Coming back from Ipswich the other day I couldn’t help noticing that the speedometer was always 4/5 mph slower than the Satnav. Given that the speedo is calibrated by mechanical influences and therefore open possible inaccuracies whereas the Satnav is being tracked by satelites and could potentailly be more accuarate. Unless there are some technical aspects of satnav I am not aware of?
sat nav should be more accurate, unless you are going up or down hill, then it will under read and you could be speeding… I have to admit I drive on satnav speedo in my roadster, as its easier to see then the dials which are behind my small steering wheel, or and my roadster speedo is also way out…
Neither. Only laser or radar is able to give a truly accurate speed. Or a good stopwatch and a couple of fixed points.
The speedo is just plain inaccurate. To account for that, they make them all such that they naturally over-read, so that even the most inaccurate example shouldn’t under-read for obvious reasons.
GPS is theoretically better. But it’s not in reality…it measures your distance covered in a given time period. It then assumes you took a perfectly straight (and horizontal) line between the two points, which will never be the case even on the straightest of roads.
If you think about it, a car manufacturer would be liable if they gave you a speedo that said you were going slower than you actually were and so resulted in your breaking the law. To cover themselves all car manufacturers ensure that their speedo units over estimate your speed, usually by around 4-5mph. Once a GPS has a lock on 3 or more satellites (and I believe most SatNav units use around 5) it can calculate your movement in 3D and so it doesn’t assume completely horizontal movement. Having said all of that, I imagine the likely error introduced into your speed calculation by an undulating and/or winding road will be pretty small given that the GPS updates are pretty frequent. Yes the figure you see is an average, but I believe it’s an average of a set of data points captured by the device, not the average of the speed between two points some distance apart.
The mk2 speedo is very consistent and fairly quick to respond. A sat-nav can take quite some time to adjust the speed it shows, but once it’s got there, it’s closer to the true speed.
However, you should take the sat-nav reading with a pinch of salt. I’ve seen mine show 2mph when I’ve been sat in the garage. It’s worth noting that the gps satelites are not quite stationary with respect to your point on the earth, so there’s some error there.
The most accurate speedo I’ve seen in a '5 was a digital speedo (run off the same feed as the built-in speedo), that was calibrated (to some margin of error) from a sat-nav. Comparing it to the sat-nav on a motorway shows how much the speedo fluctuates and how laggy the sat-navs are.
Don’t trust those speed warning signs as the side of the road. Most of the ones I’ve driven past will over-read, but some under-read. They do not, as a general rule, display the same speed for the same speed shown on the speedo. The ‘warning sharp bend, slow down’ signs also don’t seem to trigger on the same speed as the advisory ‘max speed x’ signs placed near them.
Here is what you do:
Use the satnav on a flat straight motorway to calibrate the accuracy of your own car speedo.
Interestingly our J2-Ltd which came on the wrong tyre size (nearly new Firestone 175/65/14 instead of 185/60/14) has a speedo which reads 99.8% accurate as a result…
Or if you’re very patient, put the car at a set speed and watch the ‘XXXXX 200 miles’ sign change to ‘XXXXX 100 miles’ while your passenger operates a stop-watch. 
Slightly different but anyone used trackmaster on a phone to time laps around a track. It uses GPS and i thnk its good for 100th of a second around a lap. Of course we never time our laps on track days
Why not use the driver location markers on the motorway ?
They are 100m (110yrd) apart measure those as you pass keep the speed as constant as you can and work out from there.
simple 
Or…go past a speed camera at a steady 80mph indicated, and you should receive the exact speed in the post within a few days.
I think, all in all, that’s the easiest plan.
Hi
Speedos can be pretty inaccurate. Legal limit is reading high by 10% +8kmh / 5mph for production vehicles I think. And the accuracy can vary. My NC seems fairly accurate, but it will vary with tyre wear, etc. Few years ago we had an Alfa 155 and the speedo on that read about 5mph high at most speeds. 35 was a true 30mph while 86 was a true 80mph. Although occasionally on a long journey on a hot day it would decide to read accurately.
Sat nav should be accurate. However depends on the unit. A cheap basic one will probably just use speed / distance. Most will measure speed by checking the doplar effect on the gps signal, and you can get GPS units that update at 10Hz (check a Driftbox). Accuracy of these, assuming a half decent view of the sky, should be far better than laser / radar.
Some GPS units have a max speed readabout, but wouldn’t trust these. Had one tell me I had hit 60mph, except I was on a 50cc bike. Looking at the log it had been doing 50mph a second before and was doing 50mph the second afterwards.
All the best
Keith
Speedos fitted to production cars are calibrated to indicate about 5mph faster than actual road speed. They have to do that to get through European type approval regs. I’m building a kit car at the moment and the accepted wisdom is that kit car builders should do the same to pass the dreaded Single Vehicle Approval test. Sat Navs will generally give a more accurate reading although as mentioned in previous posts, they’re not to be trusted 100%.
Stu
I’m not sure that it strictly true. My undertsanding is that they’re allowed to be out by a small amount, but that’s not always 5mph. Anacdotal evidence suggests that some manufatcurers are applying tighter tolerances. I’ve driven cars that match a satnav or measured speed with much closer speeds than 5mph. One read 71-72mph at 70. Also, the car’s speedo isn’t ussually calibrated, it’s tested to see if it passes the standard. Calibration would imply that there’s an adjuster somewhere, which (AFAIA) isn’t present on car I’ve owned.
There’s a choice of 2 non-UK standards for cars to comply with. One of them is: http://www.unece.org/fileadmin/DAM/trans/main/wp29/wp29regs/r039r1e.pdf
That states that the speedo must read greater than or equal to the true speed, and less than or equal to 10% of the true speed + 4kph. That’s 8kph @ 40kph & 16kph @120kph. Approx: 5mph @ 25mph & 10mph @ 75mph
If the speedo is reading 30 at 25mph would make it fail the test (as 8kph is approx 4.97mph). If it read 5mph at 0 it would, however, appear to pass.
Annex 3 of those regulations give more lax requirements for production vehicles.
All the best
Keith
Yes it does. The way I read it cars can differ by 2kph, but the Type Approval still needs to be done at the more accurate tolerances. So the cars can’t be designed to exhibit the wider tolerance. It looks like it allows a tolerance on the production run of 2kph. E.g to allow for minor variation in the placement of the speedo dial.