[Purplehazeffc]

Has anyone else don't a speedo accuracy with a GPS?? Down the road from my house, the council has put up a speed radar sign to keep drivers at around the 50kph speed limit.
Went past it today & sat the digital readout at 50kph. The sign showed I was doing 45kph. So it got me thinking.
So I downloaded an GPSspeed app to my phone to see the difference. It looks like at whatever speed I'm doing, the car is showing 5kph faster.
So if I'm doing 40kph, dash shows 45kph. If I'm doing 70kph, dash 75kph. 100kph dash shows 105kph. 115 dash shows 120kph.
So it's no like it is a set % difference. So then it got me thinking again.
If I spent all its time just doing suburb driving & up & down the freeway. Say averaging 50kph dash.
Then the Real average would be 45kph. Which is 10% difference. So with the 10% difference, at 10,000ks on the odo. Then the car has only done 9000ks.
Or even after a few years with doing 100,000ks. The car has only done 90,000ks. Reselling the car. 90,000ks looks a lot better than 100,000ks. Am I looking & working this out right

Again, has anyone else done a GPS speed test????
I have been thinking about putting my car on a dyno.
To see, if any power  increase from the mods I have done.
As a dyno will also give a calibrate accurate speed readout..

Cheers.
Interested in others thoughts..

[CraigB]

All vehicle speedos are set to read above what you're actually doing, it's set like that from the factory to help prevent speeding

The speedo reading fast shouldn't affect the odometer, no matter what speed you are doing you are still travelling the same distance.

[tw2005]

It's also ADR requirement they be set to indicate more than actual. Interesting to see how your other tests go. For my FD series the speedo is deadly accurate I reckon my error maybe 1 oe 2 ks at 100. I tend to go over on the dial forgetting this, lucky so far with the cameras up here but when i go past one of those council signs I have to be bang on 60

[Dazzler]

Mine is the same, as you would expect. I tested it early in the piece with a phone app.

Interestingly if you have your built in GPS on the map setting any speed limit shown will only flash red when you are exceeding the actual speed limit!

For example I have to get to about 56 kph on my digital readout before the built in GPS will start making the a 50 speed limit shown on the GPS flash brighter read. This is quite handy and I use it as a rough guide when setting my cruise speeds.

Allowing for errors and a bit of speed float I usually set my cruise to 54, 64, 84, 104 etc... 

I'm not stressed about actual distance covered because most of my previous 32 cars have been optimistic by a similar amount.

[The Gonz]

The speedo reading fast shouldn't affect the odometer, no matter what speed you are doing you are still travelling the same distance.

Sorry, Craig, my Engineer senses are tingling. In principle, if the odometer is geared off the same mechanism as the speedomoter, then there is every possibility that they are reading the same error. You can't say the odometer is fine without looking more closely.

As for Mrs G's PD, I ran the DigiHUD app on my phone while she drove and the speedometer reads a consistent 5km higher than GPS reported speed.

[CraigB]

The speedo reading fast shouldn't affect the odometer, no matter what speed you are doing you are still travelling the same distance.

Sorry, Craig, my Engineer senses are tingling. In principle, if the odometer is geared off the same mechanism as the speedomoter, then there is every possibility that they are reading the same error. You can't say the odometer is fine without looking more closely.

As for Mrs G's PD, I ran the DigiHUD app on my phone while she drove and the speedometer reads a consistent 5km higher than GPS reported speed.

Does the speedo an odometer read from the same sensor

[The Gonz]

Nothing to say it does, nothing to say it doesn't. My old HQ definitely ran both off the same spinning cable out of the tranmission. My point is that you can't say the odometer is any more accurate.

[CraigB]

My thinking was that the odo gets the correct reading from gearbox and the speedo is manually offset in cluster, another cause can be due to the 3% diameter variation limit for wheel/tyres setups though that mostly applies to when we start playing with aftermarket setups.

[The Gonz]

True. I'm reminded of the 235-wide Firestone monstrosities I ran on my Dunnydore in the 80s and 90s. However, wheel and tyre variations would affect the discrepancy between gauges and GPS, not between each of the gauges.

[tw2005]

All pulse driven. Each 360 deg will have an exact number. The rolling diameter is known and then there is the interval (frequency ) Digest all that with electronics and send a voltage to the dial for the indication.

Could be from the transmission or wheel sensors.

Both odometer and dial will be using the same data processed in different ways.


Regardless of the wheel circumference  the distance and speed  indicated will be a constant however true speed will always depend on how close to the OEM spec your wheel package is

You could jack the car up .run it in gear at say 2000 rpm in top and indicate 100kph and accumulate distance.

[Cookie Thumper]

My thinking was that the odo gets the correct reading from gearbox and the speedo is manually offset in cluster, another cause can be due to the 3% diameter variation limit for wheel/tyres setups though that mostly applies to when we start playing with aftermarket setups.

It will be read from a wheel sensor, tyre rolling diametre "shouldn't" make a difference.

[Purplehazeffc]

I would have thought that the speed difference in actual speed to dash readout speed would have been variable..
I know in all of the cars I have done a GPS speed were like this.. IE: 5 kph at 100 would be 2.5 kph at 50 kph.
But this is a constant 5 kph speed difference at all speeds. Not just at say 100 kph. It's even 5 kph difference at 40 kph..
So to me that counts out the speed coming of the gearbox. Cause then it would have to be variable difference ????

I also would of thought that the speedo & ODO would be running of the same sensor. But with it being a constant 5 kph difference ????
And with my car being less that 6 months old. The tyre & rim diameter would not come into play at all..

Might have to do a drive & reset the ODO in the car & start up the GPS. As the GPS app also has distance on it as well.

[The Gonz]

Yep, tells me the PDs have precise measurement but not accurate measurement.
By this I mean that it does track linearly across the range but it has intentionally been set 5kph higher, electronically, not proportionally in the analog style.

[Dazzler]

We have a few 5km distance testers down here on the main Highway Burnie to Hobart

[tw2005]

Can't comment on PD but previous series have a VSS fitted to the manual boxes



For the Auto I believe it get's it from the output speed sensor pulses.

The wheel speed sensors are more for ESP ABS etc not used for speedo, at least not on the FD.

Quoted error for FD series , 100 KPH Australia Spec is 3 KPH, at 60  a fraction over 1 kph and that's what I see on my FD

[tw2005]

Yep, tells me the PDs have precise measurement but not accurate measurement.
By this I mean that it does track linearly across the range but it has intentionally been set 5kph higher, electronically, not proportionally in the analog style.

So it's conceiveable there's an algorithm involved ?

[The Gonz]

y=mx+c where m=1 and c=5 

[tw2005]

y=mx+c where m=1 and c=5 

linear, nice one.

 Knew I should have completed an engineering degree.

[josh1990]

When I picked up my GD Tourer, I used a GPS app which not only takes instant satellite speed, it averages distance travelled (lat/long), time elapsed etc which also gives average speed for trip. Anyway, to the point I was at around 115-6 indicated for 110 on the GPS.

[xiziz]

If you hook up to torque app and read the odb speedo it will show the same as gps(rounded to no decimal). The instrument cluster is set to show 5kph over what the sensor reads. So assuming standard tire dimensions it will always be 5kph off.

So atleast the odo should read correctly. Assuming this is the same in the PD.

[beerman]

They all do it. They just can't read faster.

Speedometer Accuracy - Speedo Regulations - RACQ

They can be more accurate.  The old police Commodores and Falcons used to be tested and set to be exact. They now rely on a digital readout in addition to the dash speedo the digital shows true speed the dash shows factory setting. I would suspect that the speedos have a auto deduction to reduce them, whilst the odo gets the correct information but that is a guess.

[tw2005]

They all do it. They just can't read faster.

Speedometer Accuracy - Speedo Regulations - RACQ

They can be more accurate.  The old police Commodores and Falcons used to be tested and set to be exact. They now rely on a digital readout in addition to the dash speedo the digital shows true speed the dash shows factory setting. I would suspect that the speedos have a auto deduction to reduce them, whilst the odo gets the correct information but that is a guess.

I recall seeing a windscreen sticker with the calibration details indicating displayed and true speed once in a cop car.

5.3.               The speed indicated shall not be less than the true speed of the vehicle.  At the test speeds specified in paragraph 5.2.5. above, there shall be the following relationship between the speed displayed (V1 ) and the true speed (V2).

0 < or = (V1 - V2) < or = 0.1 V2 + 4 km/h

[The Gonz]

That '< or =' really threw me!

But I see that that's about the best we can do on a keyboard.

0 <= V1-V2 <= V2/10+4

For the non-mathematicians, this shows that they set their difference in speed to not exceed 10% true+4kph, meaning that if their indication were equal to or lower than true speed or higher than, say 114 at 100, it would fail their desired calibration.

Thanks, Gerard

[tw2005]

That '< or =' really threw me!

But I see that that's about the best we can do on a keyboard.

0 <= V1-V2 <= V2/10+4

For the non-mathematicians, this shows that they set their difference in speed to not exceed 10% true+4kph, meaning that if their indication were equal to or lower than true speed or higher than, say 114 at 100, it would fail their desired calibration.

Thanks, Gerard

could have thrown in a NOR or NAND , would that digitise things

[The Gonz]

could have thrown in a NOR or NAND

I still have cold sweat nightmares about designing half adders from discrete TTL gates!
(Well, I could say I reminisce fondly but it doesn't have the same ring to it.)

[simon@mgr]

They taught us that at RadSchool back in the seventies.
In order to make my speedo more accurate I fitted 205/65 15's, 70 & 80 kph spot on.

[The Gonz]

They taught us that at RadSchool back in the seventies.

Your were just a decade ahead of me there, Simon.