Waas

November 22, 2004

Mr HH I have known Mr Snaik for several years from afar, he is an incorrigible rogue, alas we have never met but he is well known to many people in Perth and Kinross and the Kingdom of Fife.

"NEMO ME IMPUNE LACESSIT"

LL

1.4k · November 22, 2004

from 17/11/2004 00h42 UTC to 17/11/2004 00h54 UTC : no broadcasting, then message without differential corrections.

Just a little glitch, there

12 minutes without EGNOS/WAAS at 1 o'clock in the morning.

Yes. I agree that it's a very little glitch.

What was the effect upon your GPSR's fix?

Were you inconvenienced greatly by the very little glitch, WLW?

If so, what were you doing at the time?

November 22, 2004

How many geocachers were honestly inconvenienced by not having ionospheric corrections between midnight and 04:16 last Saturday morning??

The rest of us who know how to use WAAS were not.

...and with a swish and a dull thud, "The Forester" misses the point again...

1.4k · November 22, 2004

With a single processor there is no such animal as parallel processing....only psuedo parallel processing....of course commercial outlets don't like to highlight this fact of life....to achieve true parallel processing one requires multi-processors!
A single processor can only work on one piece of information at a time....it may give the appearence of multi-tasking....but that is all...just the appearance.

Ullium,

I didn't say that there are twelve parallel processors. I said that a modern GPSr can listen to twelve (or ten or whatever) NavStar sats at once, ie in parallel fashion.

It's a bit like me listening to you while also noting the heckling at the back of the class from beavis and butthead while they sneer at science. I can do both at once, but I don't need multiple processors to do so or to differentiate between a genuine interest in science and their low-grade urge to sneer at science.

A modern GPS receiver can receive all of that data at once, unlike the old-fashioned 5-channel single series ones which could only track the nav message from one sat at a time.

It doesn't take multiple processors to interpret the multiple data which is being transmitted simultaneously with 12 (or more) sats simultaneously on the same frequency at the same time.

For an analogy, think about a fast Fourier transform.

Cheers, The Forester

November 22, 2004

:lol: :lol: :lol:

Yes Forester I know you didn't say there were twelve parallel processors...but you implied that a single processor can accomplish multi-tasking....when in fact it only appears to under low weight traffic :lol:

Also even you might have your concentration pushed to the limit to listen to more than a couple of sound sources and do them all justice :lol:

There comes a point where a limit is reached....note that one can't recieve EGNOS signals in battery save mode purely because of the overhead in heavy processing power required to compute the corrections received....and even when it is not actually transmitting there must be some overhead on the GPSr to keep testing for it ... wouldn't you say? :lol:

Even using all the advantages of vector mathematics most of the benefit lies in error reduction and ease of programming :lol:

And...

A modern GPS receiver can receive all of that data at once,

again...no single processor can receive multiple date at once....it just appears to!!

Ullium.

1.4k · November 22, 2004

Being a newbie to this and not wanting to read the full thread (read lazy!), can anybody explain in laymans terms what WAAS and EGNOS are?

A raw GPS fix is inevitably subject to several errors which degrade its accuracy.

The satellites are very accurately tracked from five tracking stations around the world and their orbit is updated at least twice a day (usually more often than that), however the transmitted position of the satellites is not entirely perfect. Inevitably there are little wobbles on the orbit and slight deviations from calculated track. These introduce small errors of a metre or two.

They all have extremely accurate atomic clocks, but these clocks are not 100% accurate either. There's another couple of metres error to be corrected for there.

Then there is the fact that the transmitted signal is slowed down when it enters the ionosphere. That rate of slowdown is not constant. It changes with time of day and other factors such as solar radiation that’ll give you yet another error of a handful of metres.

Then there is an even more erratic source of error introduced when the signal enters the troposphere (the bit where the atmosphere is mostly located). That changes considerably with varying weather conditions.

All these factors together give a variable error of something like 10 - 15 metres. If you plonk a GPSr on a fixed point and plot the co-ordinates which it indicates onto a screen, you end up with a scatter plot that looks like a shotgun hit.

Now imagine if you put a GPSr onto an accurately surveyed point such as an Ordnance Survey triangulation pillar and log all the range data from each visible satellite, you can calculate what the indicated pseudorange *ought* to be and you can compare that range with what the GPSr is actually indicating. The difference is a measurement of the error for that satellite at that time and at that place.

Now, if you were able to transmit that information about the current range error for each of those satellites to a second mobile GPSR nearby, say within a hundred kilometres or so, then that mobile GPSr's received ranges could be corrected for the measured errors and those errors effectively eliminated. That's what a local area augmentation system (LAAS) or a differential GPS (DGPS) system does.

There are hundreds of these systems for various purposes, but they are quite expensive to set up and they have limited range. For example, there is a DGPS service which gathers error correction data from several lighthouses around the country. The error correction data is transmitted to subscribed customers via Classic FM!

About 10 years ago the US Federal Aviation Administration decided to invest a fortune in setting up a wide area augmentation system to cover the whole of North America. It comprises a network of reference stations which are constantly logging error data for each of the GPS satellites. That data is sent on realtime to a central processing facility which interpolates between the referenced stations to form a mathematical model for thousands of blocks of territory. Those processed error corrections are uploaded to several maritime relay satellites which are mostly used for ship to shore and intership telecomms. They then transmit the WAAS correction data to the users’s GPSrs in a format which looks to the receiver a bit like a GPS data, but is not coming from an actual GPS satellite.

EGNOS is simply the European version of WAAS. It works the same way for your GPSr. It is still in a test phase and is not yet certified for safety-of-life missions, but it is there for most of the time on most days and nowadays is much more accurate than it was when the testing first started. It reduces a typical GPS fix error from something like 10-15 metres to something like 2-3 metres.

That’s good.

Geocaching was launched in celebration of the reduction in error following the abolition of the aboninable SA.

WAAS gives us a comparable improvement in accuracy of GPS. I think that's worth celebrating and enjoying while we have it.

1.4k · November 22, 2004

you implied that a single processor can accomplish multi-tasking

I didn't intend such a meaning.

A parallel receiver of radio traffic does not imply a multi-tasking processor at all.

Modern GPSRs can receive in parallel all the receivable GPS nav messages without needing parallel processors.

They do so all the time.

November 22, 2004

Modern GPSRs can receive in parallel all the receivable GPS nav messages without needing parallel processors.

Mythical....is not spelt MODERN...Forester :lol: :lol: :lol:

What it does is take each task in turn...allocate it some processor time...then move on to the next....OK it does this very very fast....but as I pointed out ...there comes a time when it starts to slow the processor down....as when you get too many of those little background programs running on your PC all demanding some portion of the processors time allocation :lol:

But all this is by the way....the question you keep avoiding (and very expertly let me say :lol: ) is ...am I correct in assuming that even if I do mannage to get the EGNOS signal ...can I be assured that it will be worth a curdy ALL of the time it is transmitting....because this has not been my experience in the recent past :lol:

And if not wouldn't it be more sensible to ignore it until it is up and running properly? :lol:

Ullium.

November 22, 2004

The only way you can judge the actual accuracy of your GPSr is by plonking it on a known location, such as a trigpoint, and noting the indicated co-ordinates.

Cheers, The Forester

Hmmm,

I tried that and the trig point was 15 feet to the left of where it should have been.

QED

November 22, 2004

:lol: :lol: :lol: :lol: :lol: :lol:

Ullium.

1.4k · November 23, 2004

the trig point was 15 feet to the left of where it should have been.

Did you move it to where you thought it should be? :yikes:

When I was Survey superintendent in the mangrove country of Southern Nigeria, working for a great big European oil company, one of my tasks was to survey the trig points which marked the boundary of several flow stations deep in the bush. The only way to survey there was by theodolite and distance measuring lasers. To do this meant having to cut sightlines through the dense jungle and this was very labour intensive and very time-consuming.

Local knowledge is always worth pursuing when surveying, so I offered the headman of the local Ijaw tribe village a reward of 10 Naira for every survey marker his people could find for me and I said I'd be back in a week.

A week later I went back to the village and was met by a very happy headman. He proudly said that his people had found 12 markers and that he'd show them all to me for an extra 80 Naira. I was delighted because I knew from the documentation in the office that 12 survey markers had been set out some 20 years previously and I'd been very careful not to tell him how many there were. I happily paid him 200 Naira and we retired to his hut to celebrate in the customary fashion with toasts of palm wine.

A short while later, he proudly said "I have something for you, Ogga", This is the customary preamble to the presentation of a gift to an honoured guest. With a flourish he drew back a curtain behind him and there on the floor were all 12 survey markers! Writing my Survey report and submitting my expenses claim wasn't easy.

But seriously, what did your GPSr show the position to be with WAAS? What did it show with WAAS deselected? What was the quaility indication figure in either case?

Did you use the OS grid co-ords? That would explain the 15 feet discrepancy.

Next time, try using the ETRS89 (WGS84) geographical co-ords, not the OS grid co-ords. Many handheld GPSrs (such as Garmin and Magellan) use a very crude conversion algorithm from WGS84 to OS grid and have an inbuilt error of 5 or 6 metres (7 in Scotland).

To see this for youself, input the geographical (Lat/Long in ETRS89) co-ords of your local trig point from the gps.gov.uk website into your GPSr. Then set your GPSr to display the OS grid co-ords of your input point and compare them with grid co-ords on the gps.gov.uk webpage.

You will probably (depending on which GPSr you have) find a discrepancy of something of the order of 15 to 20 feet -- though I can't tell you whether that is 15 feet to the left or to the right. :laughing:

Edited November 23, 2004 by The Forester

November 23, 2004

Good Morning Techies a question! on way to work with GPS strapped to handlebars it was rather dreich to say the least, the accuracy with 9 Sats was 145 metres.

A. Does the weather effect reception.

B. Will the reception be worse / better with WAAS on

Snaik

November 23, 2004

again...no single processor can receive multiple date at once....it just appears to!!

Hyper-Threading Technology

Pseudo but sweet.

Any Intel® Pentium® 4 powered gps out there? :yikes:

2.3k · November 23, 2004

B. Will the reception be worse / better with WAAS on

Arrrrrgh!!!! :yikes:

November 23, 2004

Modern GPSRs can receive in parallel all the receivable GPS nav messages without needing parallel processors.

Mythical....is not spelt MODERN...Forester

What it does is take each task in turn

A digital receiver can digitise the signal at the first IF stage on 12 channels in true parallel fashion and then use various digital signal processing techniques to demodulate the received signal and extract the data.

This has little impact on whether or not the CPU can mutlithread or not as the CPU is still only having to process a single output stream from the receiver module. It just that single stream may contain more data with each additonal channel that is feeding it and therefore will take more cpu cycles to process.

Therefore Forester is correct in the parallel reception can be achieved in true simultaneous fashion as this is a seperate circuitry module than that of the CPU. The output of this module is a single serial stream for the CPU to add to it poll list not many signals in parallel.

I made this up from the top of my head without research but its true dammit...maybe!

November 23, 2004

again...no single processor can receive multiple date at once....it just appears to!!

Hyper-Threading Technology

Pseudo but sweet.

Any Intel® Pentium® 4 powered gps out there?

This is achieved by duplicating the architectural state on each processor, while sharing one set of processor execution resources. Hyper-Threading Technology also delivers faster response times for multi-tasking workload environments. By allowing the processor to use on-die resources that would otherwise have been idle,

Hyper-Threading is just a clever way of dividing up the work to be done by parallel processors which have been set up in a specific way (duplicating the architectural state on EACH processor) to be able to receive and work on it!

This is not meant for a single processor computing environment as I understand it :rolleyes:

Yes computing technology has moved on at a fantastic rate since I was employed in the industry and certainly when I studied Computing Science 21 years ago the BBC micro was the home computer of choice :yikes: :yikes: :laughing:

How things have advanced since then....I am well out of touch by now :anicute:

Ullium.

November 23, 2004

A digital receiver can digitise the signal at the first IF stage on 12 channels in true parallel fashion and then use various digital signal processing techniques to demodulate the received signal and extract the data.
This has little impact on whether or not the CPU can mutlithread or not as the CPU is still only having to process a single output stream from the receiver module. It just that single stream may contain more data with each additonal channel that is feeding it and therefore will take more cpu cycles to process.

Therefore Forester is correct in the parallel reception can be achieved in true simultaneous fashion as this is a seperate circuitry module than that of the CPU. The output of this module is a single serial stream for the CPU to add to it poll list not many signals in parallel.

I made this up from the top of my head without research but its true dammit...maybe!

The fact that one can receive an electro-magnetic wave which carries more than one signal does in no way imply 'true parallel' processing :yikes:

These signals still have to be extracted from the carrier and processed by a single processor even if these signals are indeed being extracted electonically in parallel and fed to it ... well that is how it appears to me? What do you think Stonefisk?

It might be I am missing something here or I am just mega miles out of date....which would not surprise me in the least :laughing:

Ullium.

November 23, 2004

The fact that one can receive an electro-magnetic wave which carries more than one signal does in no way imply 'true parallel' processing

These signals still have to be extracted from the carrier and processed by a single processor even if these signals are indeed being extracted electonically in parallel and fed to it ... well that is how it appears to me? What do you think Stonefisk?

It might be I am missing something here or I am just mega miles out of date....which would not surprise me in the least

Ullium.

The fact that one can receive an electro-magnetic wave which carries more than one signal does in no way imply 'true parallel' processing

Who said it did? 12 signals being received on 12 receivers at the same time is true parrallel, no cpu processing involved. 12 signals received by one receiver on using 12 different channels is the same, still not CPU involved yet.

12 different coloured brooks attributing one river.

These signals still have to be extracted from the carrier and processed by a single processor even if these signals are indeed being extracted electonically in parallel and fed to it ... well that is how it appears to me? What do you think Stonefisk?

Erm yeah, that what I suggested? Carrier(s) have been recieved in parallel by modules independent of the CPU. These modules feeding a single data stream to the CPU. It does indeed need to extract the data which will take longer for each addtional channel in the SINGLE data stream. No, the CPU is not trying to process this signal data stream in parallel with other signal data streams .. as there are no other signal data streams this far into the process.

The CPU drinks some of the river and sees what colour brook it came from... one sip after the other. There is only one river carrying the coloured water, the cpu can only drink one colour from the river at a time. This is not parallel. The parallel part was over and done with long time ago... upstream.

I wish I was upsteam.

"The Forester" is correct the GPS receiver can receive all the signals in parallel.

"Ullium" is also correct the cpu is not processing in parallel (but really no one said it was or needs to), and he is aslo correct it does take longer to process as there is naturally more data to deal with.

Still haven't looked a block schematic of a GPS unit yet... please someone do, I refuse :rolleyes:

Edited November 23, 2004 by stonefisk

November 23, 2004

Yes as I said on a previous thread...this form of communicating is not the best

For the most part readers read what they think is implied...not necessary what is written. :yikes:

This is encouraged by the fact that we all try to talk in shorthand to keep our posts from becoming too long and boring :ninja:

Sorry I made my point badly Stonefisk :cool:

As I understand of the circuitry of this particular type of receiver (the GPSr) is they only have one aerial....no more....so no parallel aspect there would you agree ?

Also, the circuitry which separates out the different signals from the carrier wave is purely electronic to digital convertor(s) and nothing that could really be classed as a processor or any type :yikes:

As an old amateur Ham Radio enthusiast I could probably cobble up something myself to do this job....mark you... purchasing the valves might pose a problem :laughing: :rolleyes: :anicute:

Ullium.

3.8k · November 23, 2004

I bet Deva Duo wished they had never asked now !!

think they only wanted a "yes" or "no" :rolleyes: :yikes: :laughing:

November 23, 2004

Good Morning Techies a question! on way to work with GPS strapped to handlebars it was rather dreich to say the least, the accuracy with 9 Sats was 145 metres.

A. Does the weather effect reception.

B. Will the reception be worse / better with WAAS on

Snaik

Tsk Tsk Tsk

Does the weather affect reception ?

Yes ...but it is more likely that the induced error was due to the "patellic angular velocity " phenomenon ...a much misunderstood effect which is often experienced by those suffering from "letmeatititis"

:rolleyes: :yikes: :laughing:

1.4k · November 23, 2004

on way to work with GPS strapped to handlebars it was rather dreich to say the least, the accuracy with 9 Sats was 145 metres.

A. Does the weather effect reception.

B. Will the reception be worse / better with WAAS on

The "accuracy" (which is a misnomer but that's another issue) figure which you suggests to me that you were seeing the effects of multipath. This can be caused by the receiver accepting a signal from a reflection instead of a direct raypath. If your GPSr includes such a signal in its computation of position then fix quality will be degraded significantly.

If you were anywhere near buildings or electricity pylons or cables or lamp-posts or roadsigns or large vehicles at the time you saw the 145m figure, then I think multipath is likely to be the reason for the very high residual(s).

The answer to Question A is yes.

The L1 frequency which your GPSr receives is 1,575.42MHz. That's an ultra high frequency which has a characteristic of not being able to penetrate water to any significant extent. You can demonstrate it for yourself by looking at the signal strength page and then covering the antenna with your hand. It is not the bones in your hand which are blocking the signal, it is the water content of your hand. This is the reason why getting reception under tree with wet leaves is so difficult. Even a quite small amount of water in the raypath can absorb a great deal of the very feeble energy of the signal. Another characteristic is that radio waves at this frequency love to bounce off things, especially wet things.

In very heavy rain or below well developed towering cumulus or cumulo numbus clouds or in a heavy snow shower the signals can be attenuated so much that the GPS is struggling to obtain enough signal strength to keep track of the data from the satellites. It can easily loose lock on otherwise healthy signals and start to accept junk signals from multipaths (reflections).

The weather in Perthshire right now is a whole stack of layers of cloud, some of it very thick. Those conditions are not conducive to easy penetration by the GPS signals, but if you were getting 9 sats then low signal strength probably wasn't wasn't the main problem.

When considering signal strength, remember that the GPS satellites orbit at a very high altitude, something like 20,183 kilometers. The ones which are most useful are the ones with the lower angles of elevation from your position. They are even further away from you than the orbital altitude. Remember too that the transmitter power is a mere 50 watts. Think of how weak the signal is that your GPSr is trying to work with by thinking about how weak a 60 watt lightbulb would appear to you from 15,000 miles away.

Your very high figure of 145 metres was caused by a combination of circumstances, but was motly cause by multipath.

The answer to question B is neither.

WAAS can only augment the accuracy of a GPS signal. It cannot amplify it or eliminate mulitpath effects. It really wouldn't have been able to help you at all this morning.

WAAS can greatly improve the accuracy of GPS fix, but it can't make a silk purse out of a sow's ear.

There are a couple of things you can do when your GPSr is suffering from weak signals. The most obvious and most important is to ensure that the antenna is vertical. I often see bikers with a GPSr on the handlbars tilted as much as 60° off the vertical. Add in a very dense overcast and the proximity of buildings and its unsurprising that the wee beastie is sometimes struggling. Another thing to do on a very claggy day like today if weak signals are troubling you is to wipe the antenna dry.

BTW, I've just had a look at my Magellan, in weather conditions which here are pretty much identical to Perth, and it's getting a reasonably good signals and the "estimated position error" (for what that's worth!) is showing around 6 or 7 metres. I don't think that you could reproduce that high 145m figure even if you were to hop on your bike right now as the satellite geometry is constantly changing. Nevertheless, I'd be interested to hear what it shows on your homeward journey.

By the way, if you are ever on a cachehunt (on foot!) and you get such a high figure, the thing to do is figure out which satellite in use is suffering from multipath and then block it out of the computation. The way to do this is orient the azimuth/altitude page (which shows you where each satellite is relative to the horizon) to North and then block out each satellite in turn with your hand. When you've masked the rogue raypath the quality figure will pop down to normal levels.

November 23, 2004

As I understand of the circuitry of this particular type of receiver (the GPSr) is they only have one aerial....no more....so no parallel aspect there would you agree ?

Also, the circuitry which separates out the different signals from the carrier wave is purely electronic to digital convertor(s) and nothing that could really be classed as a processor or any type

As an old amateur Ham Radio enthusiast I could probably cobble up something myself to do this job....mark you... purchasing the valves might pose a problem

Ullium.

I'm confused now but anyway...

As I understand of the circuitry of this particular type of receiver (the GPSr) is they only have one aerial....no more....so no parallel aspect there would you agree ?

I cheated on Analogue RF electronics part of my degree because I would have failed the subject. The Antenna alone is a analogue RF device, its nether parallel or serial by definition(?) But maybe it could be stated as parallel due to the fact that as it receives pretty much EVERY radio signal in its part of the universe... everything that fits into its designed wavelength and probably more so. This reception is simultaneous, it has no choice, its only a dumb length of metal.

Also, the circuitry which separates out the different signals from the carrier wave is purely electronic to digital convertor(s) and nothing that could really be classed as a processor or any type

Again, RF electronics gah...but here goes. "electronic to digital convertors" err do you mean "Analogue to digital convertors"?

Anyway... the A/D conversion comes later in the process, its input is the filtered anlogue stream that comes out of RF convertor/frequency synthesizer, the A/D output is what ends up processed by the CPU.

Is this parallel ? I guess it all depends on is being fed into the A/D module. Is the RF convertor output feeding all the channels it has devired from the 'noise' simutanously into a multi-channel A/D? That would be parallel. Is there one anntenna but multipe RF convertor ciruits each filtering a particular channel?That would be in parallel. If the manufactures are to be believed it should be doing something like the above.

Okay peeked at a simple block schematic but it didn't detail the nature of the RF filter/convertor or A/D convertor.

Edited November 23, 2004 by stonefisk

November 23, 2004

Again, RF electronics gah...but here goes. "electronic to digital convertors" err do you mean "Analogue to digital convertors"?
Anyway... the A/D conversion comes later in the process, its input is the filtered anlogue stream that comes out of RF convertor/frequency synthesizer, the A/D output is what ends up processed by the CPU.

Is this parallel ? I guess it all depends on is being fed into the A/D module. Is the RF convertor output feeding all the channels it has devired from the 'noise' simutanously into a multi-channel A/D? That would be parallel. Is there one anntenna but multipe RF convertor ciruits each filtering a particular channel?That would be in parallel. If the manufactures are to be believed it should be doing something like the above.

Okay peeked at a simple block schematic but it didn't detail the nature of the RF filter/convertor or A/D convertor.

OK I think this must a case of an Englishman and a Scotsman being separated by a common language :rolleyes:

I too am a bit bemused....not that I don't understand what you are trying to say in general .... I think?..... but like myself your typing gets ahead of you sometimes and I think you have missed a few words and misspelt a few others which makes this reply a wee bit hard to follow (well neither of us are typists eh? :laughing: ).

I think you are claiming that the parallel aspect.... which you keep repeating....is involved in the cleaning and interpreting the incoming signal before the central processor gets it's hands on it ?

Whereas, I am attempting to make the point that this process does not involve anything that you could term as a processor....hence there is no multi-processing taking place.....albeit there may be other parallel aspects of the general process!!

So going back to your feeder burns to the main river....I don't think you are comparing like for like there.....hence the analogy is false...IMHO :yikes:

Ullium

November 23, 2004

So going back to your feeder burns to the main river....I don't think you are comparing like for like there.....hence the analogy is false...IMHO

Ullium

I too am a bit bemused....not that I don't understand what you are trying to say in general .... I think?..... but like myself your typing gets ahead of you sometimes and I think you have missed a few words and misspelt a few others which makes this reply a wee bit hard to follow (well neither of us are typists eh? ).

Probably yes. I do make mistakes and too lazy to correct on the 3rd read. it is true, tht I rely too mch on the fct tht most peopl's brians can automattcly filtr, translte and undrstnd the writtn wrd evn when as bad as ths.

I think you are claiming that the parallel aspect.... which you keep repeating....is involved in the cleaning and interpreting the incoming signal before the central processor gets it's hands on it ?

Yes.

Whereas, I am attempting to make the point that this process does not involve anything that you could term as a processor....hence there is no multi-processing taking place.....albeit there may be other parallel aspects of the general process!!

True, no processor involved, true there may be other paralled aspects, all true.

But I say again, I dont think anyone was suggesting that the processor was mutli-tasking, expect you kept repeating claims that the CPU are not true multi-tasking. Fine, you're correct but where is the argument? My understanding is that The Forester said that a GPS recieved all channels in parallel, this is true, no CPU involved on that part... you both agree, I agree too based on my assumptions. Still no argument.

I think everyone would agree that if WAAS data plus normal GPS channel data is in the stream there maybe (**see comment below) more data to process. This will take more processing time. Does this increase have a detrimental effect on the performance of a given GPS'R? I suggest that the increase has a negligible effect and well within the design spec of a GPS. This is esp true when you consider that a unit will hardly ever be in a geographical position to process the data of all 12 channels, because it does not see all 12 at once.

** it could be the case that there is no increase in the data hitting the CPU if the GPS signals that the WAAS negates are intelligently filtered/dropped upstream.

Edited November 23, 2004 by stonefisk

November 23, 2004

OK I think this must a case of an Englishman and a Scotsman being separated by a common language

I'm Welsh ya b****d :yikes: :laughing: :rolleyes:

November 23, 2004

OK I think this must a case of an Englishman and a Scotsman being separated by a common language

I'm Welsh ya b****d

My humblest apologies StoneFisk :anicute:

I keep forgetting that there are other Celts apart from us Scots on the forum :yikes:

And yes Forester didn't actually say that the GPSr was multi-tasking.....but I was left with the impression that he certainly implied it ??? Otherwise why make the comment (to be honest I now forget the exact words :yikes: ) parallel processing???

However, going back to what I was saying about not this not being the best medium for communicating ... perhaps this thread bears witness to that

Ullium.

November 23, 2004

Thanks Forester for an informative description that i actually understood,and GPS now horizontal on the bars

Snaik

Edited November 23, 2004 by Snaik

November 23, 2004

And yes Forester didn't actually say that the GPSr was multi-tasking.....but I was left with the impression that he certainly implied it ??? Otherwise why make the comment (to be honest I now forget the exact words ) parallel processing???

An online definition of Processing states "obtaining, recording or holding the data or carrying out any operation or set of operations on the data. It includes organising, adapting and amending the data, retrieval, consultation and use of the data, disclosing and erasure or destruction of the data"

Maybe one should not have assumed that a statement of "Parallel processing" involves a device known as a microcomputer central processing unit. Naughty assumer.

Edited November 23, 2004 by stonefisk

1.4k · November 23, 2004

parallel

Ullium,

Here's paste of the specifications page of the manual of my Magellan GPSr

>>>>

CHARACTERISTICS

Performance

Receiver 12 parallel-channel technology, tracks up to 12 satellites to

compute and update information with quadrifilar antenna

Acquisition Times (under optimal conditions):

Warm approximately 15 seconds

Cold approximately 1 minute

Update Rate 1 second continuous

Accuracy

Position 7 meters, 95% 2D RMS

w/WAAS <3 meters, 95% 2D RMS

Velocity 0.1 knot RMS steady state

Limits:

Speed 951 mph

Altitude 17,500 meters

<<<<

I haven't checked out the 951mph capability, but I know that the rest of the specs are spot on.

The only quibble I would have with their claims is that they don't make it clear that the acquisition times do not refer to the WAAS acquisition which takes much much longer due to the way the data stream is sent.

Cheers, The Forester

November 23, 2004

Yeh my Garmin Marine 76 is pretty similar....the question which bothers me is why it takes longer to get a fix on a geostationary satellite than it does an orbital one??

Unless of course you cycle very very fast like Snaik which negates somewhat the advantage of the geostationary concept :anicute: :yikes: :laughing: :rolleyes:

Ullium.

1.4k · November 23, 2004

why it takes longer to get a fix on a geostationary satellite than it does an orbital one??

OK, Ullium, lemme try to expalin explain that one without slipping into too much technowibble jargon.

The NavStar GPS sats transmit a message saying something like:

"Hi, my name's SV22. Here are my Keplerian orbital parameters which describe my flightpath today. I crossed over the equator (Northwards) at lunchtime and the time by Accurist is now half past lunch pip pip pip". It then has a long chat to you about the positions of all its brethren, very kindly and considerately so you don't have to wait until they all appear above your horizon.

From that, your GPSr can calculate exactly where the satellite is now, or at least, where it was at the instant in time when it sent that message. Your GPSr calculates its own Latitude, Longitude, Height and the Time by solving the simultaneous equation in four unknowns once it has locked onto at least four GPS satellites (only three if you've got an atomic clock in your geocaching goodie bag, but you haven't cos they make lousy swaps in microcaches and they cost more than your house).

That's a fix.

The WAAS sats don't send fix data as such. All they do is speak in a heavy GPS accent, so that your GPSr will pay attention and then they tell you all about the errors that GPSrs all over Europe (and America and bits of Africa and Asia Minor etc) are suffering from. Telling you that takes an awfully long time because the continents are split up into bazillions of areas rather like the pixels of a fine digital image.

Think of it as being like sending a high resolution linescan image of the Mona Lisa. It starts at the top left hand corner and scans to the right and drops down to the next line. Let's say you are geocaching somewhere up Mona's left nostril. Now, if you switch on your WAAS-enabled GPSr just as the thing is painting her right nostril, you will get your local area's WAAS correction data quite quickly. However, Sod's Law dictates that you switch on just as the thing is painting the lipline of that inscrutable smile. The thing then has to go on and do her chin and her boobs and the slender hands and then go back to the top lefthand corner and start all over again. You've got a long wait while it does that because there are so many pixels in that fine picture and you are just a wee pixel which has to wait its turn just like all the other pixels and the bouncers will thump you if you try to jump the queue.

So, the WAAS relay sats are sending an awful lot more data than the NavStar high flyers.

Now, there's another reason why your GPSr can lock onto the GPS birds faster than the InMarSat relay pigeons. Your GPSr is clever enough to remember what the state of play was when you so cruelly switched off its telly. It's got a clock and it knows the time and it remembers where it was when you put it to sleep. It is therefore ahead of the game because it knows, or at least can make a decent guess if you haven't moved more than a few hundred miles since then, which birds will be above the horizon. It knows which part of the lengthy navigation message should be being transmitted just about now so it can rapidly lock onto the datastream at whatever part of the cycle it happens to be. It's a bit like stepping onto an escalator. You judge the moment to step onto the thing without having to wait for a particular step to come around the cycle.

With the longer and more data-dense WAAS message it is much more complicated. Since your GPSr was last awake, the whole picture has changed. The space weather conditions have changed because the solar emission conditions have changed and because the time of day has changed the ionospheric corrections are completely different. The tropospheric conditions have changed because they do that sort of thing, just like the Govan weather. SV23 has had its clock rewound by the lighthouse keeper and it's been put forward to British Summer Time. SV24 has had a flat tyre and the AA hasn't arrived yet. SV25 had a rough passage over the South Atlantic gravitational anomaly and its orbit has gone quite wobbly. The bloke in Cheyenne Mountain spilt coffee over the keyboard and now SV26's ephemeris data is a load of cobblers. There's a new GPS sat in the array and it is transmitting junk. The whole picture of corrections is completely different.

The WAAS pigeon is now carrying a message which is no longer about that enigmatic bint Mona; it's now about a cartfull of hay painted by Constable. Your GPSr is smart enough to know where it is on earth, but it still has to wait until the wheel hub (which is where you're geocaching this time) of the haywain gets painted. That can take a long time, Sod's Law will see to that, especially if there is a bit of noise just as your pixel is being painted and you have to sit through the whole damned movie again until your bit comes around again.

Sometimes the GPSr can have to wait a long time to get the picture just right. Until it does so, it won't tell you WAAS stuff other than the fact that there's a WAAS bird sitting fat dumb and happy on a wire above the equator.

Now, have I mixed enough metaphors to explain it?

Cheers, The Forester

I reckon that's at least a pack of fags for WLW!

November 23, 2004

Sod's Law will see to that, especially if there is a bit of noise just as your pixel is being painted and you have to sit through the whole damned movie again until your bit comes around again.

Sod's Law, also known as Murphy ’s Law. If anything can go wrong, it will.

O'toole's commentary on Murphy ’s Law. Murphy was an optimist.

The first corollary to Sod's Law. Anything that is to go wrong will do so at the worst possible moment.

The unspeakable law. As soon as you mention something, if it's good, it goes away; if it's bad, it happens.

Zymurgy's first law of evolving system dynamics. Once you open a can of worms, the only way to re-can them is to use a larger can.

Skinner's constant. The quantity which must be multiplied by, divided by, added to or subtracted from the answer you get to give the answer you should have got.

Law of selective gravity. An object will fall so as to do the most damage.

Jenning's corollary. The chance of the bread falling with the buttered side down is directly proportional to the cost of the carpet.

Barth's distinction. There are two types of people: those who divide people into two types and those who do not.

November 23, 2004

I think I got the picture Forester :anicute: :yikes: :yikes: :laughing:

I should have thought of that myself :rolleyes:

Ullium.

1.4k · November 23, 2004

Murphy’s Law.

Murphy's Law is different from Sod's Law, at least in air force parlance.

Murphy's Law says that if anyone can incorrectly fit a component to an aircraft, then sooner or later somebody will.

Sod's Law says that if anything or anyone can go wrong at the worst possible time: they will.

Forester's newly discovered Law says that thread drift is inevitable!

November 23, 2004

To bring the thread back onto topic a bit. After reading the Forresters last explanation I almost understand part of the WAAS/EGNOS theory although I am still a bit confused.

We are looking into buying a second GPS purely for geocaching - we prefer to use paper map and compass for mapping- so would it be beneficial to us to get one which can be WAAS enabled? ie would it ever get us closer to the cache and if so how often would we actually be able to get the EGNOS signal?

1.4k · November 23, 2004

we prefer to use paper map and compass for mapping

So do I.

would it be beneficial to us to get one which can be WAAS enabled?

If all you want to do is plot your realtime GPS-obtained co-ords onto a 1:50,000 Ordnance Survey map, then WAAS is of no advantage to you.

WAAS gives you an accuracy which is unplottable at such a scale on a paper map.

would it be beneficial to us to get one which can be WAAS enabled?

I'm heavily biased. I love the improvement in GPS accuracy which WAAS can give us for the same reason that I campaigned so heavily and for long for Selective Availability (SA) to be abolished. WAAS gives us a comparable improvement in GPS accuracy which the abolishment of SA gave us. I think that is worth having.

would it ever get us closer to the cache

Only if the given co-ordinates are as good as GPS can give us.

If the given co-ords are junk, then all the WAAS and DGPS in the world cannot help us.

how often would we actually be able to get the EGNOS signal?

On most days, these days, you can get a WAAS (EGNOS for us in Europe) signal for most of the time. That will get better over the next few months and will soon become a matter of routine, 24x7.

At present, the service is not constant because they are still testing the system.

Cheers, The Forester

November 23, 2004

Gentlemen, Gentlemen, I'm sure this topic has well and truly been discussed and that there is nothing else to say here without repeating yourselves. We are now well informed on WAAS and EGNOS until it oozes out of our ears! Perhaps this thread has reached its capacity now.

Just tell me, If you ran out of a pint of milk, how many satellites would it take to reach Tesco before closing time? :yikes:

November 23, 2004

Gentlemen, Gentlemen, I'm sure this topic has well and truly been discussed and that there is nothing else to say here without repeating yourselves. We are now well informed on WAAS and EGNOS until it oozes out of our ears! Perhaps this thread has reached its capacity now.

Just tell me, If you ran out of a pint of milk, how many satellites would it take to reach Tesco before closing time?

When do the Pubs close in Perth pp?

November 23, 2004

Personally I find quite a few of the threads on this forum quite boring too PP :yikes:

But what I do is just ignore them.... :laughing:

Ullium.

Edited November 24, 2004 by Ullium

1.4k · November 24, 2004

Thanks Forester for an informative description that i actually understood

Y' welcome, Snaik.

I've been thinking some more about Question B and I may have to revise my answer for users of the Magellan SporTrak GPSr.

The quality indication figure (they call it EPE on my Magellan) only appears if WAAS is not on. If you are moving while receiving WAAS, the line which otherwise shows EPE simply says "WAAS". If you are stationary (defined by Thales as a speed less than ½mph) the EPE line simply says "WAAS averaging" and it shows how long the averaging has been going on.

This brings up an interesting flaw in the top level application software of this machine.

You were able to see that your fix was having a problem because the "accuracy" figure showed a huge standard deviation of the residuals. This indicated what was probably a multipath problem with the fix.

If you had been using a Magellan SporTrak and if there had been WAAS signals available at the time, your machine would not have told you that there was a major problem with the fix.

I don't advocate a dependency on the reliability or worthiness of those quality figures which some people believe are an indication of accuracy, but in the case of a gross error caused by multipath, they are invaluable.

In the specific case of a SporTrak suffering from severe multipath at a time when also receiving WAAS data, perhaps it would be better not to be receiving the WAAS data after all. Unfortunately, the problem with the machine is that you cannot deselect WAAS by using any of the menus shown on the screen or in the manual if you want to have a quick look at the quality indication figure (EPE). There is a complicated undocumented procedure to do so, but it involves shutting the machine down at least once (actually, twice if you want to go back to using WAAS) and there is a risk that the machine will not reboot properly at all if you make a simple keystroke error at any stage of the non-standard procedure.

This whole sub-issue of how to assess the accuracy of a GPS fix is an important one, though not necessarily related to WAAS (except for viewers in Scotland who use the SporTrak!).

Too many co-ordinates published by cache-placers contain gross errors. There are several reasons for this: most of which are a reliance on the co-ordinates shown instantly on the screen at the push or wiggle of a button. Perhaps a sensible interpretation of an indicated fix needs a bit of knowledge.

Unfortunately, the otherwise excellent manuals of some of the major producers of retail GPSrs are very poor at telling people how to use the data or even how to assess it.

Cheers, The Forester

November 24, 2004

Morning Forester took this photo this morning. both WAAS enabled:

Sport Trax EPE 6 M coords N56 24.354 W003 25.818

Garmin Accuracy 6M coords N56 24.356 W003 25.822

kept a forward motion goig to stop the Magellan averaging

Snaik

November 24, 2004

As usual I agree FFB with all your comments above :lol:

And you are correct that there is very little information on what procedure one should apply to be reasonably certain that the co-ords one posts are at least reasonably accurate!

Now I know this has been discussed before and what I understood from that was to walk away and return a few times to a specific location and take a different waypoint on each return....then average the results!

I've read what appears to suggest that some GPSr's can do this averaging on the spot?? Or am I mistaken ??

In any event....is there anything else I can do to ensure I get as accurate an estimate of location co-ords as I can??

(In language everyone can understand :huh: )

Apart from switching WAAS on that is :laughing: :huh:

Ullium.

1.8k · November 24, 2004

I've read what appears to suggest that some GPSr's can do this averaging on the spot?? Or am I mistaken ??

My new toy (Garmin Map60CS) does just that - when setting a mark the option is to accept the immediate fix, go to map or average. The latter appears to sample every second until you save the result.

Previously I did as suggested above - leave to settle for a while then make two or three fresh approaches to the spot allowing to settle each time. Or of course under trees etc take an offset from a spot with good reception.

1.4k · November 24, 2004

Morning Forester

G'morning, Snaik.

Thanks for the data point(s).

What the piccie shows is that the standard deviation of the residuals of the simultaneous fixes received by both the Garmin and the Magellan were similar.

It also shows that neither was receiving WAAS signals at the time of the displayed fix.

It also shows that the difference between the two displayed co-ords was 5.54m which is an excellent match with the quality indicator but says almost nothing about the real difference between the indicated co-ords and the true value unless you take extraneous data into account.

Your data is good data and very useful, please keep it coming.

Cheers, The Forester

November 24, 2004

So what is the comparison coords to metres? I guess this changes at every location but is there a rough guide that would cover the UK for us amateurs? I'm all in favour of learning new gps tricks.

November 24, 2004

I've been thinking some more about Question B and I may have to revise my answer for users of the Magellan SporTrak GPSr.

The quality indication figure (they call it EPE on my Magellan) only appears if WAAS is not on. If you are moving while receiving WAAS, the line which otherwise shows EPE simply says "WAAS". If you are stationary (defined by Thales as a speed less than ½mph) the EPE line simply says "WAAS averaging" and it shows how long the averaging has been going on.

If you had been using a Magellan SporTrak and if there had been WAAS signals available at the time, your machine would not have told you that there was a major problem with the fix.

In the specific case of a SporTrak suffering from severe multipath at a time when also receiving WAAS data, perhaps it would be better not to be receiving the WAAS data after all. Unfortunately, the problem with the machine is that you cannot deselect WAAS by using any of the menus shown on the screen or in the manual if you want to have a quick look at the quality indication figure (EPE). There is a complicated undocumented procedure to do so, but it involves shutting the machine down at least once (actually, twice if you want to go back to using WAAS) and there is a risk that the machine will not reboot properly at all if you make a simple keystroke error at any stage of the non-standard procedure.

This whole sub-issue of how to assess the accuracy of a GPS fix is an important one, though not necessarily related to WAAS (except for viewers in Scotland who use the SporTrak!).

Not quite correct ...... on the Sporttrack Pro , there is a customise option which allows you to add or remove any of the list of parameters ...so one of my navigation screens shows:-

postion (including sat status...which would normally say averaging )

Bearing and EPE

Also the option to disable Waas whist not a menu option is a very simple change to make see attached .....

Magellan Codes

hope this helps

regards FFB

1.4k · November 24, 2004

FFB,

Can you tell me how I get this cutomise option to display EPE while WAAS averaging?

Is it in the manual or is it one of those undocumented funnies?

Cheers, The Forester

November 24, 2004

From the navigation screen which shows :-

position

altimeter

time

date

search

datum

alternative co-ords ( in my case OS)

datum

use the left or right arrow to select the alternative screen ....this shows :-

position

altimeter

time

date

search

and 3 configurable options:-

in my case EPE ; Bearing

and trip Odometer .

from this screen select menu ;customize; and et voilà all becomes clear

regards FFB

edited ..for Bruichladdich induced spelling errors

Edited November 24, 2004 by Flyfishermanbob

1.4k · November 24, 2004

et voilà

Je vous remercie, Monsieur!

Waas

Recommended Posts

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment