What is correct, Speedo or sat nav?

epninety

Strider590 wrote: »

http://gpsinformation.net/gpsclouds.htm

True, I wasn't quite right in my statement. The loss in the direct path (satellite directly overhead) is very low, but increasing for paths nearing the horizon, due to a greater portion of the path being inside the atmosphere.

GPS accuracy is degraded in foggy weather, probably more due to increased multipath distortion than absolute loss. This reduces the available satellite constellation for accurate solutions. In other words signal flutter and fading become more disruptive, even though the absolute signal level is not greatly reduced.

Probably more relevant for extreme latitudes, and those of us who remember 8 and 12 channel (or even less!) receivers, than the latest chipsets with massive processing gain in their signal budgets.

Imp

epninety wrote: »

It does not receive 'personalised' data, however it does receive identity and time information from each satellite, updated 'almanac' data (describing the latest orbit corrections for each available satellite), and health information for each satellite. Since it knows something about the positon of each satellite, it solves the system of equations describing the signal delay from each satellite to arrive at a solution for both position and current time.

And, interestingly enough, if there are at least 7 satellites available to the GPS receiver, it can directly calculate its speed from the received signals, very accurately. And if you go with a reasonably modern receiver operating a 5Hz, its lag during acceleration/braking is probably no worse than the speedometer fitted to your car.

mikey72

Have we done the 2D/3D speed calculations yet?

birkee

adouglasmhor wrote: »

An electronic or the earlier eddy current speedometer is not "linked" to the vehicle though is it? You just made that up.

And I have allready mentioned tunnels twice now.

It's like trying to knit fog isn't it?

An eddy current speedometer IS directly connected to the vehicle. The fact that the indicator works from magnetic induction is neither here nor there. The cable is rotated according to the vehicles speed.
The electronic one also works from sensors that rotate according to the vehicles speed. The fact that there is no mechanical connection doesn't change the fact that the vehicles actual speed controls the information fed to the instrument.

I know tunnels have been mentioned, I mentioned them myself, as being the reason why sat navs wouldn't meet requirements, because they can't indicate a vehicles speed at any given time, i.e. in difficult reception conditions, or upon satellite failure.
(Suns magnetic storms wiping them all out? Meteor shower? Enemy action?.... What do you all do? Park your vehicles up till they get new satellites?)

Flyboy152

jase1 wrote: »

Are speedometers, especially digital ones all that much different? Just because the sample rate is higher doesn't make the measurement any more of less accurate.

As has been said, if you are driving along at a constant, indicated 70mph on a flat, straight road there is no reason to believe that the (unchanging) speed as read by the satnav is anything other than accurate.

I'd also say that on my car, the average speed measure consistently shows a speed slightly lower than the main speedo, and the same as the satnav, given the above.

But the status of the road is not the only thing that affects the readings from your satellite navigation. The set itself, the quality of signal, the hysteresis* effect from the time lag from the satellite to the set, wind speed and direction; even the weather, both terrestrial and cosmic, plays its part.

*Not sure if that is the correct use of the concept, but I am sure you will get what I mean.

mikey72

Flyboy152 wrote: »

But the status of the road is not the only thing that affects the readings from your satellite navigation. The set itself, the quality of signal, the hysteresis* effect from the time lag from the satellite to the set, wind speed and direction; even the weather, both terrestrial and cosmic, plays its part.

*Not sure if that is the correct use of the concept, but I am sure you will get what I mean.

The quality of the set should be the only thing that affects the readings, most satellites give a better reading. But even then, if it's a constant speed, even a low number with be accurate, it'll just take longer to get there.

Flyboy152

mikey72 wrote: »

The quality of the set should be the only thing that affects the readings, most satellites give a better reading. But even then, if it's a constant speed, even a low number with be accurate, it'll just take longer to get there.

But then you have to work out what the indicated speed was at the time the reading relates to. The time lag does not allow for real-time speed. Ergo, the indicated speed is inaccurate in relation to the speed the vehicle is currently travelling at.

jase1

Correct. But you could say the same for a digital speedometer as well -- these often only update every second or so and I have seen evidence on a Citroen I was a passenger in that there was a time lag of a couple of seconds there too.

Any digital readout will be this way -- you take a sample, calculate the effective rate, and display it. Even an "analogue" speedo is not directly connected to the gearbox is it?

Lum

jase1 wrote: »

Any digital readout will be this way -- you take a sample, calculate the effective rate, and display it. Even an "analogue" speedo is not directly connected to the gearbox is it?

Depends how old the car is. Mine is directly connected to the gearbox via a rotating cable, and the speedo unit also serves to translate this into a series of electronic pulses that are then sent on to the ECU.

jase1

Yes, I believe this was the case with my Primera as well (when the cheap clutch I bought exploded it gave me a chance to look inside the gearbox bell-housing). But even so, there is a point at which the rate is digitally sampled.