Speedometer accuracy

[grzewnicki]

I believe I read that factory speedometers (cable driven) had to be within +/- 3% from the factory. I imagine today with electronic speedo that accuracy is probably closer to .5%. My 2004 Explorer speedo was always off about 3 MPH (fast). Then when I got new tires and decided to replace the factory spare I found out why. There was a good .75" -1" difference in tire height between spare and new, in exactly the same size (rim-width and profile) tire.

 

[biker]

I believe I read that factory speedometers (cable driven) had to be within +/- 3% from the factory. I imagine today with electronic speedo that accuracy is probably closer to .5%. My 2004 Explorer speedo was always off about 3 MPH (fast). Then when I got new tires and decided to replace the factory spare I found out why. There was a good .75" -1" difference in tire height between spare and new, in exactly the same size (rim-width and profile) tire.

After 30 years of running radar on all kinds of roads, I found speedos to be about 2-3 km/h optimistic (reading too high, including "factory calibrated" police speedos). This was accepted as expert evidence at a speeding trial where a guy tried to argue that his speedo was reading too low. In the end, the J.P. accepted the radar and my use of it as reasons for conviction, but also mentioned in his decision my experience with radar and speedos, and the accused's lack of testing or documentation to support his low speedo theory.
This, of course, only applies to a car with stock gearing and tire size.
My new dash has a gps speedo, hoping to install soon.

 

[ez2cdave]

Something to remember . . .

Even if "calibrated by NASA" (LOL), a cable-driven speedometer will only be accurate at the time it is calibrated.

WHY ?

As tires wear, their diameter decreases . . .
As a result, engine RPM increases . . .
As engine RPM increases, the rotational speed of the drive gear in the transmission increases . . .
As the RPM of the drive gear increases, the indicated MPH on the speedometer increases !

Basically, the effect is the same as progressively lower ring & pinion gears . . . Acceleration should improve as the tires wear and the car will "slow down" with new rear tires, because of the change in height.

Then, the process repeats . . .

Also, any cable-driven speedometer discrepancy is always a % ( percentage ) and the inaccuracy will increase, as speed increases !

Dave F.

 

[Desertfox63]

Using the GPS on your cell phone via Google Maps, or Waze works well. The city has set up some photo radar traps in my area, and since my speedo is in MPH, my wife tells me how fast I am going in KPH so I stay under the photo radar set limit.

 

[Patstuff28]

Now that everyone has a GPS on their phone, you can pick up a dedicated GPS very cheaply now. I have a TomTom unit. It is very accurate in either mph or kph. It also shows the speed limit of the road that I am travelling and alerts of red light cameras.

 

[scrap--metal]

The speedometer in my Camaro seems accurate, but it has some level of hysteresis. When I get going on the interstate, I know I'm going faster than my speedo indicates for probably a few miles. Then it seems to catch up. If I decelerate, it seems to take a little bit to find the accurate speed again. It doesn't bounce or anything, so it doesn't bother me that much.

The 140mph range could definitely be real.

 

[wped777]

Thanks everyone for good tips. I found a site called Omnicalculator.com. It will calculate RPM, speed, gear ratio and tire diameter. Enter three of the fields and it will calculate the fourth. Saves headache doing the math. I also bought an app called SpeedTracker. It gives a readout like an analog speedometer and you can record highest speed.

 

[FS87LT]

A friend had a new '78 Z/28 4sm. He said he got tired of TAs outrunning him in high gear, UNTIL he added a 2nd full cat converter to the lh side. END of that problem!

The Chevy parts book has a very good speedometer gear chart in it, related to production tire sizes and rear axle ratios.

Back then, I'd use the mile markers on the Interstate to check the speed and distance calibration of my cars. As there were allegedly some irregularities in their placement to each other, I'd use a 10 mile run, for good measure. Easier to figure things that way too! Quite easy in that 60mph = 60 seconds/mile. For 5 seconds either side of that, the speed increase/decrease is pretty linear, but after that, you need a digital wrist watch (in a time well prior to GPS).

Only thing with the "highway driving" method, you had to be very vigilant to maintain a constant speed, up hill and down, for what I determines to be better accuracy. Especially for a 10 mile run. 5 miles can work too, depending on finding a flat section of road that long.

The old AC Speedometer catalog mentioned a calibration window of something like "-1 to +3mph", as I recall.

Police speedometers, i.e., "Calibrated" were supposed to be -/+ 1mph from indicated speed. A special clockspring allowed for this tighter variation. At ambient interior temps of below 85 degrees F, IIRC.

On the earlier than middle 1980s cars, when they went to more-teeth gearing, they also used adapter boxes to fill the gaps. On my '77 Type LT OEM 14" radials and 2.56 PosiTrac, it had an adapter box. When I went to the P225/70R-15 Z/28 sizes, after using the GM speedometer gear charts, we removed only the adapter box and the indicated speed was within 1mph or indicated, in the Interstate speed range. Which meant 2000rpm on the factory tach was 62mph indicated.

Back when my car was newer, when a customer came in with a speedometer speed inaccuracy complaint (or speeding ticket issue), it was common for the tech to just ask for a "1 tooth less" driven gear, as a guess. Might have gotten the speed indicated to be more accurate, but the distance would now be off. Which might later relate to "poor fuel economy" complaints, possibly.

I prefered to use the numbers approach, which included the revs/mile of the tire size, the rear axle ratio, etc. to get a "revs mile" of the speedometer cable end at the speedometer head. Then alter the gearing combination of the speedo gears to get to where you desired to be. Which might also need to include a new adapter box.

Rather than talk about tire diameter, instead go into the TireRack.com "specs" page for your tires and get that "Revs/Mile" spec. Then work from there. Revs/mile is a more accurate way to do it, from my experiences. There is a tire size comparator in the Miata website, and a few other similar places.

Sorry for the length, just my experiences,
FS87LT

 

[wped777]

I had to read it twice before it all sunk in. Good info. Another way to skin the cat.

 

[Twisted_Metal]

Just hang your phone out the window and record the exhaust note @ 60 MPH.

Then download a graphing frequency analyzer and measure the Dopler shift as a percentage of the speed of sound (at current temp and air density) and use it to calculate your actual speed.

Yeah… That would work.