Inside the Volkswagen emissions cheating

Posted Jan 7, 2016 14:51 UTC (Thu) by paulj (subscriber, #341)
In reply to: Inside the Volkswagen emissions cheating by alonz
Parent article: Inside the Volkswagen emissions cheating

EU rules are that the speedometer must _always_ read higher than true speed, I think, possibly by at least some small percentage (I don't quite remember). In part to ensure motorists can not argue their speedometer didn't tell them they were speeding. So it's probably related to that, and so I don't think you can, least not in the EU.

Is my vague memory and guess. ;)

Inside the Volkswagen emissions cheating

Posted Jan 7, 2016 15:35 UTC (Thu) by nybble41 (subscriber, #55106) [Link] (2 responses)

I don't know about the EU, but my speedometer here in the US seems accurate enough when compared to a GPS, at least at highway speeds. In any case, if it's commonly known that speedometers are required to read 5-10% higher than true speed, then motorists *could* well argue that they took this into account when interpreting the speedometer results, since the law is that your true speed can't be over the speed limit, not that your speedometer can't read over the speed limit. For example, if the law is that speedometers must read at least 5% over true speed, and yours said that you were 3% over the limit, you could argue that—knowing how speedometers must be calibrated by law—you had a reasonable expectation that your true speed was at least 2% under the limit, and that if that was not the case the issue must be with the speedometer and not the driver.

Inside the Volkswagen emissions cheating

Posted Jan 8, 2016 13:58 UTC (Fri) by NAR (subscriber, #1313) [Link]

Actually some drivers are more "clever", they also take into account the tolerance of the police - i.e. they start to fine only if the driver went substantially faster. For example in Hungary the some (most?) of the police only fines the driver if he went faster than 69 km/h on a road where the maximum allowed speed is 50 km/h. So a few drivers go at 68 km/h (according to their GPSs).

Interestingly there are also minimum speed limits, for example 100 km/h on motorway in left lane on some steep parts - there the driver might say that he was over the minimum limit according to the speedometer, but in fact was slower. I don't think anyone ever fined for this.

Inside the Volkswagen emissions cheating

Posted Jan 8, 2016 20:12 UTC (Fri) by anselm (subscriber, #2796) [Link]

IIRC, here in Germany the speedometer in a car may not display a speed that is less than the speed the car is actually travelling at, while in the other direction the maximum allowed tolerance is something like 10% of the highest speed printed on the speedometer scale. So if the highest speed on the scale is 200 km/h, the speedometer may display a speed that is up to 20 km/h faster than the actual speed. This leads to even small cars with wimpy engines having speedometers that go up to fairly impressive speeds which are way beyond what the car could actually reach – even, as we say “at full throttle, downhill with the wind at its back, and homesick”.

If you're cited for speeding, there is an automatic allowance for “manufacturing tolerances” that would let you get away with somewhat more than 50 km/h in a 50-km/h zone. Even so, you don't get to claim that your speedometer showed less than the actual speed of the car. Manufacturers calibrate the speedometers to display a higher speed on purpose, so people that drive such that the speedometer value matches the posted speed limit are actually safely below the speed limit.

Inside the Volkswagen emissions cheating

Posted Jan 7, 2016 17:02 UTC (Thu) by pjones (subscriber, #31722) [Link] (2 responses)

Starting with FMVSS 127 in 1983, new passenger cars for retail sale in the USA must be configured such that at 50 miles per hour: a) the speedometer cannot be wrong by more than 5%, and also b) the indicated speed must be higher than the actual measured speed. 127 has sense been supplanted by other regulations, but AIUI the rule is still the same.

If you look on VW and Mini web forums (merely the two vendors I've had cause to look at in the past), the first thing people notice when they hook phones and such up to OBD2 to display real time data is that GPS and OBD2 display the same speed more or less, but the speedometer readout is usually ~2.5mph faster when moving at speed.

I don't know if OBD2 is required to show the real measured speed or not, but on those vehicles it does.

Indeed

Posted Jan 8, 2016 13:03 UTC (Fri) by mchouque (subscriber, #62087) [Link]

On my car, the speedometer reports a value that is always higher by 4 km/h than what I read on the OBD bus.

Inside the Volkswagen emissions cheating

Posted Jan 9, 2016 18:45 UTC (Sat) by jdulaney (subscriber, #83672) [Link]

My speedo in my 1988 Ranger showed about 3 MPH over what my actual speed was, as measured by a combination of radar and timed distance. I corrected this, of course.

Inside the Volkswagen emissions cheating

Posted Jan 7, 2016 19:12 UTC (Thu) by jhhaller (guest, #56103) [Link] (1 responses)

Speedometers were typically driven by wheel rotation, so worn tires and under-inflated tires will show a higher speed than the same car with new tires which are overinflated or normal. Given that the air pressure rises as the tire gets warm, you will get different values for the same tire pressure and wear in the same trip. Presumably, the ECU could supply some correction for the pressure/temperature, particularly if equipped with a Tire Pressure Monitor as newer cars are. I'm not sure if any cars drive the speedometer by the GPS in cars with GPS (including those used just for showing a direction), but those could be somewhat more accurate, excluding the GPS wander factor which overestimates instantaneous distance traveled.

Inside the Volkswagen emissions cheating

Posted Jan 7, 2016 19:34 UTC (Thu) by pjones (subscriber, #31722) [Link]

Typically in the post-OBD2 world there's a Hall Sensor or an optical sensor in the wheel hub and it's counting axle rotation, and then the computer is using some method (who the hell knows if it's just multiplication or something more complex...) to arrive at a number. And then it's displaying a different number.

There certainly *are* GPS speedometers - they're common for kit cars, and google finds plenty of them for sale. It's worth noting those cars don't have to comply with the FCR/NHTSA/FMVSS requirements for a new retail consumer vehicle; they still have to have a speedometer to be street legal to get a VIN assigned, but the rule that says it must reflect a higher-than-measured value at 50MPH, AFAICS, does not apply. Emissions standards are also applied differently in many cases.

Inside the Volkswagen emissions cheating

Posted Jan 7, 2016 20:47 UTC (Thu) by mstone_ (subscriber, #66309) [Link]

IIRC, the permitted margin of error (permitted as in "massive fine for noncompliance") is +10%-.001%. If you were an engineer, how would you implement this given hardware with a margin of error something like +-2.5%? Most would implement it as "set it to read high" because that's a heck of a lot easier than re-designing the hardware to actually have a .001% margin of error.