|
Satnavs are accurate if travelling on a straight, level section of road. Anything else will build in errors.
|
|
And those errors will always be under-reads, won't they? Since anything but a level road will make the actual distance travelled greater than that perceived by the GPS.
|
|
And if you are going round bends the point to point distance may be shorter than the length of road travelled. I don't know how often the satnav takes a position check, so it may not be a problem.
|
|
|
|
Your sat nav IS accurate. The variations in height are not enough to cause any effect , sharp bends however introduce a slight "as the crow flies effect" but as your speed is being measured 1 a second you can work out for yourself how much that affects the reading - its not much,
------------------------------
TourVanMan TM < Ex RF >
|
|
It's small but measurable - about 0.5% under on a 10% gradient.
|
0.5%? at 70 mph thats 0.35 mph variance. I doubt its as high as 0.5% though
------------------------------
TourVanMan TM < Ex RF >
|
|
It is. Think of the gradient as a right-angled triangle. The satellite can measure the base but not the height, and the car actually travels along the hypotenuse. If the base is 10 and the height 1, the hypotenuse is sqrt 101, or 10.0499. Divide that into 10 to get the ratio of measured distance to true distance and you get 0.9950.
|
The satellite can measure the base but not the height,
My gps gives me the height above sea level, so I think they could factor in hills. But I doubt that any do!
|
|
|
which is fine is you had one satelite directly over head. With your fix from 8 satelites the line of sight from some of them is not seeing your right angled triangle the same way. From a side on view (which a proportion of the satelites will be) the gradient is considerable less.
------------------------------
TourVanMan TM < Ex RF >
|
|
The most recent GPS chipsets (including SiRF III) don't only use line of sight GPS signals. They use reflected signals to improve accuracy in weak signal areas, including under foliage and urban canyons. This is why they are typically able to process 20 GPS signals even though the most you'd see in line of sight is 8. And depending on where you live in the world things are improved with WAAS EGNOS. Latest units can get a fix indoors even - so much for line of site ;-)
|
|
It doesn't matter where the satellites are. Current GPS's work out speed by measuring the horizontal distance travelled in a given time. Horizontal distance given by lat/long points which do not take into account distance travelled up or down.
|
|
So a GPS calculates its own position on, and measures its progress across, an imaginary ideal, smooth sphere? No wonder they cope so badly with roads that end in lakes or docks! ;-)
|
|
|
|
I don't understand. Why can satellites only measure the horizontal distance? You seem to be suggesting that they only work in 2d? But a lot of receivers can give a height - presumably this is calculated form the satellites as well?
Does it not work in 3 dimensions on a vector principle? Therefore is the consideration of height gain irrelevant?
|
Satelites work out nothing. They are only very accurate clocks sending out time signals. The device does all the work from those signals
------------------------------
TourVanMan TM < Ex RF >
|
|
|
|
|
I have to say that there aren't many 1:10 hills that I worry about the speed accuracy of my GPS system on.
|
try this in a mondeo - set cruise at an indicated 77mph, which satnav will confirm is really 70mph. now reset the average speed measure on the computer, you will see that it returns 70mph! so the car, and hence, odometer, know the real speeds, it's just the dials are setup to over-read. So don't worry about higher mileages recorded as this is accurate.
interestingly the over-read doesn't seem to be linear, e.g it says 51 at true 50, 64 at true 60, 77 at true 70 and about 90 at true 80. so they've built in a step up in the dials, i guess to try and save your licence....
Rich
|
|
If I am right, no one has answered the OP's question about whether a misreading speedometer means that the odometer is inaccurate. I'm going to guess that the answer to that is "no" - but I'd like that confirmed by an expert.
|
re earlier. Yes, when calculating speed the normal GPS software works on a flat plane,not even a sphere. The routing software AFAIK likewise. The only software I've seen that takes height into account is 'memory map' that uses OS mapping and aerial photography for 3d mapping.
re speed/odo. When I've checked both bikes and the car, the speedos seem generally to be around 5% + a couple of mph over optimistic and the ODO's seem to be a lot more accurate. There is a legal requirement that the speed MUST NOT read low, so play safe and add a bit whereas there is no such requirement for the odo.
|
|
|
If I am right, no one has answered the OP's question about whether a misreading speedometer means that the odometer is inaccurate. I'm going to guess that the answer to that is "no" - but I'd like that confirmed by an expert.
Cheers Tyro, I think you're correct. I must say there's alot of very technical information about how GPS works which is welcome ----- I've certainly learnt something from the thread. I think I'll also believe that the odometer is accurate ---- unless someone thinks otherwise!
|
|
|
|
If you run the Mondeo in diagnostic mode and have the speed on the digital display it very closely matches sat nav speed.
And I can only describe the slow change in speed on the analogue display as inertia or latency becasue lift off and it slowly drops but the digital and sat nav displays update immediately.
|
If you run the Mondeo in diagnostic mode and have the speed on the digital display it very closely matches sat nav speed.
The same applies to my V70; switch the HU-803 head unit to the "hidden" diagnostics menu, and then choose the speed sensitive volume setting. This includes a speed indicator (albeit in kph) but which is very close to "Sat Nav speed". The speedo on the other hand overreads by 6mph at 70mph, this degree of inaccuracy then being maintained no matter how much faster you go despite being fed from the same source as the speed sensitive volume function. No doubt all very intentional and designed.
|
Sat nav is likely to be far more accurate than speedo, remember it relies on making calculations on billionths of a second time differences, if it couldn't get your speed right it wouldn't know where you are.
There's some formidable number crunching involved when a sat nav locks on to a signal when moving, astonishes me that it doesn't amaze more people.
|
I am amazed you are astonished that not more people are amazed.
------------------------------
TourVanMan TM < Ex RF >
|
|
I know, appalling, isn't it? I did read it afterwards but with no edit button, what can I do? ;-)
|
> The same applies to my V70; switch the HU-803 head unit to the "hidden" diagnostics menu...
How do you do that, then? I've never been convinced that mine takes any notice of speed in setting the volume.
|
> The same applies to my V70; switch the HU-803 head unit to the "hidden" diagnostics menu... How do you do that, then? I've never been convinced that mine takes any notice of speed in setting the volume.
Here's a post I made on the subject back in August 2003: www.honestjohn.co.uk/forum/post/index.htm?v=e&t=15...4
The hidden speed-sensitive volume option includes the ability to override the default speed related gain and either increase, or reduce, the effect. If I'm honest though, having played with it mine is now set back to the default gain at which it works perfectly; music volume appears the same regardless of road speed, which is how it should be. The cabin is a sufficiently quiet and relaxed place to be that any extra gain becomes immediately noticeable because it's too strong.
Have fun.
|
> Here's a post I made on the subject back in August 2003...
Thx. The link in your old post no longer works but a quick Google of "HU-803 hidden menu" turned it up first time. Looks like there's stuff in there that could have some seriously weird effects, so before I dabble I'm going to read it properly for the Reset All to Factory Settings option!
If nothing else, an accurate km/h readout will be useful on French motorways.
|
GPS positional information at the military standard of signal is accurate enough to deliver payloads to very accurate locations travelling at very high speed and at varying altitude - think cruise missile. I realise military GPS systems have more accurate (and encrypted feeds) but a cruise missile or GPS guides bomb travel pretty fast and need precise updates to get even near to where they are meant to go.
And again I know the wrong targets often hit... but the projectile no doubt hit accurately the target it was sent to but the wrong target selected. And lots of mis-hits also down to laser guided bombs too.
I remember seeing footage from the first Gulf War of cruise missiles flying down streets and turning at junctions. Cool? Maybe. But accurate yes.
For me sat nav has always been pretty accurate and from my example above it has to track speed, location and height very accurately or it would have been an issue for US military. After all they are the main customer for GPS and they own and maintain it.
|
rti70 - agreed, gps positioning can be very accurate. It's all down to the software in the receiver or navigation software.
Memory and processing power is limited in retail devices and full spherical trigonometry is very processor intensive.
With our simple requirements for in-car stuff, pretending that things are on a 2d flat world is normally sufficient to let the device spend most of its time drawing a fancy display, which is what most punters want.
|
|
It can't be that difficult to make a device that bounces a signal off the road and measures the true speed that way?
|
If I am right, no one has answered the OP's question about whether a misreading speedometer means that the odometer is inaccurate.
The odometer is directly driven via the speedometer cable, such as via a gearbox 'take off'. So the odometer is geared to the road wheels and will be accurate subject to tire wear and wheel slippage.
The speedometer however, is operated via a rotating magnet beneath a metal disk (or did in my day) that is attached to the pointer and the magnetically induced current in the disk reacts with the rotating magnet to produce a twisting force that pushes the disk round in the same direction as the magnet against a spring. In other words, the speed reading is less tightly coupled to the road wheel rotation. The reading is affected by the magnet's strength and its spacing from the disk.
|
|
Both electronic nowdays, would have thought that they would be similarly inaccurate, rather similarly accurate.
|
|
|
|
|
|
|
