speedometer

So, just a small thing. My 2007 Taco's got about 3200 miles. The speedometer has always been off by about 4 or 5 miles an hour. It reads high. Thus, when it reads 65, i'm actually doing 60... and so on. I guess I'm just happy it doesn't read low.

I talked to a guy at the dealership and he said that Toyota allows for a 10% variance. That sounds high. That makes it sound like if I was doing 100mph the gauge could either read 90mph or 110mph...

Anybody have any experience with this. Comments?

 

All manufacturers end up with some cushion on the speedo. Some are more exagerated than others. It's a bid deal that the odometers not reading LESS than you have actually travelled (more doesn't matter so much) so the manufacturers add a little extra to the calculation. This calc is used in the speed calc (distance/time) so you get a speedo variance.

That's how it was explained to me by a much wiser person than I.

 

One cool, easy way to correct that would be to get tires that are 10% larger than stock (unless you are 2wd and lowering your vehicle). Then your speedo will match your speed.

 

thats very true, it will offset the speedo enough to where it will balance it back to right where u should be as long as the tires are too much bigger, but you have to add the 10% to the size by the circumference, not the tire height or width

 

Mines about 2 miles faster than what my GPS says.

 

Mine is "dead on balls accurate" according to my GPS.

 

What other speed measurement are you guys using to determine how off the speedometer is?

 

You could repeatedly drive past one of those police radar trailer things that says "your speed is" and then flashes up a number, or have someone pace you in a vehicle whos speedometer is known to be accurate, or personally I am using the speed function on my GPS unit, though I do not suggest that you go out and buy a GPS just to check the calibration of your speedometer, but they are a wonderful tool.

 

Yeah, I'm not going to buy a GPS for this (plus I'd wonder how accurate it can really measure speed). I'll watch for one of those automated speed signs. They sometimes put one in the school zone near my house.

 

The "your speed is" thing, and the GPS and the truck all agree. (I've checked). Not that it is at all related, buy to answer your wondering, it depends on how many satelites the GPS is receiving siginals from at a time. The best I have ever seen it get was one day when it was receiving the signal from 12 satelites and/or ground stations and it was accurate to +or- 4 feet. Normally it is about +or- 12-15 feet. I don't know how that translates as far a margin of error for speed, but it should be pretty low.

 

See, that's the thing. You're not giving me any more confidence in how accurate GPS can measure speed. I thought they could get down to 1 foot accuracy routinely. I wonder if you got lucky that your GPS agreed with the other two sources. Let me throw a bit of math out (nd, don't worry, just take a couple of aspirin and the headache will subside).

30 mph = 30*(5280 feet/mile) / (3600 seconds/minute) = 44 feet/sec

To measure speed, GPS has to record at least two locations and accurately time the interval (their clocks are extremely accurate, so no worries about that part).

Worst case (with 12 satellites) is -4 feet at the beginning and +4 feet at the end. That gives a 8 foot variance. Divided by 44 means 18 % uncertainty. It gets better at higher speeds though. 60 mph would be 88 feet/sec. So the 8 foot variance would be 9% uncertainty.

On the other hand, I'm sure GPS speed is not done by a single measurement. They can improve their accuracy by having larger statistics. I'm just guessing here, but supposing they sampled 4 times per second, over 5 seconds (you'd have to be sure to maintain a consistent speed) they'd have 21 pairs of numbers. They can improve accuracy by averaging those and their uncertainties are divided by the sqrt(# of measurements). sqrt(21) = 4.6.

So, maybe it's not so bad as I thought. That 30mph +/- 9% becomes 30 mph +/1 4% and the 60 mph becomes 60 mph +/- 2%

Oh, and for reference, on a normal day (+/- 12-15 feet, let's pick 15 for worst case) you'd have a variance of 30 feet, which at 30 mph (and 21 measurements) would be +/- 14.8%, and at 60 +/- 7.4 %.

I'll trust the police positioned radar gun signs more than GPS.

PS. nd, have those aspirin kicked in yet?

 

I do physics equations for fun in my head. That said. ND can I have some of those asprin.

I am going to completely agree with you here Sawdust, not because I have checked your calculations, but simply because I am currently drunk (as per the alliance by-laws) and I can't begin to even process what you just said or check your math right now.

In all seriousness, though I have checked the GPS against the radar gun sign in both of my Tacoma's, one F-150, two Honda Accords, and two of my work trucks (I am a trucker), and in all of them the GPS read-out has agreed with the radar sign read-out, though the 2007 Tacoma is the only one where the speedometer has also agreed with the other two devices. So I suspect that there is some other factor in the speed calculation of the GPS that makes it more accurate, but I can neither prove it or tell you what the equation for it is.

 

dont take the readings off of those speed signs too seriously, around here they always read about 5mph under what my speedo says...maybe its the speedo, but it's been like that with every car i've ever owned. i went past one on my bike doing about 120 in a 40mph zone haha

 

yeah i'm having the same problem.. my speedo is of by 5 mph and my odo is .5/15miles fast... i'm fighting them right now about it.

 

uuuhhhh.... what?

 

Look, i have no idea what Sawdust just said but i saw alot of numbers in it and it confused me. On a side note, i'm proud of you for following the rules and being to drunk to do math. Kudos! Following the rules is cool

 

i have no idea what that texan said mine is 5mph faster than my spedo says but that is from the bigger tires

 

No kidding? I learned this stuff from getting my degrees in physics. I'm the alliance's worst nightmare: a Texan with an education

I could be seriously underestimating how often they sample position. If they did it 4 times as often as I guessed, that would reduce the uncertainties by half. Electronics should be capable of even doing millesecond readings, but I don't know if that's cost effective. That would amount to 1000 readings per second and reduce the uncertainties to about one thirtieth of my estimates.

 

A texan with an education is not scary at all if that education came from Texas. Now if you went out of state i might be worried....

 

So it's been bugging me, so I dug out the box for the GPS and looked up the specs.

it's sample rate is twice per second.

It's listed margin for error in determining velocity is +/- .05 meters/second at a steady rate of speed. Since they put that last part in there I assume it's error is greater under acceleration or deceleration.