Speedo Error - Garmin proves it

The error is linear, not that surprising from a calibration standpoint.
Image Unavailable, Please Login

 

This is pretty typical of 80's vintage Ferraris. Even the newer ones have generous speedos. My 430 is generous by about 2-3 mph though it stays constant at least at speeds under 100.

Dave

 

Do not trust GPS speeds as absolute. I learned the hard way that my GPS was lower than actual, despite the 'conventional wisdom' that says GPS speedos (esp at highway speeds) are more accurate than a speedometer. I have further supported this (and w.two different Garmins btw, of different vintage) by running at two different automated radar speed limit signs.

Or if you do trust it, good luck.

 

Is it possible to recalibrate these things?

 

Yes, to a degree. I had Palo Alto speedometer recalibrate the one from my 328 when I restored it. It made it MUCH more accurate, though there is still a variance. I think it's just the nature of these gauges; they're not that good.

 

This seems completely backwards to me.... if I understand right he is saying his speedo reading is low. The differences seem about right to me except backwards to my car. I show about 10 mph (16 Kph) slow at 70 mph in my 328 and this seems consistent with what others say.

 

Something must have been wrong with ducowti's GPS units. GPS velocities are typically computed using the Doppler shift on the observables ... they are VERY accurate. In fact, GPS-derived velocities (even from relatively low-cost chipsets such as those found in Garmin and TomTom units etc.) are typically accurate to about 0.2m/s (i.e., about 0.5 mph).

308-newbie is right when he says that RTK GPS positions are far more accurate than stand-alone positions, however, as noted above, stand-alone velocities should still be accurate to about 0.5 mph, even when using inexpensive consumer-grade GPS receivers.

One word of caution however is that GPS positions and velocities will always be at their most accurate when the GPS antenna has an unobstructed view of the sky and when plenty of satellites are visible. In other words, on the open highway your GPS-derived velocities will usually be spot on. However, in the middle of downtown New York City (i.e., a so called "urban canyon") the story is different.

 

I like to use roadside radar signs to check. Ferraris are a little generous, Toyotas spot on.

Dave

 

I have the opposite problem, when doing 60kph the nedle on the guage sits at 70kph, at 110kph the nedle sits on 125kph, I am running 17" rims with 40 series tyres, how do you recalibrate the speedo?

 

At least here in europe a 5-10% overrate of the speedovalue is a standard factory setting. I see it every few years when I have done the mandatory car test and the car is on the dyno run. I saw that on my fiats, alfas and lancias and even on my wifes toyota...

 

I had palo alto fix/calibrate mine.
Now the speedo reads low under 25mph and off above it. And it certainly has it's better days being more accurate some than others. As I approach 65, it seems to be off linearly by as much as 8mph and as much as 15mph at higher speeds.
I checked against a GPS & local speed sign.
I called Palo & they told me they calibrated it up to 100 & it was spot-on.

Hmmm, who do you believe? "Honest officer, the GPS said ..."

 

As stated earlier, as long as your GPS receiver is treated with a reasonable degree of respect (i.e., it has an unobstructed view of the sky and there are lots of satellites ... say 5 or more), the GPS-derived velocities will be absolutely spot on, and will be way more accurate than ANY speedometer in ANY car ... new or old.

 

I checked the speedo in my 328 this morning using GPS and a side-of-the-road radar indicator.

The speedo on my 1989 328 reads 4MPH FASTER than the car is traveling at any speed up to 100MPH. So there is no percentage variation at all in this case - it's just a straight 4MPH which indicates to me that the speedo is inherently quite accurate. Almost seems like somebody at Veglia stuck the needle on the speedo shaft slightly out of position...Surely that could never happen!

 

It's my understanding that cars sold in the U.S. (may be the same for other parts of the world) must have the speedometer read the exact actual speed or up to 10% fast (in which case you're actually going slower than indicated). If this is indeed the case, then manufacturer's are likely to make them read a bit fast (rather than right on) so that manufacturing tolerances, wear etc. can be accounted for. Reading low/slow might cause problems with the Feds., lawsuits for speeding tickets, insurance issues etc.

I tend to doubt that the "courtesy radar signs/speed indicators" along roads are super accurate since they're not used for tickets or legal proceedings. They're probably not far off but I would think radar guns and fixed units used for actual ticketing are more precise.

 

That must be some scene driving by the same road a few times to verify the speedo at different MPH. You actually brought it up to 100MPH? Hope it was a wide and empty road!

Ace

 

Not Palo Alto.

 

Q: They "they calibrated it up to 100"...Against what?
To do this with any degree of accuracy, they would need to know:
-The final drive ratio
-The number of actual pulses coming out of the speed sensor at a given speed
-The actual rolling diameter of the tires

Then again, having the car on a rolling dyno would be even more accurate...

GPS (consumer) accuracy, in a steady speed state, is better than 0.5 MPH. I suspect that Johnny Law is aware of this...as both his radar and GPS technologies use the basic principal of Doppler shift.

 

I just went by the radar once, to see how the radar and GPS compared. At an indicated 60 in the car, the radar showed 56 and the GPS showed 56.3. That was the only flyby of the radar site. The rest was done with the GPS. The lowest speed I checked was 30 - GPS said 26.4. Every 10 MPH from there up to and including 100MPH had the same error within .2 mph. Of course there could be some error in my own visual/foot coordination trying to hold the speed exactly at an indicated increment but basically, there was no accumulating instrument error as the speed went up/down.

Road was plenty wide and traffic sparse enough to hit 100 (or more) but I didn't stay there any longer than necessary since I didn't think that my explanation of speedometer-checking would be sufficient if I was pulled over (and I removed my Valentine 1 after my trip from Vegas to here last year so their would have been no alerts...)

 

I have much more trust in the GPS system then I do in the accuracy of the radar speed limit signs.

 

GPS is extremely accurate if, as noted previously, you have sufficient satellites in view (and you are using a WASS-enabled GPS), to get position accuracy down to a few feet. Normally, here on the Chesapeake and surrounding area, position accuracy on my boat's (WASS-enabled) GPS and the computer nav software (also WAAS enabled) is within 10 feet or less. I have seen it as low as 3.7 feet. If you are doing this without WASS-enabled GPS, then position error can be 50 feet or more and you can't tell what the error is anyway which would not really allow accurate speedometer checking. WASS enabled GPSs usually have a readout window available with the position accuracy (x feet) at all times.

If sufficient satellites are not in view or not yet locked in, I have seen speed readings on a stationery GPS unit of up to 497 knots!

 

My GPS managed to log a max speed over the speed of sound
Image Unavailable, Please Login

 

My 328's speedo is really very accurate - indeed much more accurate than the one on my Toyota!