Speedometer calibration with 3.27 rear end

[Audiophile]

I am going to have Palo Alto Speedometer calibrate my Kph speedometer after I changed my rear end from the original 3.92 to a 3.27.  The kph speedometer came out of a car that had, I believe 3.69 rear end. I am looking for the number of turns the speedometer cable makes in a mile with the standard rear end (3.92). From that I should be able to do a ratio from the 3.92 to the 3.27 which should mean the cable is going to turn 83.4% or 16.6% slower with the 3.27.  Once Palo Alto knows the number of turns the speedometer cable makes in a mile they then can do a calibration.

If it makes any difference the car is a 1970 280SL manual 4 speed.

Any help would be greatly appreciated.

[lurtch]

Hello M - - -
     
   Wayzata,MN is a long way from Palo Alto Speedometer.  Not so for me - - -  I live less than a mile away from this shop. I have used them a number of times on several other of my collectible cars. I am sure they will get this job done right.

 This post interests me because I have the same speedometer/3.27 ratio issues. I am sure there are a bunch of other 113' s that do too.

Could you do a follow up post when you get it back from them?

Larry in Palo Alto, CA

[Richard Madison]

This ratio calculator may help:

http://www.dakota-truck.net/TIRECALC/tirecalc.html

Richard M, NYC

[J. Huber]

Hey Mflaten -- is this something Palo Alto requested from you? Did they mention the "ratio test" they discuss on their website?

Reason I ask is, I had my speedometer "mileage" counter fixed at PAS in 2009-- they repaired it perfectly. Almost. They also fixed my capillary tube on the center gauge. Almost.

Problem is -- when I got it back, my speedometer was way off. And my temperature gauge reads 20 degrees C low when warmed up!  I called them immediately, they said do the ratio test -- then send them both back.

I have not done it! I didn't feel like tearing the dash apart -- and that ratio test looks like a PIA. So I put it off. I'm an idiot I know. I'll be surprised if they honor the mistake years later but they might.

[66andBlue]

Mark,
I don't quite understand why you should do the work for PAS.
They do have conversion tables for axle ratios, all they need to know is which ratio your current speedometer fits and that should be indicated on the back. Are you not able to read the ratio number of your current speedometer? Also if your current speedometer was matched to a 3.69 axle it must have been off with your previous 3.92 axle, correct?
Do you have non-original tire sizes?
If the speedo came from a car with a 3.69 axle why would they need to know the cable rev/mile for a 3.92 axle?  
Makes little sens to me.
If you are not happy with their customer service you might try North Hollywood Speedometer: http://www.nhspeedometer.com/

[DavidBrough]

When I had mine done I compared a few indicated speed readings on the new axle with a sat nav to come up with a percentage error which was then adjusted out by the repair shop.

[Audiophile]

Thanks for all the input.  I spoke to Palo Alto Speedometer and they say do not have information on the output shaft rotation of the various models.  I do know what the speedometer came out of (3.69) and which rear end it was for and I do know what the rear end was before the swap (3.92) and the swap was 3.27.  They say they do not know with out the ratio test they cite on their website.  Since all my gauges are out and I can't start the car with the oil line disconnected, I am going to jack the rear axel and measure the tire circumference and determine how many turns of the tire to go 52 feet 9 1/2 inches and how many revolutions of the speedometer cable.  They they should be able to calibrate the speedometer.

I will let you know what I find.

[Audiophile]

I performed the Palo Alto Speedometer test whereby you count the number of revolutions of the speedometer cable in 52 ft 9 1/2 inches.  Instead of driving or pushing my car this distance I calculated the circumference of the tire (79.48 inches), tires are 205/70/14 and determine I would need to make 7.97 revolutions of the tire to travel the 52 ft 9 1/2 inches.  With one side of the car jacked up I turned the tire and came to 6.0 revolutions of the speedometer cable.  Does this sound right?  Palo Alto seem to think the number should be about 12.

Input is appreciated.

[J. Huber]

Hey I wish I could help -- I am in a similar boat. Have to perform that blasted test before I can get mine re-calibrated. If I can figure out the directions... What stinks is it was fine before I sent it in, and I made it clear mine was a 3.75 but they did something to change it. Now I am at about 120 something KMH at 3500 I think. That ain't right.

If anyone has done the test, I'd appreciate some tips myself -- after you help mflaten of course!

[gugel]

I performed the Palo Alto Speedometer test whereby you count the number of revolutions of the speedometer cable in 52 ft 9 1/2 inches.  Instead of driving or pushing my car this distance I calculated the circumference of the tire (79.48 inches), tires are 205/70/14 and determine I would need to make 7.97 revolutions of the tire to travel the 52 ft 9 1/2 inches.  With one side of the car jacked up I turned the tire and came to 6.0 revolutions of the speedometer cable.  Does this sound right?  Palo Alto seem to think the number should be about 12.

Input is appreciated.

(Posted this yesterday, but for some reason it didn't appear on the site)
When you turn only one wheel, the driveshaft rotates only half as many times as when you rotate both (because of the differential).  So your 6 rotations of the speedometer cable would have been 12 if you had rotated both wheels.

BUT -- while your computation of 797 tire revolutions per mile (7.97 for a hundredth of a mile) is theoretically accurate, in practice the tire rotates a few more times per mile because the weight of the car flattens it slightly on the bottom.  For example, for a few 205/70/14 tires I selected at random, the Tire Rack gives 821, 819, and 822 revolutions per mile (respectively 8.21, 8.19, and 8.22 per hundredth of a mile).  The number is slightly different for different tire makes, and of course  changes a bit with wear.  If your speedometer cable rotates 12 times in a hundredth of a mile with 7.97 tire revolutions for that distance, it would rotate about 12.4 times with 8.22 tire revolutions for that distance.  This would seem enough of a difference to affect speedometer calibration, so you might think about actually pushing the car the hundredth of a mile suggested by Palo Alto to get a more accurate result for your particular tires.

Chris

[Audiophile]

Chris,

How confident are you that when you turn only one wheel, the driveshaft rotates only half as many times as when you rotate both.  Do the spider gear ratio have any affect on that or is it always 1/2?

Thanks Chris.

[gugel]

Chris,

How confident are you that when you turn only one wheel, the driveshaft rotates only half as many times as when you rotate both.  Do the spider gear ratio have any affect on that or is it always 1/2?

Thanks Chris.

It's always 1/2. The ring gear in a differential rotates an amount equal to half the sum of the rotations of the two wheels.  For example, when driving in a straight line, each time the wheels make one rotation each (two rotations total), the ring gear makes one complete rotation ( half of two).  In a turn, say one wheel rotates .8 times and the other 1.2 times, that also produces one rotation of the ring gear (half of .8 + 1.2).  If one wheel is on the ground and the other is rotated, it takes two rotations of the one wheel to rotate the ring gear once (half of 0 + 2).  

Each rotation of the ring gear of course rotates the driveshaft a number of turns equal to the differential ratio -- 3.27 in your case.

For a clear explanation of this in somewhat more detail, you can look at http://www.jedi.com/obiwan/jeep/misc/gearratio.html  For some simulations of a differential in action, see http://auto.howstuffworks.com/differential2.htm

[Audiophile]

Thanks Chris.  Just what I needed.  My speedo is off to Palo Alto.