So, which one of these is more accurate?

[Tally Ho]

sailed my O'Day 322 on the south end of Lake Michigan today. Had a nice south wind at about 12 kts, gusting to 17-18 kts. Lake was fairly smooth with the south wind. No currents or tides to deal with.

I have a Data Marine knot log, a Garmin Chartplotter with GPS and ran 2 sailing apps on my iPad.

I reset the maximum speed on the Chartplotter after we killed the engine, and started the 2 apps after I set the sails.

I would have assumed that a GPS is the more accurate than my speed log, but which of the GPS speeds is more accurate (and why is there so much variance)?



It was a great sail, and maybe I am just bragging a little, but still curious about the variation in the GPS speeds.

Greg

 

[SFS]

Are all three speeds GPS derived? If the chartplotter speed is from a through hull, then you would have to account for the difference in boat speed and SOG.

The average GPS speeds agree pretty well. 5.75 knots in the middle picture and 5.8 knots in the third picture are the same number. Tough to determine an average from all the instantaneous readouts in the first picture, but something close to 6.1 maybe?

 

[smokey73]

For the GPS, the SOG will include yaw and pitch. Depending on where the gps sensor is located, how often the SOG refreshes, etc it can have a pretty big effect on SOG. The first chart seems like it is taking the gps location and calculating the SOG very often accounting for the rapid change and variation. In winds like that and a boat that size I would imagine you were pitching a lot which accounts for the great variation in short times. Imagine putting your handheld gps on the end of a stick and waving it fore and aft while walking. At any given instant in time, depending on when the GPS refreshes the data, it will sense the motion of the gps on the stick and add or subtract that from your walking speed. That is why a gps without some pretty sophisticated software would not be worth much at the top of the mast for SOG. Just watch how fast your masthead moves forward when you pitch forward! It would be okay for position since that only changes a very little with the pitch but it changes in relative position and thus velocity very quickly. Just my 2 cents - your mileage may vary.

 

[Ron20324]

Motor along at 5 knots according to the paddlewheel speed into the tide. Do a 180 without touching the engine speed. You'll be going "slower", though the GPS will have you faster by maybe a knot. I have your same instrument setup, and go by the GPS if it makes and difference how fast I go.

 

[Tally Ho]

Are all three speeds GPS derived? If the chartplotter speed is from a through hull, then you would have to account for the difference in boat speed and SOG.

The average GPS speeds agree pretty well. 5.75 knots in the middle picture and 5.8 knots in the third picture are the same number. Tough to determine an average from all the instantaneous readouts in the first picture, but something close to 6.1 maybe?

They are all GPS speeds, but from difference sources ( Chartplotter and iPad).

Greg

 

[Tally Ho]

For the GPS, the SOG will include yaw and pitch. Depending on where the gps sensor is located, how often the SOG refreshes, etc it can have a pretty big effect on SOG. The first chart seems like it is taking the gps location and calculating the SOG very often accounting for the rapid change and variation. In winds like that and a boat that size I would imagine you were pitching a lot which accounts for the great variation in short times. Imagine putting your handheld gps on the end of a stick and waving it fore and aft while walking. At any given instant in time, depending on when the GPS refreshes the data, it will sense the motion of the gps on the stick and add or subtract that from your walking speed. That is why a gps without some pretty sophisticated software would not be worth much at the top of the mast for SOG. Just watch how fast your masthead moves forward when you pitch forward! It would be okay for position since that only changes a very little with the pitch but it changes in relative position and thus velocity very quickly. Just my 2 cents - your mileage may vary.

I think the Chartplotter has a "speed filter" that sort of averages out the big spikes. I think that must be the answer.

Either way, it was a fast sail and a lot of fun.

Thanks for all the info.

Greg

 

[dlochner]

Handheld GPS devices will measure the speed that the unit is moved. If you pick up the unit and move it forward the speed at which you move the unit will be added to the speed of the boat, giving a higher reading. The GPS speed indicates the speed the GPS is traveling, not the boat speed.

Some GPSs and chart plotters will allow you to set the number decimal points to be shown. If one device is set to 1 decimal place (tenths) and one set to 2 decimal places (hundredths) they will never read exactly the same due to rounding errors; 6.75 rounds to 6.8, 6.74 rounds to 6.7.

GPSs also read speed over ground not speed through water. Knot meters read speed through water, not over ground. Current, leeway, impeller placement, angle of heel, tack are all factors that can contribute to different readings.

In terms of general navigation, small differences are not that meaningful. A tenth of a knot difference in your calculations will only yield a 600 ft difference in distance after an hour. After 10 hours it would only mean a distance of 1 nautical mile.

 

[capta]

A free or low cost 'app' versus a unit designed for this exact purpose and fairly costly? I know which one I'd rely on.

 

[Benny17441]

The knot meter reads speed over water and the GPS reads speed over the bottom. Some GPS units are more precise than others and the signal integrity may also vary.

 

[Tally Ho]

Handheld GPS devices will measure the speed that the unit is moved. If you pick up the unit and move it forward the speed at which you move the unit will be added to the speed of the boat, giving a higher reading. The GPS speed indicates the speed the GPS is traveling, not the boat speed.

Some GPSs and chart plotters will allow you to set the number decimal points to be shown. If one device is set to 1 decimal place (tenths) and one set to 2 decimal places (hundredths) they will never read exactly the same due to rounding errors; 6.75 rounds to 6.8, 6.74 rounds to 6.7.

GPSs also read speed over ground not speed through water. Knot meters read speed through water, not over ground. Current, leeway, impeller placement, angle of heel, tack are all factors that can contribute to different readings.

In terms of general navigation, small differences are not that meaningful. A tenth of a knot difference in your calculations will only yield a 600 ft difference in distance after an hour. After 10 hours it would only mean a distance of 1 nautical mile.

I understand the difference between the paddle wheel and the GPS. I generally assume the GPS is more accurate. Often the paddle wheel and and GPS are the same on one tack, but different on the opposite tack. I assume this is due to the placement of the paddle wheel on one side of the hull.

In this particular case, I did not move the iPad, so it should have been "boat speed". But it was placed in the cabin, while the Chartplotter is mounted at the helm.

Like you said, small differences don't mean much, but was surprised by the variance in the 2 different GPS devices, and even the different apps on the same iPad.

Thanks,

Greg

 

[kloudie1]

It has to do with the sample speed and averaging speed of the unit.. The app looks to be sampling faster which would account for it "catching" the peaks more often where the Garmin is averaging those peaks into a longer time spot.. Many apps and plotters have an option in "settings" that allows you to set the sampling interval.. If you set them all the same, the results will not vary as much .. Trust the chartplotter.. having a one knot spike for the time you are running down the front of a wave (maybe 30 feet) is pretty much meaningless except for bragging rights.
EDIT: Example.. for max speed, chartplotter averages speed every 3/10 sec. so it sees 3, 6, 3 as knots in the 3/10ths that it is looking at. it reports 4 knots max (the average over the 3/10th.. .. app looks every 2/10ths so it sees 6, 3.. the average is 4.5 knots.. so it reports that as max speed.. not quite that simple, but works as an example..

 

[Jackdaw]

GPS chipsets report lat, long, and speed every second, or 10x a second. Errors in any of these three values can be created via several circumstances, and the software of the device needs to sort it out. To do this, most will average speeds for a period of time. The best ones will do that, along with throw out impossible values. Impossible values are created by impossible levels or acceleration or deacceleration. Kind of the way ABS brakes work, a way-to-sudden burst of speed ( say 5.0-5.0-5.0-7.0-5.2-5.0) is an impossibility in a sailboat and the software should just throw it out.

Good ones do, bad ones do not.

 

[Bill Roosa]

First things first. GPS does not measure speed. It does calculate the speed based on two positions it measures and the time between those measurements. the formula is: (position1-position2)/(time between 1&2)=speed. That is why mounting the antenna at the top of the mast will yield a very erratic speed calculation.
As far as which one is more accurate?..... well if by accurate do you mean how well do they measure your position that is pretty easy to determine. Just compare your known position with what the GPS reports. If you are trying to determine which one calculates the most accurate speed then the one with the antenna fixed to the boat near the center of gravity and updating more frequently AND has the most satellites in view will have more accurate results. Bigger antennas and GPS units that can track more satellites genarally have better accuracy. Course all should measure position to about 30 feet so you are not going to get much variati

 

[Jackdaw]

First things first. GPS does not measure speed. It does calculate the speed based on two positions it measures and the time between those measurements. the formula is: (position1-position2)/(time between 1&2)=speed.

Ah... No.

For the last 10+ years, GPS chipsets have used Doppler shift to calculate speed during a single fix.

 

[Bill Roosa]

No, they use dopple shift to account for the speed of the satellite and correct for errors introduced when the radio waves travel through the ioniosphere where the speed of the waves change. Global POSITIONING System -GPS

 

[Jackdaw]

Bill, Please read up. I used to run the Product Planning Group at Magellan. My name's on 6 of the patents. I'm making this up; this was my job. Modern GPS devices use 'apparent doppler' to calculate speed.

Start here.

GPS Determination of Course and Speed
br Tom Clark, W3IWI
First -- let me describe how your GPS receiver measures position and velocity. Your receiver tracks (at least) 4 GPS satellites. The DSP tracking loops built into the receiver have to lock onto two parts of the signal from each satellite -- the 1.023 Mb/s C/A code and the 1575.42 MHz carrier frequency. The dual DSP operation involves two tracking loops -- a CLL (Code Locked Loop) for the C/A code, and a PLL (Phase Locked Loop) tracking the carrier.
The CLL produces the timing of the GPS signal WRT (with respect to) the receiver's internal clock -- this is normally expressed in distance units called a Pseudo-Range (PR) and it includes the error in the receiver's internal clock added to the real geometric range of the satellite. Similarly, the PLL measures the carrier phase rate (i.e. apparent satellite frequency with respect to the receiver's local oscillator, usually derived from the same crystal oscillator that is used as the timing clock) -- this is called the Pseudo-Range Rate (PRR) and includes the frequency error of the receiver's LO and the contribution from the Doppler shifts associated with all the motions. The Doppler shift includes the vector sum of the satellite's ~7 km/sec orbital velocity plus the 400 m/sec (at the equator) rotational velocity of the earth plus your receiver's motions (in a moving car, ~10-50 m/sec).

In early GPS receivers, four PRs from 4 satellites was converted into a 3-D (XYZ, Lat/Lon/Hgt or whatever) position plus the calibration of the timing bias of your receiver, and 4 PRRs were converted into a 3-D velocity plus a measurement of the frequency error of the oscillator. More modern receivers take all the PR+PRR data from all the N satellites in view for the past T seconds and feeds the 2*N*T PR+PRR samples it into a single mathematical "black box" (BB) (usually a Kalman filter) to produce an over- determined estimate of the same 8 parameters. So in modern receivers, this BB is using both the combination of past & present PRs and PRRs from many satellites to improve the Position, Velocity & Time (PVT) estimate. So when you look at the equations inside the BB), the measured "apparent Doppler" frequencies are even more important.

 

[Bill Roosa]

well the current version of Bowditch would disagree with you.
I'm thinking that a satellite would be approaching/departing from me at some random angle. The Doppler shift would therefore also be random so how would you know what Doppler shift to use to calculate your speed? It does not make sense from a math and geometry viewpoint.
Perhaps we are talking about a different kind of Doppler shift. I'm talking about the shift caused by the speed between the satellite and antenna. You do that at least 4 times (4 different satellites) and note the time hack sent with each satellite transmission and you can account for all the errors introduced by the radio waves not traveling at the speed of light through the ionosphere and atmosphere (highly variable), the speed of the satellite and the speed of the boat. The time hack also allows the unit to account for the time it takes to receive each of the 4 satellite transmissions and any movement of the boat (very small but they use GPS on jets!!) in that time.
Or did I miss something
For the record I used to work in the Joint Staff J65 which handles among other GPS and I was the senior action officer on the GPS program.

 

[Squidd]

If you really want to get screwed up...compare your direction (cog) on three different GPS and your compass reading.

 

[Jackdaw]

Bill,

There is is nothing 'random' about the the approach angle or amount of Doppler shift at all... It has to be known to the reciever find the pseudo-code signal sent from the bird in the first place. Otherwise, the reciever would have no idea where to find the signal... The so-called 'cold start' which can take 30 minutes. Once tracking, the VERY SMALL amount of shift from the expected frequency (the so-called apparent-Doppler) is used to calculate speed. Run this stream through the Kalmann filter and the speed estimate is better than the delta model you describe. This capability started with the SiRF STAR 3 chipset introduced in (about) 2005.

 

[Bill Roosa]

Ok, I'll admit I was in the J65 a few years ago. so what you are saying is the satelite(s) ephemerist that you get during cold start is used to let the receiver know which direction the satellites are coming from by knowing the time/date and my position (which I don't know to begin with) ????? I suspect that it needs to get at least one position first then do the BB stuff right?
Amazing stuff in any case.
as for GPS and compass agreement, I've always found that my GPS course and compass agree quite well once you account for variation, deviation and current.