> > the dial seems linear, and the pointer is driven by a stepper

> motor.

> >

> > I do remember coming across a web page which mentioned that

> > the electronic speedometer can be calibrated by the dealer

> > 'service and maintenance module', presumably by changing some

> > scale factor in EEPROM (they use HC11's for body controller).

> >

> > I also remember reading (decades ago) that car makers deliberately

> > made mechanical speedos read high at high speeds to fool people in

> > driving lower. This seems to be the reverse of that. Makes you

> drive

> > slower in town.

> >

> > Robert

> >

> > Don Hackler wrote:

> > >

> > > I've found the easiest speedometer check is using a GPS. The

> Magellan

> > > units (and I assume the others) have a display mode that will

> show

> > > speed. Toss the GPS on the dashboard, set the cruise, and see

> how far

> > > off your speedo is. In my car, when the actual speed (by the

> GPS) is

> > > 65, my speedometer shows 70. This works best on reasonably

> straight and

> > > level roads..

> > >

> > > - Don.

> > >

> > > Dave Mucha wrote:

> > >

> > > >--- In piclist@yahoogroups.com, "rtstofer" rstofer@p...> wrote:

> > > >

> > > >

> > > >>I haven't seen the device you are looking for but I can tell

> you

> > > >>that it is common for the mechanical cable to rotate at either

> 800

> > > >>or 1000 revolutions per mile. Which it is depends on the

> > > >>manufacturer.

> > > >>

> > > >>

> > > >

> > > >

> > > >Unfortunatly, it also depends on tire inflation and wear as

> well as

> > > >the car being fitted with proper tires.

> > > >

> > > >The last time I worked on a car was an '83 cutlass that kept

> breaking

> > > >cruse control cables.

> > > >

> > > >Almost everywhere you go, there are calibrated mile markers.

> > > >Typically a 2 mile stretch where you get your cruse control up

> to

> > > >speed, pass the first marker, then a mile later, a second

> marker.

> > > >And, often those have reflectors that you can use for feedback.

> > > >

> > > >Dave

> > > >

> > > >>I don't know what car you have and even if I did I wouldn't

> know

> > > >>anything about it. The last car I messed with was a '73

> Porsche

> > > >>

> > > >>

> > > >914

> > > >

> > > >

> > > >>that I used in sports car rallies. I had just graduated with a

> > > >>

> > > >>

> > > >BSEE

> > > >

> > > >

> > > >>and I built a rally computer with dimmable incadescent

> displays and

> > > >>all that stuff. In rallies there is a calibration leg where

> you

> > > >>have the opportunity to match your odometer to the rally

> master's.

> > > >>Hence, I know a little about the old ways of doing things.

> This

> > > >>thing had a barn full of 74xx logic and could easily drain the

> > > >>battery in an hour or so.

> > > >>

> > > >>But it was fun!

> > > >>

> > > >>Anyway, if you have a strictly digital system it should be

> easy to

> > > >>grab the pulses coming from the transmission, clean them up and

> > > >>stuff them into a PIC. Then you have to do some type of

> > > >>

> > > >>

> > > >computation

> > > >

> > > >

> > > >>and output a similar pulse to the odometer.

> > > >>

> > > >>If you don't get more help here, look into taxi meters. They

> must

> > > >>do the same kind of thing.

> > > >>

> > > >>--- In piclist@yahoogroups.com, "Roy Wheaton"

> roy_wheaton@y...>

> > > >>wrote:

> > > >>

> > > >>

> > > >>>I've noticed several products for autos and trucks that

> provide

> > > >>>

> > > >>>

> > > >>speed

> > > >>

> > > >>

> > > >>>corrections for electronic speedometers. Is this something

> that

> > > >>>

> > > >>>

> > > >>can

> > > >>

> > > >>

> > > >>>be done with a PIC? How does this type of device work? I'm

> > > >>>

> > > >>>

> > > >already

> > > >

> > > >

> > > >>>using a PIC in my Toyota for various other functions, so it

> would

> > > >>>

> > > >>>

> > > >>be

> > > >>

> > > >>

> > > >>>a great fit.

> > > >>>

> > > >>>Thanks in advance for your help.

> > > >>>Roy

>

>

>

>

> to unsubscribe, go to www.yahoogroups.com and follow the

instructions

> Yahoo! Groups Links

>

>

>

>

>







----------------------------

#6280 Jun 26, 2004

Agreed. On a recent road trip on a straight, flat freeway with

cruise control engaged, I tested this by noting milage markers and

using a stopwatch to time the distance between them. I tossed my

notes but recall the GPS velocity as being within 1% - er, ok, my $2

stopwatch and the Garmin IQue agreed to within 1% (no SA). Pretty

suprising to me.

--- In piclist@yahoogroups.com, "Michael Puchol" mpuchol@s...> wrote:

> Hi,

>

> With current GPS without SA, the velocity accuracy that can be

achieved on

> flat ground is quite high, much higher than by other means. It is

true that

> GPS has errors, but I find it WAY more accurate than any vehicle

tachometer

> I have tried. In general, at least in the vehicles I've driven, the

tacho

> shows a figure some 10% higher than the true speed. Doing distance

tests

> with a GPS and the car's tacho, the GPS achieved a much more

accurate result

> in terms of velocity.

>

> Another factor is that GPS doesn't rely just on position changes

over time

> to measure velocity, the calculation is actually done using the

> pseudo-ranges and pseudp-range-rates (PRs and PRRs), which include

the

> apparent doppler shift in the signals.

>

> I know of a Danish-made device which reports distance travelled

using SMS

> messages, which all it has is a tachometer sensor which measures

turns. To

> calibrate it, you drive around for a while, and send it the distance

> readings on the car's tacho, from there onwards it can tell you

distance

> travelled.

>

> Best regards,

>

> Mike

>

>

> ----- Original Message -----

> From: "rtstofer" rstofer@p...>

> To: piclist@yahoogroups.com>

> Sent: Saturday, June 26, 2004 4:34 PM

> Subject: [piclist] Re: Speedometer Calibrator GPS cal

>

>

> >

> > There is a mathematical fallacy in trying to get speed from a GPS

> > and it has to do with the uncertainty of the distance covered

during

> > a sample period. Remember that EVERY point given by the GPS has

an

> > error - less now that selective availability has been removed but

> > still real.

> >

> > Take 60 MPH for example - that is 88 feet per second. Now, if the

> > GPS measured 95 feet (only a 7 foot difference) that would be 65

> > MPH. If the measurement was 81 feet the speed would be 55 MPH.

And

> > again, EVERY point was an error. The more distance you cover

during

> > the sample period the more accurate the reading will be because

the

> > sample error is a smaller percentage of the distance.

> >

> > I suppose you could use the CEP (circular error probable) and kind

> > of draw the circles and geometrically figure the range of error

but

> > it is sufficient to know that it exists.

> >

> >

> > --- In piclist@yahoogroups.com, Robert Rolf robert.rolf@u...>

wrote:

> > > You should do this at the speeds you commonly drive.

> > > I was rather surprised to find that my '93 Caravan gave accurate

> > > speeds (GPS) for values greater than 70km/hr, but read high

> > > by 5km or so for values lower than this. Bizarre, given that

> > > the dial seems linear, and the pointer is driven by a stepper

> > motor.

> > >

> > > I do remember coming across a web page which mentioned that

> > > the electronic speedometer can be calibrated by the dealer

> > > 'service and maintenance module', presumably by changing some

> > > scale factor in EEPROM (they use HC11's for body controller).

> > >

> > > I also remember reading (decades ago) that car makers

deliberately

> > > made mechanical speedos read high at high speeds to fool people

in

> > > driving lower. This seems to be the reverse of that. Makes you

> > drive

> > > slower in town.

> > >

> > > Robert

> > >

> > > Don Hackler wrote:

> > > >

> > > > I've found the easiest speedometer check is using a GPS. The

> > Magellan

> > > > units (and I assume the others) have a display mode that will

> > show

> > > > speed. Toss the GPS on the dashboard, set the cruise, and

see

> > how far

> > > > off your speedo is. In my car, when the actual speed (by the

> > GPS) is

> > > > 65, my speedometer shows 70. This works best on reasonably

> > straight and

> > > > level roads..

> > > >

> > > > - Don.

> > > >

> > > > Dave Mucha wrote:

> > > >

> > > > >--- In piclist@yahoogroups.com, "rtstofer" rstofer@p...>

wrote:

> > > > >

> > > > >

> > > > >>I haven't seen the device you are looking for but I can tell

> > you

> > > > >>that it is common for the mechanical cable to rotate at

either

> > 800

> > > > >>or 1000 revolutions per mile. Which it is depends on the

> > > > >>manufacturer.

> > > > >>

> > > > >>

> > > > >

> > > > >

> > > > >Unfortunatly, it also depends on tire inflation and wear as

> > well as

> > > > >the car being fitted with proper tires.

> > > > >

> > > > >The last time I worked on a car was an '83 cutlass that kept

> > breaking

> > > > >cruse control cables.

> > > > >

> > > > >Almost everywhere you go, there are calibrated mile markers.

> > > > >Typically a 2 mile stretch where you get your cruse control

up

> > to

> > > > >speed, pass the first marker, then a mile later, a second

> > marker.

> > > > >And, often those have reflectors that you can use for

feedback.

> > > > >

> > > > >Dave

> > > > >

> > > > >>I don't know what car you have and even if I did I wouldn't

> > know

> > > > >>anything about it. The last car I messed with was a '73

> > Porsche

> > > > >>

> > > > >>

> > > > >914

> > > > >

> > > > >

> > > > >>that I used in sports car rallies. I had just graduated

with a

> > > > >>

> > > > >>

> > > > >BSEE

> > > > >

> > > > >

> > > > >>and I built a rally computer with dimmable incadescent

> > displays and

> > > > >>all that stuff. In rallies there is a calibration leg where

> > you

> > > > >>have the opportunity to match your odometer to the rally

> > master's.

> > > > >>Hence, I know a little about the old ways of doing things.

> > This

> > > > >>thing had a barn full of 74xx logic and could easily drain

the

> > > > >>battery in an hour or so.

> > > > >>

> > > > >>But it was fun!

> > > > >>

> > > > >>Anyway, if you have a strictly digital system it should be

> > easy to

> > > > >>grab the pulses coming from the transmission, clean them up

and

> > > > >>stuff them into a PIC. Then you have to do some type of

> > > > >>

> > > > >>

> > > > >computation

> > > > >

> > > > >

> > > > >>and output a similar pulse to the odometer.

> > > > >>

> > > > >>If you don't get more help here, look into taxi meters.

They

> > must

> > > > >>do the same kind of thing.

> > > > >>

> > > > >>--- In piclist@yahoogroups.com, "Roy Wheaton"

> > roy_wheaton@y...>

> > > > >>wrote:

> > > > >>

> > > > >>

> > > > >>>I've noticed several products for autos and trucks that

> > provide

> > > > >>>

> > > > >>>

> > > > >>speed

> > > > >>

> > > > >>

> > > > >>>corrections for electronic speedometers. Is this something

> > that

> > > > >>>

> > > > >>>

> > > > >>can

> > > > >>

> > > > >>

> > > > >>>be done with a PIC? How does this type of device work? I'm

> > > > >>>

> > > > >>>

> > > > >already

> > > > >

> > > > >

> > > > >>>using a PIC in my Toyota for various other functions, so it

> > would

> > > > >>>

> > > > >>>

> > > > >>be

> > > > >>

> > > > >>

> > > > >>>a great fit.

> > > > >>>

> > > > >>>Thanks in advance for your help.

> > > > >>>Roy

> >

> >

> >

> >

> > to unsubscribe, go to www.yahoogroups.com and follow the

> instructions

> > Yahoo! Groups Links

> >

> >

> >

> >

> >







----------------------------

#6281 Jun 26, 2004

Most modern GPSs derive velocity directly from the ranging signals

(relative doppler)

It is only absolute position that is subject to SA

(selective availability).

If your postulate were correct, then a stationary GPS would be seen

to be moving about (non-zero velocity) as each sample point randomly

moved about due to measurement error and SA. How else can it

tell me that I am moving at 0.1km/hr as I walk at a slow pace,

when the SA error is 3M or more?



Robert

No GPS geek, but somewhat aware because of GPS timing interests.



rtstofer wrote: >

> There is a mathematical fallacy in trying to get speed from a GPS

> and it has to do with the uncertainty of the distance covered during

> a sample period. Remember that EVERY point given by the GPS has an

> error - less now that selective availability has been removed but

> still real.

>

> Take 60 MPH for example - that is 88 feet per second. Now, if the

> GPS measured 95 feet (only a 7 foot difference) that would be 65

> MPH. If the measurement was 81 feet the speed would be 55 MPH. And

> again, EVERY point was an error. The more distance you cover during

> the sample period the more accurate the reading will be because the

> sample error is a smaller percentage of the distance.

>

> I suppose you could use the CEP (circular error probable) and kind

> of draw the circles and geometrically figure the range of error but

> it is sufficient to know that it exists.

>

> --- In piclist@yahoogroups.com, Robert Rolf robert.rolf@u...> wrote:

> > You should do this at the speeds you commonly drive.

> > I was rather surprised to find that my '93 Caravan gave accurate

> > speeds (GPS) for values greater than 70km/hr, but read high

> > by 5km or so for values lower than this. Bizarre, given that

> > the dial seems linear, and the pointer is driven by a stepper

> motor.

> >

> > I do remember coming across a web page which mentioned that

> > the electronic speedometer can be calibrated by the dealer

> > 'service and maintenance module', presumably by changing some

> > scale factor in EEPROM (they use HC11's for body controller).

> >

> > I also remember reading (decades ago) that car makers deliberately

> > made mechanical speedos read high at high speeds to fool people in

> > driving lower. This seems to be the reverse of that. Makes you

> drive

> > slower in town.

> >

> > Robert

> >

> > Don Hackler wrote:

> > >

> > > I've found the easiest speedometer check is using a GPS. The

> Magellan

> > > units (and I assume the others) have a display mode that will

> show

> > > speed. Toss the GPS on the dashboard, set the cruise, and see

> how far

> > > off your speedo is. In my car, when the actual speed (by the

> GPS) is

> > > 65, my speedometer shows 70. This works best on reasonably

> straight and

> > > level roads..

> > >

> > > - Don.

> > >

> > > Dave Mucha wrote:

> > >

> > > >--- In piclist@yahoogroups.com, "rtstofer" rstofer@p...> wrote:

> > > >

> > > >

> > > >>I haven't seen the device you are looking for but I can tell

> you

> > > >>that it is common for the mechanical cable to rotate at either

> 800

> > > >>or 1000 revolutions per mile. Which it is depends on the

> > > >>manufacturer.

> > > >>

> > > >>

> > > >

> > > >

> > > >Unfortunatly, it also depends on tire inflation and wear as

> well as

> > > >the car being fitted with proper tires.

> > > >

> > > >The last time I worked on a car was an '83 cutlass that kept

> breaking

> > > >cruse control cables.

> > > >

> > > >Almost everywhere you go, there are calibrated mile markers.

> > > >Typically a 2 mile stretch where you get your cruse control up

> to

> > > >speed, pass the first marker, then a mile later, a second

> marker.

> > > >And, often those have reflectors that you can use for feedback.

> > > >

> > > >Dave

> > > >

> > > >>I don't know what car you have and even if I did I wouldn't

> know

> > > >>anything about it. The last car I messed with was a '73

> Porsche

> > > >>

> > > >>

> > > >914

> > > >

> > > >

> > > >>that I used in sports car rallies. I had just graduated with a

> > > >>

> > > >>

> > > >BSEE

> > > >

> > > >

> > > >>and I built a rally computer with dimmable incadescent

> displays and

> > > >>all that stuff. In rallies there is a calibration leg where

> you

> > > >>have the opportunity to match your odometer to the rally

> master's.

> > > >>Hence, I know a little about the old ways of doing things.

> This

> > > >>thing had a barn full of 74xx logic and could easily drain the

> > > >>battery in an hour or so.

> > > >>

> > > >>But it was fun!

> > > >>

> > > >>Anyway, if you have a strictly digital system it should be

> easy to

> > > >>grab the pulses coming from the transmission, clean them up and

> > > >>stuff them into a PIC. Then you have to do some type of

> > > >>

> > > >>

> > > >computation

> > > >

> > > >

> > > >>and output a similar pulse to the odometer.

> > > >>

> > > >>If you don't get more help here, look into taxi meters. They

> must

> > > >>do the same kind of thing.

> > > >>

> > > >>--- In piclist@yahoogroups.com, "Roy Wheaton"

> roy_wheaton@y...>

> > > >>wrote:

> > > >>

> > > >>

> > > >>>I've noticed several products for autos and trucks that

> provide

> > > >>>

> > > >>>

> > > >>speed

> > > >>

> > > >>

> > > >>>corrections for electronic speedometers. Is this something

> that

> > > >>>

> > > >>>

> > > >>can

> > > >>

> > > >>

> > > >>>be done with a PIC? How does this type of device work? I'm

> > > >>>

> > > >>>

> > > >already

> > > >

> > > >

> > > >>>using a PIC in my Toyota for various other functions, so it

> would

> > > >>>

> > > >>>

> > > >>be

> > > >>

> > > >>

> > > >>>a great fit.

> > > >>>

> > > >>>Thanks in advance for your help.

> > > >>>Roy

>

>

> to unsubscribe, go to www.yahoogroups.com and follow the instructions

> Yahoo! Groups Links

>

>

>

>






S
e
n
i
o
r
T
u
b
e
.
o
r
g