Interesting Civic tech quirk

[California98Civic]

I have long noticed that my speedometer is about 2.5 MPH too fast, relative to roadside police radar readouts and more recently against my Ultra Gauge. I just stumbled across a possible explanation--though I am not sure. Using this calculator I estimated the difference in my car's stock wheel/tire combo with a 5th Gen Civic's stock wheel/tire combo. The difference in MPH readings is about 2.279 lower with my 14" wheels as opposed to the previous gen's smaller 13" wheels. I know the transmissions often just bolt onto one another, and I have learned that most gearing is about the same 5th/6th gen, but the final driven gear is often taller on the 5th gen cars. Did I figure this correctly?

[cfg83]

California98Civic -

This may not apply to your-gen Civic, but Civic odometers have been incorrect in the recent past :

Odo Uh-Oh: Honda extending warranties on 6 million cars

Honda has decided to settle a class-action lawsuit that alleges its odometers were racking up miles too fast. The automaker says odometers on some 6 million Hondas affected by the suit were accurate to within 3.75% on the high side. The NHTSA doesn't regulate odometer accuracy, and the only industry standard is a voluntary one set by the Society of Automotive Engineers that says odos should be within +/-4%. While the car's affected by the suit fall within that range, Honda recognized that its customers expected their odometers "would be based on zero," and they weren't.

The settlement will lengthen the warranty mileage of affected vehicles by 5% and Honda will pay lease-mileage penalties incurred by owners, which is expected to cost the automaker around $6 million. If you own a 2002 to 2006 Honda or Acura bought between April 12, 2002 and November 7, 2006, then you're eligible for the benefits of the settlement.

A too-fast speedo racks up miles on your odometer (early warranty voiding) and makes you think you are getting better MPG.

CarloSW2

[PaleMelanesian]

My Odyssey was part of that mess. The speedometer is the usual 1-2% fast, but the odometer is more like 5% fast. They're not the same, which makes it even more sneaky.

I have the 175-70-13 tires and my speedo reads about 2.5% fast. That sounds about the same as yours, but we have different tire sizes.

[California98Civic]

Originally Posted by PaleMelanesian

I have the 175-70-13 tires and my speedo reads about 2.5% fast. That sounds about the same as yours, but we have different tire sizes.

Different tire sizes but similar diameters: my 185/65-14 and your 175/70-13 are close, though yours are slightly smaller.

[PaleMelanesian]

Right. 65 vs 70 sidewall almost covers the difference. I always assumed the 14's would have a close-to-accurate speedometer due to the slightly bigger tire.

Are you reading the speedometer mph or odometer miles? They're not always the same.

[California98Civic]

Originally Posted by PaleMelanesian

Right. 65 vs 70 sidewall almost covers the difference. I always assumed the 14's would have a close-to-accurate speedometer due to the slightly bigger tire.

Are you reading the speedometer mph or odometer miles? They're not always the same.

I'm reading the speedometer. It seems 2.5 mph HIGHER than actual speed (not percent but a consistent number of miles per hour, regardless of speed). The ODO I tried to calibrate against Google maps and found a 0.2%-0.4% UNDERcount by the ODO, but I think that test hopelessly flawed because the route was very very hilly and only 11 miles long. Later this month I try again with google maps, a good GPS that gauges altitude, and a 69.2 mile route that is fairly flat.

[Peter7307]

Yep found the same thing with Google and hilly terrain.
The results are far better if you can find either flat ground or gently rolling hills.

On the error in the speedo it may be a built in factor. Some makers side on the "reading higher than actual speed" aspect to help keep drivers legal and avoid any clams against them for speeding fines etc.
it may also be a manufacturing error with the actual speed of the needle being accurate but the scale behind being slightly off if it is truly consistent at all speeds.

Peter.