[rich8566]

Originally Posted by Viffermike

This is actually an unofficial mandate by the Feds that goes back to the 55 MPH days in the 1970s. Car manufacturers calibrate speedometers to read slow for two primary reasons:
1). To account for the loss in rolling diameter as a tire's tread wears (lower rolling diameter = faster RPM = higher speedometer reading)
2). To prevent lawsuits and such against car manufacturers, municipalities, states, etc. for false speed readings that lead to unlawful behavior, i.e. speeding tickets

Sorry, but I disagree with your statement that this is a mandate by the Feds.

I currently have two Hondas that have dead on accurate speedos. All of our prior Hondas had dead on accurate speedos. My 2000 SL5O0 has a dead on accurate speedo. Our Audi A4 has a dead on accurate speedo. Our 2014 Ford Escape has a dead on accurate speedo. Every BMW I've had (5 since 2007) were inaccurate to the tune of 3-5%, or 2-3 mph at normal road and highway speeds. .

[Xi-Xi]

Originally Posted by Viffermike

This is actually an unofficial mandate by the Feds that goes back to the 55 MPH days in the 1970s. Car manufacturers calibrate speedometers to read slow for two primary reasons:
1). To account for the loss in rolling diameter as a tire's tread wears (lower rolling diameter = faster RPM = higher speedometer reading)
2). To prevent lawsuits and such against car manufacturers, municipalities, states, etc. for false speed readings that lead to unlawful behavior, i.e. speeding tickets

Most U.S. law enforcement agencies are instructed to NOT pull over vehicles that are traveling less than 10 percent over the posted speed limit. Why? A combination of three reasons:
1). User error/improper calibration of speed-reading apparati (timers, laser/radar, etc.)
2). The aforementioned speedometer calibration tendencies
3). The relatively low income that comes from most low-level speeding tickets (i.e., below 10 MPH, not in a construction zone, etc.), and the likelihood that income will be mitigated by defendant challenge in court. In other words, it's simply not cost effective.

If it's "unofficial", it's not a "mandate". Regardless, I think most would agree that BMW speedos read high, not "slow". In which case, the smaller rolling diameter of the tires would exacerbate the problem, i.e., at a given speedo reading the vehicle is going even slower than with new tires.

[Viffermike]

Originally Posted by rich8566

Sorry, but I disagree with your statement that this is a mandate by the Feds.

I currently have two Hondas that have dead on accurate speedos. All of our prior Hondas had dead on accurate speedos. My 2000 SL5O0 has a dead on accurate speedo. Our Audi A4 has a dead on accurate speedo. Our 2014 Ford Escape has a dead on accurate speedo. Every BMW I've had (5 since 2007) were inaccurate to the tune of 3-5%, or 2-3 mph at normal road and highway speeds. .

You're just flat wrong. Honda in particular has had a reputation for decades for providing readings that are higher than the actual speed.

I don't know how old you are, but I was alive in the 1970s and 1980s when this was an actual issue discussed not only in government but in Car and Driver and Road & Track, if my memory serves me correctly (I subscribed to both as early as 1980). Back then, many speedos were mechanical -- driven by a cable -- and could not be calibrated electronically in any way. You had to change physical gearing to change the calibration. To allow for the factors I mentioned earlier, manufacturers were compelled -- again, unofficially -- to calibrate speedos to read higher than the actual speed. Among gearheads and auto enthusiasts, this was an accepted thing.

Also consider that the average passenger car tire loses 1/2" in diameter as its tread wears. This converts to roughly 2 percent of its rolling circumference -- and with tires with deeper tread such as trucks, the difference is greater. Presto!: your speedometer is reading 2 percent faster with worn tires. This is a physical fact that no electronic calibration can correct, either.

[Luftwaffe1O1]

There is a European law that no speedo can read lower than the actual speed.

This article details it nicely:
https://www.thrillist.com/cars/your-...-japanese-cars


Here is the link to the official page regarding the regulation:

http://www.unece.org/fileadmin/DAM/t...gs/r039r1e.pdf

Basically, in Europe, under no circumstance may a speedo read BELOW the actual value, which is why most german cars read slightly on the optimistic side...


The More You Know...

[rich8566]

Originally Posted by Viffermike

You're just flat wrong. Honda in particular has had a reputation for decades for providing readings that are higher than the actual speed.

I don't know how old you are, but I was alive in the 1970s and 1980s when this was an actual issue discussed not only in government but in Car and Driver and Road & Track, if my memory serves me correctly (I subscribed to both as early as 1980). Back then, many speedos were mechanical -- driven by a cable -- and could not be calibrated electronically in any way. You had to change physical gearing to change the calibration. To allow for the factors I mentioned earlier, manufacturers were compelled -- again, unofficially -- to calibrate speedos to read higher than the actual speed. Among gearheads and auto enthusiasts, this was an accepted thing.

Also consider that the average passenger car tire loses 1/2" in diameter as its tread wears. This converts to roughly 2 percent of its rolling circumference -- and with tires with deeper tread such as trucks, the difference is greater. Presto!: your speedometer is reading 2 percent faster with worn tires. This is a physical fact that no electronic calibration can correct, either.

Sorry - but you are totally wrong.

[Viffermike]

Originally Posted by rich8566

Sorry - but you are totally wrong.

Sorry, but you need to supply proof beyond your own personal accounts.

Beyond what's been supplied by others, here's one of those Car and Driver articles I alluded to earlier. A quote from that article (emphasis added):

In the U.S., manufacturers voluntarily follow the standard set by the Society of Automotive Engineers, J1226, which is pretty lax. To begin with, manufacturers are afforded the latitude to aim for within plus-or-minus two percent of absolute accuracy or to introduce bias to read high on a sliding scale of from minus-one to plus-three percent at low speeds to zero to plus-four percent above 55 mph. And those percentages are not of actual speed but rather a percentage of the total speed range indicated on the dial. So the four-percent allowable range on an 85-mph speedometer is 3.4 mph, and the acceptable range on a 150-mph speedometer is 6.0 mph.

:: drops mic ::

[3.0L]

Originally Posted by rich8566

Sorry - but you are totally wrong.

Reminds of a saying my dad had way back when:

"Don't bother me with the facts, my mind is made up."

[PlayItLoud]

My wife's comment after we got home from picking up our 228i xdrive for a well check. (She followed me home in our slightly lifted Hummer H2). "Why were you driving so slow?"

My answer, "The speedo said 65. I was driving the speed limit."

Her answer, "you replace your 350Z that you drove like the Bandit running from smokey and get a Bimmer. Now you drive like a granny."

My answer, "I was trying not to attract uneccessary attention."

5 weeks later we pickup my M240i xdrive and she follows me home in our lifted Jeep TJ. Same conversation. Only difference....

I answer back, "yes honey. And after she is done with break-in I guarantee I will continue to drive it like a granny. Just like the little old lady from Pasadena."

[PlayItLoud]

Originally Posted by Viffermike

Sorry, but you need to supply proof beyond your own personal accounts.

Beyond what's been supplied by others, here's one of those Car and Driver articles I alluded to earlier. A quote from that article (emphasis added):

In the U.S., manufacturers voluntarily follow the standard set by the Society of Automotive Engineers, J1226, which is pretty lax. To begin with, manufacturers are afforded the latitude to aim for within plus-or-minus two percent of absolute accuracy or to introduce bias to read high on a sliding scale of from minus-one to plus-three percent at low speeds to zero to plus-four percent above 55 mph. And those percentages are not of actual speed but rather a percentage of the total speed range indicated on the dial. So the four-percent allowable range on an 85-mph speedometer is 3.4 mph, and the acceptable range on a 150-mph speedometer is 6.0 mph.

:: drops mic ::

Well known since the early 80's at least. (Yeah, I got grand kids). The oddity was the 2 Nissan 350Zs I owned. In both cases the digital speedometer always read faster than the analog by about 2% to 3%. I tended to rely on the digital more than the analog with those two cars.

[rich8566]

Originally Posted by Viffermike

Sorry, but you need to supply proof beyond your own personal accounts.

Beyond what's been supplied by others, here's one of those Car and Driver articles I alluded to earlier. A quote from that article (emphasis added):

In the U.S., manufacturers voluntarily follow the standard set by the Society of Automotive Engineers, J1226, which is pretty lax. To begin with, manufacturers are afforded the latitude to aim for within plus-or-minus two percent of absolute accuracy or to introduce bias to read high on a sliding scale of from minus-one to plus-three percent at low speeds to zero to plus-four percent above 55 mph. And those percentages are not of actual speed but rather a percentage of the total speed range indicated on the dial. So the four-percent allowable range on an 85-mph speedometer is 3.4 mph, and the acceptable range on a 150-mph speedometer is 6.0 mph.

:: drops mic ::

Just because there is leeway it does not mean that there is any built in error. Sorry but I've confirmed in all of the vehicle I've mentioned that the speedos are dead on accurate.

[rich8566]

Originally Posted by 3.0L

Reminds of a saying my dad had way back when:

"Don't bother me with the facts, my mind is made up."

Just because there is leeway it does not mean that there is any built in error. Sorry but I've confirmed in all of the vehicle I've mentioned that the speedos are dead on accurate.

[aerobod]

With 5 to 8mm of tread wear from new to worn out on a typical car tyre and rolling radiuses in the 300 to 350mm range, I normally expect to see a speedometer variation of 2% due to tyre wear alone, let alone the variation between manufacturers for tyres of the same size and tyre pressure changes. The only vehicle I have ever had from Toyota, Subaru, Honda, Nissan, GM, Ford, Chrysler, Fiat, Triumph, MG, BMW, VW and Caterham that can have a speedometer that is 100% accurate all the time is the Caterham R400. This is due to the fact that I can adjust the number of pulses per kilometre to account for variations in tyre size, it receives about one pulse every 5cm of tyre movement, so can be adjusted in increments of about 0.005 km/h at 100km/h. Given the fact that the Caterham is very hyperactive and hates constant speed or direction, can accelerate 0-100km/h in 3.8 seconds or out brake a Corvette and the speedometer can't keep up with the speed changes when driving it like it was built for, it doesn't really matter that the speedometer can be accurate on that car anyway.

[aerobod]

A few thoughts on variations in speedometer readings due to wear, tyre size variations and pressure variations:

Taking a Michelin PSS tyre in 225/40-18 size, from Tire Rack it has a new tread of 10/32", a new diameter of 25.1" and revolves 829 times per mile. This means that it's radius when new is 318.8mm, rolling radius is 308.9mm and tread wear from new to worn out is 6.35mm (8/32").

Basic trigonometry gives a contact patch length of 158mm in the quoted config. Let's say we have a comfort tyre pressure when the tyre is cold of 2.0 bar (nominally 29psi), when the tyre is heated up this will easily rise to 2.5 bar (nominally 37psi). The contact patch area will stay fairly constant under the same load, so with a weight of approximately 4000N (about 900 pounds) on a tyre and assuming the contact patch is fairly rectangular, it will be 127mm wide at 2.0 bar pressure (this assuming a rectangle, but it will be more oval and wider, but the relative width at a given pressure is more important than the true full width). If the pressure is increased to 2.5 bar, the contact patch area will decrease by 20%. If it shrinks evenly in width and length, it will now be 141mm long and 114mm wide. This means the rolling radius will increase to 310.9mm, or by 2.0mm. This means the rolling radius will increase by 0.65% due to pressure.

The tread depth worn to the minimum will give a rolling radius of 302.6mm, 2.1% less than a new tyre.

Looking at Tire Rack for variation in the max performance summer tyres between manufacturer and tyre model, 823 to 833 rpm seems to be the range, or 1.0%

Taking all these variation together, we see 3.75% change in speedometer reading due to pressure, tyre manufacturer and wear.

To be sure the speedometer never reads low, a manufacturer would need to build in a reading about 4% above the actual speed.

[3.0L]

Originally Posted by rich8566

Just because there is leeway it does not mean that there is any built in error. Sorry but I've confirmed in all of the vehicle I've mentioned that the speedos are dead on accurate.

... until the tires wear.

[Viffermike]

Originally Posted by rich8566

Just because there is leeway it does not mean that there is any built in error. Sorry but I've confirmed in all of the vehicle I've mentioned that the speedos are dead on accurate.

The 'leeway' is designed to start, at its low end, at dead-on accurate when a car is at speed. How is that not built-in bias? Did you even read the article I linked to? (Which, by the way, dates to 2002 and discusses then-relatively-new electronic speedos that use sensors, etc. that are processed by the ECU.)

Once again: You need to supply proof beyond your own personal accounts. You are not every driver, and you do not own every car. Congrats for having accurate speedos on your vehicles, but that only makes you a speedo-accuracy deity in your own world, not the real one.

(Also, another consideration: Underinflated tires will have a significantly lower rolling circumference, and most drivers aren't nearly as obsessive about keeping inflation pressures evened out as many of us on this board ... )

As a professional know-it-all, I can state that the first rule of being such is to realize that you don't know anything close to it all -- and to quit and acquiesce at that point, listen, and learn. Apologies, but you're digging yourself a deeper hole every time you post. Even when you double-post. Keep on shovelin'.

[Maynard]

Talked myself out of a couple of speeding tickets with this, back in the days when speedos were fragile and mechanical - pointing out to the officer that my ride had either a broken or mis-fit speedo, so I was just guessing. Once they were even kind enough to 'help me figure it out' by letting me drive by and radar'ing me. I probably should feel worse about misleading law officers, but I suspect that is giving myself too much credit, and in reality I should be glad they didn't want the extra paperwork and were giving me credit for a creative approach.

[bulletrs]

Buy a Vbox and try again, these apps are not accurate at all. Fastest i got on my M235i fully tuned 500+ HP 0-62mph (0-100kmh) was 4.04s. Checked with a VBOX sport. 100-200 in 6,97s.

Fully stock with a vbox my m235i did 5,1s. Could not get it faster.

There is so much BS around 0-60 etc.. times. That is my experience.

[Bmw240xi18]

Originally Posted by 3.0L

I think it's reasonable to assume that the speedometer gets its info from the computer. In any case, the roadside radar readouts consistently prove that my speedometer is off.

Next time out, I'll try to remember to do a timed run between mile markers. One mile traveled at 60 MPH indicated should take exactly 60 seconds. Probably set the cruise control to 60, first.

And for what it's worth, every car I've owned during the past 20+ years has had an optimistic speedometer, including my current MB GLK-350.

The roadside speed radars are also off.

[XutvJet]

Originally Posted by bulletrs

Buy a Vbox and try again, these apps are not accurate at all. Fastest i got on my M235i fully tuned 500+ HP 0-62mph (0-100kmh) was 4.04s. Checked with a VBOX sport. 100-200 in 6,97s.

Fully stock with a vbox my m235i did 5,1s. Could not get it faster.

There is so much BS around 0-60 etc.. times. That is my experience.

Mags have tested the M235x to run near 4 flat 0-60 and they all use VBox and those 0-60 run times are an average of around 10 runs. The M240X has about 20whp and 30wtq more than the M235. It's certainly is plausible that a stock M240x can get into the upper 3 second 0-60 range. X's advantage is it's abundant AWD traction and ability to build boost prior to launch.

Heck, even a stock M235/M240 6MT could do upper 3 second 0-60s if you strapped a set of slicks on the back to make the traction comparable to that of the X. Also, the RWD is a ~160lbs lighter which is huge. In stock form, the M235 6MT has routinely been tested to do 4.6 0-60s. I haven't seen data for the M240 6MT yet. I would suspect it's in the 4.4 range.

[samej]

Originally Posted by Viffermike

This is actually an unofficial mandate by the Feds that goes back to the 55 MPH days in the 1970s. Car manufacturers calibrate speedometers to read slow for two primary reasons:
1). To account for the loss in rolling diameter as a tire's tread wears (lower rolling diameter = faster RPM = higher speedometer reading)
2). To prevent lawsuits and such against car manufacturers, municipalities, states, etc. for false speed readings that lead to unlawful behavior, i.e. speeding tickets

Most U.S. law enforcement agencies are instructed to NOT pull over vehicles that are traveling less than 10 percent over the posted speed limit. Why? A combination of three reasons:
1). User error/improper calibration of speed-reading apparati (timers, laser/radar, etc.)
2). The aforementioned speedometer calibration tendencies
3). The relatively low income that comes from most low-level speeding tickets (i.e., below 10 MPH, not in a construction zone, etc.), and the likelihood that income will be mitigated by defendant challenge in court. In other words, it's simply not cost effective.

Most modern cars' gauges are computer controlled, so what the OBDII says (and what the digital readout would say if it's coded) is almost certainly going to mirror what the analog gauge says. That's definitely the case with the speedometer on my P3Cars gauge, which is an OBDII device.

And now y'all know.

There was an article in Car and Driver years back. Germany mandates that their speedos MUST read at the indicated speed or lower. Combined with the various wheel and tire options from the Germans that are common they intentionally read high. VW was the most notorious for reading high (some 5mph and higher IIRC), but all the German manufacturers read high. On the other hand the odometers are very accurate so the correct speed is known to set that.

[aerobod]

Here is the EU specification of how the speedometer needs to be calibrated and tested: EU speedometer specifications. Section 5.3 provides the formula for the displayed speed: 0 ≤ (V1 – V2) ≤ 0,1 V2 + 4 km/h, where V1 is displayed speed and V2 is true speed.

[Lew Carrum]

Originally Posted by martopoulos

App-based 0-60 times are widely inaccurate-- the GPS hardware is just not meant for something like this. Someone on this forum also claimed their m235i with a few mods was getting 3.6 seconds when measured with an app. Keep dreaming. If you want to know your time, go to a track or get a real accelerometer. Here's one that has been consistently accurate (i.e. before mods it consistently registered 4.6 seconds, and after Dinan mods is consistently showing 4.3 - 4.4):

K.A.T. Matrix 3-Axis Car Performance Accelerometer - ~$110 on Amazon, though out of stock at the moment

The video is the proof of this though, and if you time it with a stopwatch you will record a reading of about 4.3 seconds.