Slightly OT? Speedo accuracy vs GPS

There's one on the A488 at Pontesbury. Only thing is that it overreads by about 10%..shown me as doing 33 when the speedo says 30MPH in three different lorries.

We're looking at how the device calculates your speed. Doppler shift in the signal from the satellite is taken into account as part of the calculation used to time how long the signal took to reach your device, and therefore how far from the satellite you are.

3 satellites and you have a postion. 2 positions, and you have the speed. Positions is all the devices software is using to calculate it.

OK, Have it your way.

Plenty more of those if you like.

GPS uses the predicted Doppler shift to adjust it's receiver frequency (by up to 2400Hz) to lock on more quickly to all the satellites, once it has picked up the first one.

It also feeds it's known position estimates and the doppler shift readings it obtains from the PLL decoders on the 50Hz carrier signal of each channel into a Kalman filter to get a more accurate speed than an instantaneous (pure Doppler) or long term (just position based) speed reading.

But don't just take my word for it. The info is all out there.

Sorry Paul, but I don't think any of the links you provide actually state whether velocity is calculated directly from doppler shift in common or garden GPS units or not, and you fail to quote a source for your statement.

Your post makes me feel like I'm being sold an ecotek valve, a Surbo, some magnetic insoles to reduce my risk of coronary heart disease and some zen odour reducing fuel economy increasing signal booster necklaces for my mobile phone.

(c:

Well the one I left in talks about Garmin basic and more advanced units. If you like I could find out which single chip GPS is used in TomTom Go and the like, is that common or garden enough?

But I can assure you nothing on the market today uses simplistic 'take two positions and average them' algorithms. Mostly it's a single chip jobbie that takes in an antenna feed and power and outputs NMEA. What goes on inside isn't that widely publicised.

I put the technobabble in to summarise what actually happens, in real life, in pretty much all GPS units on the market. If you want to know whether it uses Doppler shift or not you have to expect some technical speak in return. (Answer - yes it does, two different ways.)

BTW, what statement do you mean? If you mean the Kalman filters and stuff, I didn't find it on the web somewhere. I *know* about GPS. I wrote what I wrote off the top of my head.

I'd be happy to go into all the reasons Ecotek and all the other dodgy friends don't work in the same sort of detail someday - but I'm off to bed.

Night all.

It is illegal for a spedometer to read under at all. But it is also impossible for a speedometer to deliver an absolutely accurate reading, there will always be a +/- X% margin of error. To reconcile these two, speedos are therefore deliberately designed to read *over* by X%. X will vary depending on speedo type and technology but 10% is regarded as par for the course on olde worlde mechanical speedos. So you will always be travelling more slowly than your speedo tells you. In a straight line with a good view of the sky and at least 4 satellites acquired, GPS will always be more accurate (unless the Yanks decide to turn selective availability back on - unlikely).

There is an advisory 40mph on the A1 in Gateshead. Going North there is no trouble taking it at 80mph+

The most accurate I have had was 0.5km/h out at 56MPH and that was because the tyres were part worn. Would have been spot on otherwise. Not bad for a 4 year old 300,000km vehicle.

Bigger vehicles seem to throw them; I was following a bus that went into the town at 35mph on my speedo (which would be about 32 on my GPS - the RX8 seems to be +3mph on the digital speedo), the sign said it was doing 45!

Richard

And going south, most of the time you'll be lucky if you get through the bugger at 8mph.

I want a Newcastle bypass.

Richard

True!

You've already got one, and it's crap!

I've checked perhaps 15 speedos over the years using a steady speed and a stopwatch against the motorway blue/red 1/10km markers. The worst was an Astra which read 80mph at a real 70mph, the best was a 1.5 Nissan Sunny which read 71mph at 70. Most cars seem to read 2-5 mph fast but a few will be 10% out.