First, let me start by throwing out a big thank you to Kleistang for hooking me up with a set of MT ET Streets for my 05.

Changing a wheel/tire combo is always a challenge with our cars, because it seems like every diameter change (16s to 18, 17s to 20s, etc..) results in a speedometer that reads wildly off.

In my case, I'm going from 20" wheels/tires for DD duty all the way down to a 17" wheel for Drag Radial duty. The difference is significant.

You end up holding a programmer in your hand that's asking you for a number you might not know. My SCT X3, for example, asks for the number of tire revolutions per mile.

Thankfully, the math behind figuring out the circumference of your wheel is pretty simple stuff.

All you have to do is take the diameter of the wheel/tire combo and multiply it by Pi, which we all know to be 3.14. Your answer is the circumference in inches.

Then, to figure out how many revolutions per mile you turn, you simply take the number of inches in a mile (63360)and divide by the value you get from the above formula. (Pi*D)

Now, where I ran into an issue was when I went to Pirelli's website and actually looked up the RPM value. What I found didn't make sense.

According to www.tirerack.com, the RPM value for this tire is actually 728.

So you have to ask yourself; how do they get a different value than I do when I simply do the math?


:115: The answer is in the weight of the car. :115:


When you put the weight of the car on the tires, two things happen: the sidewall compresses, and the tread flattens out. Flattening out the rubber tread against the ground effectively decreases the overall circumference of the tire. This gives you more revolutions per mile than using the simple calculation above.

Think "Silly putty on a news paper." (An over-simplification, I know, but you get the point.)

So, now we realize a need to use a different calculation to get the value you need to properly correct the speedometer. This leaves you with two choices.

1. Mark the tire and the ground. Then roll the car forward one complete revolution, measuring the distance between the marks.
2. Take the distance from the center-line of the wheel down to the surface and multiply it by 6.283.

I use the second method, because while it's hard to get a measurement that's 100% accurate even to 1/10th of an inch, you can get pretty close with a simple tape measure. It's also a lot less work than marking the ground/tire and rolling the car around.

So, when you drop a tape from wheel center to ground, you get a measurement of roughly 14". When you do the math, you get the following:
I'm sure I'd be much closer to that number if I could get a more accurate measurement, but I'm lazy right now.

The next time you think about changing tires, hopefully now you know how to figure out the proper diameter and revolutions per mile you have to enter into your programmer.

 

I just got back from my first fishing trip of the year and I'm really tired. You made my head hurt. LOL!

 

When you use PI you really should use more than two decimal places. Otherwise calculation error will be larger than you'd like and it'll just propagate as you do more and more calculations. Don't fear the precision.

 

.......

 

Generally, I use 3.14159. For whatever reason, that's how much of Pi has stuck in my head all these years.



Also, and I'll edit the original thread if I can, I misspoke regarding the effect weight has on the tire.

The weight of the car DECREASES the circumference, thereby giving you more revolutions per mile.

I'm surprised no one caught that.

FWIW: It's much easier to do with an Excel spreadsheet.

If the speedometer is off, won't that give you an incorrect reading?

 

I use my GPS to get my true speed, then tweak the tire revs per mile with my Xcal2 (which only allows me to set a number at 10 RPM intervals). Even as an engineer, I'll take real-time numbers over calculations any day.

 

I am an engineer too, but I just use what ever the manufacturer quotes. Reason being the speedometer is always wrong. The algorithm they use to calculate speed is only true at one speed, 55 I think but I am not sure. So every other speed is wrong. As long as I am close to my actual speed I am fine. When I run a GPS unit in my car it states my speed is about 3mph faster than the speedo at 80.

Close enough for Government work.

 

WHAT
Mathmatical equations that are only correct at 55 MPH?
How did we ever put a man on the moon with thinking like this?

 

Maybe Algorithm wasnt the best choice of words. I am no speedometer expert so maybe there is some fancy electronic way of doing now. My comment is based on the old speedo gear from the trans. Basically it was all mechanical so therefore it was only "perfect" at one point and like I said I believe it was 55mph as that was the prevalent speed limit at the time.

The main point of my thread is that close is good enough for me, but I can understand if the OP wants things optimized. I am overly precise on alot of other things just not speedo calibration.

 

Oh and 157db, if that pisses you off dont even start reading about Tachometers. They are off by a certain percentage also.