Speed and distance logs-do we need them?

[lw395]

Speed is available directly from GPS using doppler analysis of the signals from the satellites in use. It can be done from differentiating positions, but that's a pretty crude method (differentiated error is never pretty) and I'd be mildly surprised if any GPS sets today did it that way.

I think you will find all consumer GPS works out speed from position.
Then they tend to put a bit of processing in to reduce the indications of wandering about at 1 knot when you are stood still.
A good log with 0.01 knot resolution is a big help with sail trim.
Being a dinghy racer, I'm used to not having a log, but then we have really sensitive speed indication in the form of gaining/losing against an identical boat.

 

[AntarcticPilot]

I think you will find all consumer GPS works out speed from position.
Then they tend to put a bit of processing in to reduce the indications of wandering about at 1 knot when you are stood still.
A good log with 0.01 knot resolution is a big help with sail trim.
Being a dinghy racer, I'm used to not having a log, but then we have really sensitive speed indication in the form of gaining/losing against an identical boat.

+1 My understanding is that it is done as an output of the Kalman-type filtering done on the position data. However, I am not privy to the internal software design of GPS units; I am basing my understanding on my knowledge of signal processing in analogous fields (e.g. window tracking in satellite altimeters).

 

[JumbleDuck]

I think you will find all consumer GPS works out speed from position.

I don't think they ever have. There's a rather old article about it at http://www.aprs.net/vm/gps_cs.htm - you'll see that it refers to "the newer Garmins (GPS-25, GPS-48, GPS-12, GPS-12XL, GPS-II+, GPS-III etc)" making use of Pseudo Range Rates (PRR), the Doppler speed information, from several satellites, and those receivers are ancient history now.

Working out speed from position is bad enough nowadays, because of the hazards of differentiating noise, but would have been even worse when SA was active.

 

[JumbleDuck]

+1 My understanding is that it is done as an output of the Kalman-type filtering done on the position data. However, I am not privy to the internal software design of GPS units; I am basing my understanding on my knowledge of signal processing in analogous fields (e.g. window tracking in satellite altimeters).

They use Pseudo Ranges (PRs) and Pseudo Range Rates (PRRs) from as many satellites as they can see, then combine present and historical data to get the best estimates for current values. However, speed comes from the corrected PRRs and not from changes in the position found from the PRs.

 

[AntarcticPilot]

They use Pseudo Ranges (PRs) and Pseudo Range Rates (PRRs) from as many satellites as they can see, then combine present and historical data to get the best estimates for current values. However, speed comes from the corrected PRRs and not from changes in the position found from the PRs.

Thanks, that's interesting. You mention combining present and historical data - I presume that's some form of predictive filtering? That would introduce a time lag to the response to changes of velocity.

My experience is from the software used in satellite altimeters to position the range-gate using predictive filtering of previous ranges.

 

[KellysEye]

I took ours our years ago I got fed up with cleaning it. If you have sailed a lot you don't need it we know the speed of the boat over the ground, tides or not. We use a GPS for navigation.

 

[JumbleDuck]

Thanks, that's interesting. You mention combining present and historical data - I presume that's some form of predictive filtering? That would introduce a time lag to the response to changes of velocity.

It surprised me when I first found out how GPS does speed, but on reflection I suppose that if you can measure distance from a satellite to an accuracy of under a wavelength then measuring frequency changes in the incoming signal isn't any more complicated - and possibly less complicated.

As regards the filtering, I have done vaguely similar things in modelling non-linear heat flow, but I don't know any details of how GPS does it. I imagine that the lag will be down to the speed of processing, rather than the frequency of fixes ... which in any case I suspect is determined more by the need to avoid flickering displays than by hardware limitations.

 

[AntarcticPilot]

It surprised me when I first found out how GPS does speed, but on reflection I suppose that if you can measure distance from a satellite to an accuracy of under a wavelength then measuring frequency changes in the incoming signal isn't any more complicated - and possibly less complicated.

As regards the filtering, I have done vaguely similar things in modelling non-linear heat flow, but I don't know any details of how GPS does it. I imagine that the lag will be down to the speed of processing, rather than the frequency of fixes ... which in any case I suspect is determined more by the need to avoid flickering displays than by hardware limitations.

Predictive filtering, by its nature, reduces the sensitivity of the system to changes in the measured quantity. This manifests itself as lag in the output compared with reality. It's OK with smoothly changing variables, but tends to go off a bit when confronted with a sudden change. The output graph tends to look a bit like the wake of a boat under autopilot responding to a sudden change of course

I've got several papers on my bookshelf related to this - I happened to be involved with the ERS-1 altimeter mission.

 

[JumbleDuck]

Predictive filtering, by its nature, reduces the sensitivity of the system to changes in the measured quantity. This manifests itself as lag in the output compared with reality. It's OK with smoothly changing variables, but tends to go off a bit when confronted with a sudden change. The output graph tends to look a bit like the wake of a boat under autopilot responding to a sudden change of course

Oh, absolutely. I just don't think that the systems are inherently limited to a position or speed fix every two second - I think that's likely to be a power saving measure as much as anything, and I expect that the GPS systems in airliners do much more frequent fixes.

In Ye Early Days of GPS, standard Garmin units blanked the display if your speed exceeded 100kt, unless you bought an aviation version, like the GPS90 which cost twice as much as its non-aviation counterpart. The only differences between them were (a) the case colour and (b) an extra link on the circuit board of the restricted version. Cut that and you were good to go as fast as you wanted. There was no reception issue, by the way, because the set kept on fixing position and speed at 100kt+ - it just wouldn't tell you what they were.

One of the nice things about moving from flying back to sailing is how cheap everything is!

 

[AntarcticPilot]

Oh, absolutely. I just don't think that the systems are inherently limited to a position or speed fix every two second - I think that's likely to be a power saving measure as much as anything, and I expect that the GPS systems in airliners do much more frequent fixes.

In Ye Early Days of GPS, standard Garmin units blanked the display if your speed exceeded 100kt, unless you bought an aviation version, like the GPS90 which cost twice as much as its non-aviation counterpart. The only differences between them were (a) the case colour and (b) an extra link on the circuit board of the restricted version. Cut that and you were good to go as fast as you wanted. There was no reception issue, by the way, because the set kept on fixing position and speed at 100kt+ - it just wouldn't tell you what they were.

One of the nice things about moving from flying back to sailing is how cheap everything is!

The 100 kt limit never bothered us - we were using it for survey. But the vagaries of the incomplete constellation, coupled with the natural problems of ionospheric variations, meant that in the very early days survey work could require a lot of planning beforehand, with a fairish chance of not getting any results anyway! Of course, that was in Antarctica, where the ionospheric difficulties were greater. I was at arms-length to it; it was colleagues who were using it, not me. But I still remember the awe with which Trimble's first hand-held GPS was received! We used systems that recorded the data for post-processing against a fixed station.

 

[Buck Turgidson]

If you're interested in the boat's response to sail trim, you need one. GPS is slow to respond to changes in speed - the interval between fixes is two seconds, and internal filtering of the positions to increase accuracy means that the speed output will be even slower to respond accurately to changes in speed. A paddle-wheel log responds pretty much instantaneously to changes in sail-trim, you can see the effects of your latest tweak immediately. If you rely on SOG, the wind may change over the period that it takes for the GPS to respond to your adjustments.

As others have said, it measures a different thing from your GPS SOG, and that may well be useful, especially if sailing in an area where there are back-eddies and turbulent flow below the resolution of tidal charts - most of the tidal gates on the West Coast of Scotland are like that. Knowing the tide vector is a different problem to knowing the height of the tide and is a much more difficult problem, and less tractable using tables etc. Combining log and GPS means you have potentially complete knowledge of the various vectors affecting your boat.

Not sure where you got this info from but it's about 25 years out of date.
10hz is quite common these days.

 

[AntarcticPilot]

Not sure where you got this info from but it's about 25 years out of date.
10hz is quite common these days.

When I were a lad, it could be 10 hours between fixes, if the satellite positions weren't right! We were using it in the polar regions before the constellation was complete.

Perhaps I should have said, the interval between independent fixes is about 2 seconds. Given that the bit-rate of the data is about 50 bits/second, that's going to be (roughly) the interval at which the satellite data is refreshed. Of course software can output positions more frequently, but as noted already, the fixes aren't independent of each other, and it probably takes something in the region of 4 seconds (I don't know the exact time-constants) before fixes are totally independent of each other.

 

[Vara]

Use paddle log a lot in trim mode, ideal for sail trimming.

 

[Buck Turgidson]

When I were a lad, it could be 10 hours between fixes, if the satellite positions weren't right! We were using it in the polar regions before the constellation was complete.

Perhaps I should have said, the interval between independent fixes is about 2 seconds. Given that the bit-rate of the data is about 50 bits/second, that's going to be (roughly) the interval at which the satellite data is refreshed. Of course software can output positions more frequently, but as noted already, the fixes aren't independent of each other, and it probably takes something in the region of 4 seconds (I don't know the exact time-constants) before fixes are totally independent of each other.

The C/A code is repeated every millisecond. You are talking about the nav message data rate. Difference between cold acquisition and tracking!

 

[Marsupial]

All very interesting, all I can add to this is that if I take a handheld GPS in my car the speed it displays lags the cars actual (indicated on the speedo) speed by about 4 - 5 seconds under acceleration but most noticeable when stopping- I.E. It says we are moving when stationary at traffic lights - for about 4 -5 seconds as the speed on the GPS slowly decays to zero.

 

[William_H]

I have a paddle wheel log which is invariably jammed up with little mussels after four weeks afloat. I'm far too scared to remove and clean it, so it stays that way, and I don't feel any loss. Quite the reverse, really, as I prefer to sail with as few instruments as possible. I use a Walker's log for longer trips, mainly for fun. GPS does all the important stuff.

Don't be afraid to pullo the paddle wheel out while in the water. When I had one (now long gone due to fouling) I would lie on my belly on the floor pull with right hand and then slap the left hand over the hole. I could free it one handed usually or replace it with the provided cap. Then off with the left hand sealing the hole and in with the paddle. About 1 cup of water comes in each change. Easily mopped up. You even get to admire the lovely blue light that comes from the water whne the hole is open.
Certainly around here with fouling you either clean it before every voyage or keep it out and fit before each trip and remove before leaving the boat. Too much trouble for me. olewill

 

[Bru]

Must admit that on Brigantia I pretty much gave up on the log as it was wildly inaccurate (had to max out the calibration to get it anywhere near and it wasn't that good even then)

On Erbas however the Raymarine log is normally pretty well spot on as far as I can tell and as already mentioned dead easy to remove for cleaning and when the boat isn't in use due to the cunning flap in the housing tube

I keep the blanking cap tucked in the bilge next to the fitting anyway and it is the work of but a few moments to pull the impeller and fit the cap. Most of the time it doesn't even let enough water in to trigger the auto bilge pump!

Currently, due to a nasty dose of the barnacles, with a side order of weed, the log is under reading badly and this is a minor PITA cos it affects several other useful, albeit non -essential functions (can't get True Wind on the wind instrument, Set and Drift on the plotter is meaningless, Wind Trim mode on the tiller pilot doesn't work properly (no, I don't get that one either but it doesn't) and what have you)

And, of course, it's directly useful for instant feedback on whether an alteration of sail trim has had the desired effect

So whilst I could live without it, it does serve a useful purpose

 

[Fantasie 19]

I'm glad I'm not the only one then. Being a bilgy I must have a go at this drying out lark to clean the little beastie out!

Another vote for the Walker log.

Nicely dried out at the moment, Pete..

 

[Momac]

I made a tidal river journey yesterday . I made reference to both SOG and speed through the water. It's not essential to have both figures but it can help to verify the affect of the tide. Indeed such a journey can easily be made without any instruments.
I know more of less what speed through the water my boat should be doing at certain engine speeds so this information does have a use to verify whether all is well. Very occasionally the paddle wheel sticks but a blast of speed invariably clears the problem. My speed log also shows depth so it is always on when under way , the lack of depth often being our greatest problem.