Request investigation help into odometer oddity
 

OK, here's the deal. I like to keep my trip odometer (showing tenths of miles) evenly paired to my main odometer (showing only whole miles). I've done this for my last two cars, so when I hit 1000 miles I did it with my 8.

At 999 miles I got myself to a big open parking lot (at a nearby highschool) and drove around at ~10MPH, watching the main odometer. When it hit 1000, I instantly hit the brakes. OK, so now I know that the odometer should be at exactly 1000 miles. So I reset the trip "A" odometer to 000.0 miles. So they're at the same place, but the trip is missing the thousands column.

This weekend I hit 1600 miles. When the main odometer hit 1700, the trip "A" odometer was showing 1699.7 miles.

I know .3 miles off out of 600 isn't much, but legally the main odometer has to be exactly true, or the manufactuer (if it's a manufacture defect) can get into big trouble. So can anybody else also do this, and then keep track of when the main odometer flips over vs the trip odometer? I'm curious also whether there's any way to confirm whether it's the main odometer or the trip odometer that's off.

i'll check this the next time i fill up (i don't try to match whole numbers but i reset at every fill up).

edit: my guess is the main odometer has some "fudge" built in. meaning, if it was analog, it wouldn't have read exactly 1700...that last "0" would not quite be in place yet but almost there...u know what i mean by that? that's just a guess though. if it was fairly accurate on your previous car (with analog odometers) then i'm at a loss.

I know what you mean, but the difference has been increasing. It has been pretty steady, every increment of around 200 miles I notice that it's off by another tenth.

It's hard to do at a fill-up, because you don't necessarily fill up exactly when you roll over miles.

Wait a second...

The main odometer shows whole miles, and the trip odo shows down to tenths, right?

Dude, 1699.7 rounded off to the nearest mile is 1700.

---jps

OK, please read carefully.

I synchronize them at the moment when the main odometer rolls. If it is accurate (and by law it must be), this is the exact moment it rolls from 999.9999999999... miles to 1000.00000000... miles. At this exact moment I slam on the brakes - no more miles (or even inches) are added to my car. I then clear out the trip odometer, meaning it should be set to 000.000000000... miles. Every inch I move after that point should update both odometers the same distance.

So how does it happen that the main odometer rolls from 1599.999999999... miles to 1600.000000000... miles, but the trip odometer is still at 1599.7xxxxxx... miles? An odometer is not an approximation, it is exact. By law. Otherwise, what's to stop a car from reporting a mile when it's actually only been 0.5 miles? Or 1.49 miles? Either one rounds to 1 mile, but that's not allowed. It should only report one mile when it's gone EXACTLY one mile. And they can't be reporting the same thing when they start unified and slowly change the point when they change miles.

Just because the main odometer doesn't show smaller parts of a mile, that doesn't mean that it just kinda picks some random point that feels like a mile and updates. The computer calculates fractions of a wheel revolution. And since it knows the circumference of your wheel, that translates into calculating inches.

So, to answer your specific question Sputnik, I will be happy to report back in about 2000 more miles - at which point, at this rate, it will have lost one whole mile between the two odometers. Would that satisfy you that there's something strange between the two?

Well, if I still remember how to use a calculator properly, .3 mi difference every 600 miles would equal 50 miles difference after 100,000 miles. Probably won't have a huge effect on your resale value. But, I see your point and it would be interesting to know whats going on.

Car is for driving......
 

MAN, stop doing slow circles in dark parking lots, obsessing about how accurate your 'ODO' is.......

Maybe drive over to SEARS, and buy a life......

Seriously,

doc

lol

I probably wouldn't care, but all three of the Hondas I've owned have had a much more severe problem that Honda wouldn't acknowledge.

For my Hondas, I'd lose about 5000 miles out of 100k. And that is severe and would affect resale.

So ever since then I always watch between the two.

(93 Accord, 98 Prelude, 99 Civic, in case you were wondering)

Just a note about speedometers and digital displays.

A speedometer is a measuring device and as such has an accuracy. We can't expect PERFECT readings due to many reasons such as tire size, air pressure and temperature and how much tread is left.

A 26 inch diameter tire measures 81.7 inches in circumference. If you wear 1/4 inch of tread off, the tire will be 25.5 inches in diameter and 80.1 inches around. That's about a 1% error!

Additionally, Any digital display can only be accurate to Plus or Minus 1 of the least significant digit.

As accuracy increases, cost increases. I don't know what accuracy manufacturers aim for but it certainly isn't x.99999999999.

Maybe we can hope to only have 2-5 % accuracy and drive in straight lines.

BTW, I don't suppose you know which tire is used to measure the distance while driving in your randomly sized circles?

Should put this in perspective as pulinteef noted 50 miles of 100,000 mi is the error. This represents an error of 0.05% or 5 basis pts. The world is not perfect my friends and if any instrument of measuerment provided 99.95% accuracy I wouldn't cry about it.

You should go for a drive and relax. I'd be more worried about the oil pan problem folks are experienceing. Even if that hits 0.05% of cars produced I could sleep at night. But if it went to 5% well I'd feel bad about my purchase.

Guys,

We all know that the odo/trip system has a certain accuracy. However, the problem is that there is obviously a bug in the firmware that computes the distance from the pulses coming from the speed sensor.

Reasoning (from my, as a firmware engineer's perspective):

There is a single source (the speed sensor) that is used as the base of all the computations including both the ODO and TRIP counters. Therefore, regardless the accuracy of the base signal, all three counters should be of the same accuracy. The ODO only doesn't display the tenth, but still, it keeps track of them.

So it comes down to a bug in the counter firmware no matter how it is implemented. Just my 2cents.

Erik

i can't believe this. why would .3 mi be a big concern. it won't hurt anything in the long run and it's nothing to worry about. it's .3 miles. just be happy that you have an 8 and leave the small details alone.

Thank you, Erik, for understanding what I've been asking and wording it better (and simpler) than I.

I don't care if it's off by some slight amount, I care why the two are diverging, despite beginning at the same point and generating their data from the same source.

Sputnik is probably right. What happened is that when you saw 1000 it was probably 999.7. I would imagine when you reached you next milestone of anykind you will be off by about the same amount.

If it started out the same and has gotten a little worse every 100 miles then I missed that post.

Discount.....
 

And don't discount the value of good psychiatric treatment.....

Every ~200 miles I notice that the main odometer clicks over .1 sooner on the trip odometer. As such:

They have the same input, but are generating different output. I'm just wondering if anybody wants to confirm or help me to understand why this would happen.

I see erik's point as well...

if there is a firmware problem, and say it resulted in a non-linear difference between the odometer and the tripometer then in the long run the odometer may be significantly off.

But in all cases, I don't think you will need to worry about it. I don't see the error to be that significant.

When i purchased new tires for my CL, the rounds-per-mile (as calcuated by tirerack) was 2-3% less than the OEM size. I didn't care because it's not going to be that different in the end. The performance tire is what I was after. And anyway, who knows...by the time you get 50,000 miles on your car they won't sell RE-04 (or whatever that OEM tire is) and you will have the same situation as I did...

i think its just the difference betwween the tenths on the odo and the tenths on the trip. you say you are slamming on the brakes as soon as it trips but if someone would do the math(not me i hate math ;) ) whats a couple or 3 inches difference over 200 hundred miles? would that account for the "slippage " he is seeing?

whoa man.....i finished with a statistics course this week and the last thing i need is to see or read about are numbers hahah

at least with your 8 now, you aren't worse off than your other civics, which i think you should be thankful for but good luck trying to find a solution