• Protip: Profile posts are public! Use Conversations to message other members privately. Everyone can see the content of a profile post.

My speedometer is 5mph off!

Joined
15 June 2002
Messages
179
Location
Palm Beach Florida
My car now has 6k miles on it.

I wonder how many REAL miles the car has on it.

When the Speedo shows 80 the car is going about 75 on GPS.

------------------
James
2000 Red/Tan
South Florida
http://members.aol.com/flnsx
 
There are a number of possible causes:

- Tire sizes other than stock

- Wear on the tires (the tire diameter varies 2 percent from full tread to the treadwear indicator bars)

- Inaccuracy of the speedometer, either by design or production variances

- Accuracy of the speed registered on the GPS (?)
 
Originally posted by nsxrock:
My car now has 6k miles on it.

I wonder how many REAL miles the car has on it.

When the Speedo shows 80 the car is going about 75 on GPS.


Funny, mine seems that way as well- off by ~5mph.
-Thomas


------------------
'94 Brooklands Green Pearl/Black NSX
5-Speed
#157
 
Originally posted by nsxtasy:
There are a number of possible causes:

- Tire sizes other than stock

- Wear on the tires (the tire diameter varies 2 percent from full tread to the treadwear indicator bars)

- Inaccuracy of the speedometer, either by design or production variances

- Accuracy of the speed registered on the GPS (?)

In reference to the GPS on mine (if you are speaking of the same thing) there is settings
for calibration..I might check that 1st'.
MYNSX
 
I notice the same 5mph difference on both my 91 NSX and my 01 GSR when tracking speed with my portable Garmin GPS V receiver.

Vytas
 
I have a Garmin GPS.

They are supposed to be almost 100% accurate. People claim that a GPS system is possibly the only truly accurate way to judge true speed. I don't know if I believe that.

My car has 4 new tires with only 1k miles on them. They are stock tires I would never buy non OEM tires
smile.gif


My car seems way off. I took a 1500 miles trip back from Connecticut with the car. I registered over 1700 miles on the OD and I didn't side track or get lost once. I know these maps can be off, but not by 200+ miles.

I guess I'll get the dealer to look at it.

------------------
James
2000 Red/Tan
South Florida http://members.aol.com/flnsx

[This message has been edited by nsxrock (edited 01 August 2002).]
 
Originally posted by nsxrock:
I guess I get the dealer to look at it.

Good idea. Sorry, I forgot to mention another possible cause:

- There's something wrong with the car.
 
I noticed similar. On the highway, one of those speed monitor thing showed 71-72 mph when I thought I was doing 75mph on speedo.
 
How are your tires? If they're nearly gone they can account for 2mph. On 1/2 worn OEM tires my '93 was within 1 mph at 60-80 according to my GPS set for updates every second. While many cars are off at those speeds, I'm surprised to hear it is common for the NSX.
 
Interesting. I was in a Ford Taurus once with a Garmin and was astounded at the accuracy of the thing! Seemed like it was accure to the 1/2 mph. I haven't tried it in the NSX yet. I have heard the more satellites it can see in the sky, the more accurate it is on the elevation settings. Maybe this is true with speed reading also.
 
More important than the number of satellites is their location. Assuming at least the minimum of three, the closer to the horizon and further apart the better. Although they intentionally introduce error for security reasons, those are in terms of absolute position. When polling every second or less, relative position is very accurate and if you hold your speed dead steady you will see that the GPS reading does not vary. This is probably the most accurate speed measurement available.

There is a very cool but buggy module for the HandSpring that polls at 0.1 second intervals and uses the latest correction logic. I've considered trying to develop something like the G-Cube based on it but just don't have the time, $ and motivation. Anyone want to help?
 
My speedo is incorrect as well. I don't know if this matters but it "feels" more in line if the TCS is off. Not sure if they're related....my .02
 
My comments about the accuracy of GPS mostly revolve not around the accuracy of its positioning, but how speed is calculated. Speed is difference in distance divided by difference in time. While I do not own a GPS, I believe GPS locations are accurate within something like 50 feet, which is very accurate, and differences in location would have the same accuracy. However, I don't know what the sampling frequency is for the calculation by the GPS. If the calculation is performed every tenth of a second, it might not be totally accurate. If it's averaged over, say, every ten seconds, it might be extremely accurate.

If you have a GPS and are driving at a constant rate of speed, and the GPS reading of your speed does not vary (within 1 mph), then I would assume that it is accurate to a precision of less than 1 mph, so it is not an issue.
 
Some fast refresh rate GP systems are adequate, but the best way to calibrate speedometers are radar systems. Similar to the devices police use, Stalker Progressive radar guns (with software for your laptop) will measure MPH to .0001 %. They also will allow you to chart and overlap multiple runs for extremely accurate 0 to 60 time or whatever speed combination you can think up, including breaking. The magazines use these for their tests; problem is they have a fixed cost of $2500 with only one allowed importer in the US.

[This message has been edited by NSXEE (edited 02 August 2002).]
 
I can't think of a reason to need accuracy better than 0.1 mph for most any purpose, so at 10% the cost of a fancy radar system that is far less convenient, I'll stick with GPS.

Ken, although absolute accuracy of GPS is ~50ft (sometimes more like 150), relative accuracy is far better. In other words, if you stand still they don't think you're wandering around in a 50-150 ft radius area from one second to the next. If their relative accuracy were no better than absolute position they would be pretty useless.

BTW it looks like the module I mention above is no longer made and the company failed. I'm not surprised since their own user forum was full of unhappy people complaining about bugs. Too bad.

The old site was: http://www.geodiscovery.com/

I see a few people are still selling old stock, but I think I'd pass.
 
speed measurements on GPS units are almost dead on 100% accurate. in fact, there was a case a while back in canada as i recall where an off duty officer in his personal car was following a speeder and issued him a ticked based on his GPS readings. it went to court and the usual professional witnesses were brought in from both sides and it was upheld. i'll see if i can find the link, but GPS measurement is more accurate than radar or lidar (no cosine error) and less prone to error, though it's a different methodology (pacing rather than point and shoot).

ps, my NSX speedo is right on with my GPS (Garmin III+ )


[This message has been edited by robr (edited 02 August 2002).]
 
here's some prelim info, i read more than this after it went to court (although its possible it was in the NMA newsletter which is only available on paper as far as i know)

Ottowa, Canada

(9/00) For probably the first time in Canada,
a court yesterday accepted data from a global positioning system unit as evidence of a vehicle's excessive speed. Const. Gatien was following a speeding car and decided to estimate the rate of speed on his
GPS unit instead of his speedometer because the unit is considered more accurate.
 
Actually on a 4D GPS with decent signal strength you'll find that the accuracy is down in the 18-25ft range.

My GPS reads MPH to the .1 mph accuracy and does change as my speed changes on the road.

I have also tested my Garmin against my speedo AND a police radar gun at the same time, and both read within .2 mph of each other, so Im gonna say that the GPS is fairly accurate ;P as is my speedo...

-B
 
Originally posted by sjs:
...Although they intentionally introduce error for security reasons, those are in terms of absolute position.
...
I've considered trying to develop something like the G-Cube based on it but just don't have the time, $ and motivation. Anyone want to help?

BTW, there is no longer an error introduced in GPS - civilians just don't have access to the more precise military signal but the civilian signal is error free (as of May 1, 2000).

That sounds like a fun project for a notebook or a Pocket PC (I have a Vericom but sometimes wish for a G-Analyst).
 
Originally posted by nsxtasy:
My comments about the accuracy of GPS mostly revolve not around the accuracy of its positioning, but how speed is calculated. Speed is difference in distance divided by difference in time. ...

A GPS system is also an accurate time system - it computes time as part of its calculation with atomic clock accuracy. It is accurate to about 14 nanoseconds, and should be able to use that information when computing speed :)

However, most GPS units calculate velocity more directly by measuring doppler shift of the radio signal which yields about 0.5 mph accuracy 95% of the time.

See http://www.gpsinformation.net for far more than you want to know about GPS.
 
Originally posted by wilsonp:
...See URL=http://www.gpsinformation.net]http://www.gpsinformation.net[/URL] for far more than you want to know about GPS.

Hey, great site, thanks!

No time to dig through that now, but are you quite sure about the error factor being off? I'd heard and read more than once that it was simply reduced substantially. But when you say we don't have access to the same signal, isn't that really the same thing? Hmmm.., so you're also saying that the Differential units (DGPS) no longer have any real use?
 
youre talking about selective availability, and yes, its off. they can turn it on selectively in certain locations (like in the middle east during war). the military has a more accurate version that we dont have access to.
 
Originally posted by sjs:
[BHmmm.., so you're also saying that the Differential units (DGPS) no longer have any real use?

[/B]

You'd have to read the site, but DGPS is still better than standard.

OTOH, I think WAAS is the wave of the future and is built-in to my next GPS :-D
 
Getting back to the original query...

Originally posted by nsxtasy:
There are a number of possible causes:

- Tire sizes other than stock

- Wear on the tires (the tire diameter varies 2 percent from full tread to the treadwear indicator bars)

- Inaccuracy of the speedometer, either by design or production variances

- Accuracy of the speed registered on the GPS (?)

I think it's safe to say that the last item listed is not a factor.
 
A GPS is going to be more accurate than any car speedo. But that said, My nsx is dead on to my GPS, that is dead on at 50, 70, 100, and that is about as high as I have gone reading the GPS while driving. The speedodometer also checked out dead on to the radar guns on I-5 north of Redding. My odometer was off by about 1.5 miles at the 260 mile marker in Oregon. That has it about 0.5% off. I guess I'm lucky that with the tires I have and the air pressure I run, my speedo is that accurate. I know my RL, Integra, and other vehicles aren't as good. If you are off by one or 2 percent, you can see if at your next tire change you may look at a different tire size to correct for some of that error if you choose to.

Fritz
 
Back
Top