I recently had my knotmeter overhauled by the manufacturer. I had shorted it out when working on my wind speed indicator, (s--- happens). After installing the rebuilt knot meter, I calibrated it against my handheld gps. My question is, why do the gps and knotmeter vary so much during sailing? They will sometimes be 3/4 knot or more different and sometimes they read the same. Is one more accurate than the other?
knot meter vs. gps
- Thread starter Bill C320 Papillon
- Start date
- Status
GPS does speed over ground
Knot meter does speed in water - any currents or slide slippage will effect the knot meter. Example if you sail at 4 kts against a 3 kt current you will be going 4 kts against the water but only 1 kt over the ground.
Knot meter does speed in water - any currents or slide slippage will effect the knot meter. Example if you sail at 4 kts against a 3 kt current you will be going 4 kts against the water but only 1 kt over the ground.
Differences
The knot meter measures the boat speed through the water and the GPS is measuring the boat speed over the ground. There is a slight delay in the GPS signal because the signal has to come from a triangulation of satilites.
The knot meter measures the boat speed through the water and the GPS is measuring the boat speed over the ground. There is a slight delay in the GPS signal because the signal has to come from a triangulation of satilites.
Easiest way to explain
Here is the simple answer. The GPS measures speed traveled over the earths surface. The knotmeter measures speed traveled through the water. If you have a 2 kt current you are traveling into, your knotmeter would read two knots faster than your GPS. The little paddle on the knotmeter doesn't know there is a current. It only knows how fast the water is traveling past it. PS. There are lots of other factors that affect both units. Both are accurate if you understand what they mean. Hope this helps Ward
Here is the simple answer. The GPS measures speed traveled over the earths surface. The knotmeter measures speed traveled through the water. If you have a 2 kt current you are traveling into, your knotmeter would read two knots faster than your GPS. The little paddle on the knotmeter doesn't know there is a current. It only knows how fast the water is traveling past it. PS. There are lots of other factors that affect both units. Both are accurate if you understand what they mean. Hope this helps Ward
I get it but---
I sail on an inland lake where there aren't appreciable currents. I hadn't thought about slippage occurs, especially as the wind increases. Anyway, was it appropriate to use the gps for calibration (considering speed over land vs. speed thru water and satellite dely etc.) or is there a better way to set the knot meter? Thanks Bill
I sail on an inland lake where there aren't appreciable currents. I hadn't thought about slippage occurs, especially as the wind increases. Anyway, was it appropriate to use the gps for calibration (considering speed over land vs. speed thru water and satellite dely etc.) or is there a better way to set the knot meter? Thanks Bill
Out here most harbors have a measured
nautical mile on a main channel somewhere, think it is marked by white square or triangle with black outline. Check around at your marina, I bet someone has a measured mile somewhere on the lake.
nautical mile on a main channel somewhere, think it is marked by white square or triangle with black outline. Check around at your marina, I bet someone has a measured mile somewhere on the lake.
Isn't there a different reason?
I don't have GPS for my boat so I may not be qualified to answer; but, doesn't the GPS work by measuring the speed over some increment? What you would be reading at any point in time would therefore be an average over some time increment, ending at the point in time when you are reading it. Your knot meter is measuring speed instantaneously as the paddle wheels turn. The instrument has some tolerance for variability that can be set depending upon your preference, but it probably adjusts more rapidly than GPS. I would trust GPS measured speed over a longer interval, but not as an instantaneous measurement.
I don't have GPS for my boat so I may not be qualified to answer; but, doesn't the GPS work by measuring the speed over some increment? What you would be reading at any point in time would therefore be an average over some time increment, ending at the point in time when you are reading it. Your knot meter is measuring speed instantaneously as the paddle wheels turn. The instrument has some tolerance for variability that can be set depending upon your preference, but it probably adjusts more rapidly than GPS. I would trust GPS measured speed over a longer interval, but not as an instantaneous measurement.
GPS Delay
While there is a slight delay with GPS, remember, the signal is traveling at the speed of light and the coorilation between satellites is measured in nanoseconds.
While there is a slight delay with GPS, remember, the signal is traveling at the speed of light and the coorilation between satellites is measured in nanoseconds.
WAAS makes it instantaneous
Scott, I think your impression of GPS being slow may be from earlier versions--the Current WAAS (Wide-Area Augmentation System) software in newer gps's can work out your correct speed as "instantaneously" as the knot meter (remember that the knot meter, like any electrical device, is also measuring intervals, but very minute ones; data from the satellites is available in a constant stream and only limited by how fast your ground unit can analyze it). The bottom line is, I'm pretty sure a newer model gps with WAAS is as accurate as you're going to get, probably more accurate than any electromechanical knot reading device.
Scott, I think your impression of GPS being slow may be from earlier versions--the Current WAAS (Wide-Area Augmentation System) software in newer gps's can work out your correct speed as "instantaneously" as the knot meter (remember that the knot meter, like any electrical device, is also measuring intervals, but very minute ones; data from the satellites is available in a constant stream and only limited by how fast your ground unit can analyze it). The bottom line is, I'm pretty sure a newer model gps with WAAS is as accurate as you're going to get, probably more accurate than any electromechanical knot reading device.
GPS satellites transmit every 30 seconds
so it's up to the receiver to decide what period to use for velocity calculations. I don't know if there are receivers that will let the user specify the period. Anybody know? I tend to watch the log for speed, the compass for course, and the GPS for ETA. When not engrossed in those, I sometimes keep an eye out for other boats, ships, etc. Cheers, Bob s/v X SAIL R 8
so it's up to the receiver to decide what period to use for velocity calculations. I don't know if there are receivers that will let the user specify the period. Anybody know? I tend to watch the log for speed, the compass for course, and the GPS for ETA. When not engrossed in those, I sometimes keep an eye out for other boats, ships, etc. Cheers, Bob s/v X SAIL R 8
Bob,
I don't think that's true. The GPS satellites are sending a constant signal. At least that's my understanding of the system. It's just Loran, in the sky. The receivers have to have a set time to update the signal or you would see gibberish on the screen. Updates would be too fast. Mine updates once a second. When the boat moves the readout changes. It can't do that if a signal only arrived once every 30 seconds. I suggest that what you are seeing is that your receiver is set to give an average speed at 30 second intervals. That is user changeable in your setup menu.
I don't think that's true. The GPS satellites are sending a constant signal. At least that's my understanding of the system. It's just Loran, in the sky. The receivers have to have a set time to update the signal or you would see gibberish on the screen. Updates would be too fast. Mine updates once a second. When the boat moves the readout changes. It can't do that if a signal only arrived once every 30 seconds. I suggest that what you are seeing is that your receiver is set to give an average speed at 30 second intervals. That is user changeable in your setup menu.
knotmeter
simple. the knotmeter and gps are measuring two different things. the knotmeter is measuring the speed of the hull through the water, the gps the speed of the boat over the surface of the earth. only with no current or wind will they agree.
simple. the knotmeter and gps are measuring two different things. the knotmeter is measuring the speed of the hull through the water, the gps the speed of the boat over the surface of the earth. only with no current or wind will they agree.
Get an average and compare
Choose a location where current, tide, etc. influences are minimal, and where the wind speed and direction is pretty constant. Sail the boat on a broad reach with optimum trim, and wait until both units give you constant read-outs. You might do this over, say, 4 separate periods of 30 seconds each. Now tack the boat and sail on a reciprocal course, following the same procedure. Ideally, you will be getting identical readings from each source, or if there is a discrepancy, it will be easier to spot and figure out. For example, if the GPS is reading a fairly constant 5.6 but the knotmeter shows 0.8 difference from one tack to the other, it is likely to be due to sending unit location on the hull. If the knotmeter is constant, but the GPS isn't, then there is a current or a tidal motion influencing the results.
Choose a location where current, tide, etc. influences are minimal, and where the wind speed and direction is pretty constant. Sail the boat on a broad reach with optimum trim, and wait until both units give you constant read-outs. You might do this over, say, 4 separate periods of 30 seconds each. Now tack the boat and sail on a reciprocal course, following the same procedure. Ideally, you will be getting identical readings from each source, or if there is a discrepancy, it will be easier to spot and figure out. For example, if the GPS is reading a fairly constant 5.6 but the knotmeter shows 0.8 difference from one tack to the other, it is likely to be due to sending unit location on the hull. If the knotmeter is constant, but the GPS isn't, then there is a current or a tidal motion influencing the results.
He's on a Lake
and unless there's one big drain at the other end of the lake, there's probably no current. Easiest way is to combine the good ideas discussed above. Find a measured mile or half mile or make one yourself: i.e., if there is no nautical chart for your lake, find a road map and measure off a mile along the shoreline - point A to B, translate it to a nautical mile and then sail or motor along that line. Your GPS should say it's a mile (nautical if so calibrated) and you can set up a quickie little time on distance spreadsheet to tell you how many minutes it will take to travel that mile in knots. BTW - have you calibrated BOTH the GPS and the knot meter in KNOTS? Stu
and unless there's one big drain at the other end of the lake, there's probably no current. Easiest way is to combine the good ideas discussed above. Find a measured mile or half mile or make one yourself: i.e., if there is no nautical chart for your lake, find a road map and measure off a mile along the shoreline - point A to B, translate it to a nautical mile and then sail or motor along that line. Your GPS should say it's a mile (nautical if so calibrated) and you can set up a quickie little time on distance spreadsheet to tell you how many minutes it will take to travel that mile in knots. BTW - have you calibrated BOTH the GPS and the knot meter in KNOTS? Stu
Not necessarily, Stu...
Most lakes are dammed rivers, and the river still flows through 'em. Creeks also feed 'em, which also have current. As an example, there's enough current flowing through Lake Lanier north of Atlanta to completely cleanse the lake approximately every 72 hours. The amount of current in the main channel varies, depending upon whether the sluice gates in the dam are open...which happens a couple of times a day. So technically there IS a drain at one end. Rainfall in the watershed feeding the river and streams coming into a lake can also influence the amount current going through it. So while there's not a STRONG current in most places on lakes, there's still some...enough that a measured mile is the only way to get an accurate measurement of speed over the ground.
Most lakes are dammed rivers, and the river still flows through 'em. Creeks also feed 'em, which also have current. As an example, there's enough current flowing through Lake Lanier north of Atlanta to completely cleanse the lake approximately every 72 hours. The amount of current in the main channel varies, depending upon whether the sluice gates in the dam are open...which happens a couple of times a day. So technically there IS a drain at one end. Rainfall in the watershed feeding the river and streams coming into a lake can also influence the amount current going through it. So while there's not a STRONG current in most places on lakes, there's still some...enough that a measured mile is the only way to get an accurate measurement of speed over the ground.
GPS is transmitted continuously
Fred is bang on. Your receiver processes the information one time per second. Some of the newer ones are a little faster according to the tech talk.
Fred is bang on. Your receiver processes the information one time per second. Some of the newer ones are a little faster according to the tech talk.
I thought that the timing data
was only available every 30 seconds. I could be wrong but that means that the speed would be figured over a 30 second time period. WAAS only makes the data more accurate by adding a position correction factor. It does not speed it up.
was only available every 30 seconds. I could be wrong but that means that the speed would be figured over a 30 second time period. WAAS only makes the data more accurate by adding a position correction factor. It does not speed it up.
Knotmeters even vary with different tacks
The knot meter on my SanJuan 24 varies depending on which tack- granted there is probably a faster tack but not as much as the knotmeter shows ( about 2 knts!!) Have never been able to figure out why. My first theory was that the paddlewheel was not at right angles to the direction of travel so I reinstalled it -no difference. The paddlewheel is about 6 inches on the port side of the keel and about 8 inches back of the point where the keel meets the hull at the forward part. If I remember corectly it was port tack that shows faster.
The knot meter on my SanJuan 24 varies depending on which tack- granted there is probably a faster tack but not as much as the knotmeter shows ( about 2 knts!!) Have never been able to figure out why. My first theory was that the paddlewheel was not at right angles to the direction of travel so I reinstalled it -no difference. The paddlewheel is about 6 inches on the port side of the keel and about 8 inches back of the point where the keel meets the hull at the forward part. If I remember corectly it was port tack that shows faster.
Filtering Delay
I also understand that GPS positions are computed very quickly (on the order of a second.) But because a speed is calculated as the difference between two points, if the time is quick then the distance traveled is short, and even the WAAS accuracy of 3 to 10 feet results in appreciable speed errors. Luckily, if the readings are averaged over time, they become extremely, and absolutely accurate. The averaging, or filtering can be set up on a recent GPS, with the user providing a filteing time - for instance 30 seconds. So the GPS gives a very accurate speed over ground for 15 to 30 seconds ago. There are a number of factors that can cause accuracy in a knotmeter. Flow over the paddle wheel can vary with heel, tack, and even speed. Temperature, vibration, and calibration could also cause errors. So the absolute accuracy of a knotmeter is probably around 0.5 to 1.0 knots. But the the sensitivity and speed of the knotmeter is excellent. It can measure small differences in speed very quickly. So it is the best instrument for understanding the impact of changing winds or sailing tactics on speed. If you understand the filtering of your GPS, it is an excellent tool for accurately calibrating your knotmeter, provided there is no current! Lakes can easily have 1 knot currents from water flow, and wind driven flow or sieches. David Lady Lillie
I also understand that GPS positions are computed very quickly (on the order of a second.) But because a speed is calculated as the difference between two points, if the time is quick then the distance traveled is short, and even the WAAS accuracy of 3 to 10 feet results in appreciable speed errors. Luckily, if the readings are averaged over time, they become extremely, and absolutely accurate. The averaging, or filtering can be set up on a recent GPS, with the user providing a filteing time - for instance 30 seconds. So the GPS gives a very accurate speed over ground for 15 to 30 seconds ago. There are a number of factors that can cause accuracy in a knotmeter. Flow over the paddle wheel can vary with heel, tack, and even speed. Temperature, vibration, and calibration could also cause errors. So the absolute accuracy of a knotmeter is probably around 0.5 to 1.0 knots. But the the sensitivity and speed of the knotmeter is excellent. It can measure small differences in speed very quickly. So it is the best instrument for understanding the impact of changing winds or sailing tactics on speed. If you understand the filtering of your GPS, it is an excellent tool for accurately calibrating your knotmeter, provided there is no current! Lakes can easily have 1 knot currents from water flow, and wind driven flow or sieches. David Lady Lillie
Well, the Coast Guard wrote the spec
(see link below). The spec calls for 1500 bits of data transmitted at 50 bits per second.... Interestingly enough, the satellite actually tells the receiver how many 1.5 second epochs "have occurred since the transition from the previous week". And it actually does this every 6 seconds (at the beginning of each of the 5 subframes) so a receiver could update its location every 6 seconds. Anything it does faster than 6 seconds it is doing internally - such as updating distance to a waypoint - mine does that in hundredths of a mile at ranges less than 10 miles. But that has to be using the velocity data derived from the most recent time cut from the satellite. For accuracy, I should point out that the full 'navigation message' requires 25 cycles X 30 seconds, because sub-frames 4 & 5, which contain the almanac data and other stuff which isn't needed all the time, are sent over 25 cycles. And since this thread has to do with accuracy, with a GPS up to 50 meters is reasonable, so if I am sailing at 5 knots, in 6 seconds I have traveled 1852 meters * 5 miles/hour * 1/600 hour = 15 meters OK, even if the GPS is accurate to 5 meters (which is sometimes is, but you and it don't know when that is unless you and it are standing on a benchmark) that still gives an accuracy of +/- 33% So for moving objects, I think the full 30 seconds would be appropriate. That gives 75 meters traveled and 5 meters accuracy = +/- 7% Hmmm, that Cheeto isn't looking toooooo bad in the accuracy game ;D For a stationary object, one might hope that repeated satellite fixes would increase accuracy, but I don't think so - the receiver doesn't have any way to know which is a more accurate position and which isn't, so I would expect only a modest 'learning' curve, if in fact the receiver trys to 'learn' and doesn't just display the current value. Cheers, Bob s/v X SAIL R 8
(see link below). The spec calls for 1500 bits of data transmitted at 50 bits per second.... Interestingly enough, the satellite actually tells the receiver how many 1.5 second epochs "have occurred since the transition from the previous week". And it actually does this every 6 seconds (at the beginning of each of the 5 subframes) so a receiver could update its location every 6 seconds. Anything it does faster than 6 seconds it is doing internally - such as updating distance to a waypoint - mine does that in hundredths of a mile at ranges less than 10 miles. But that has to be using the velocity data derived from the most recent time cut from the satellite. For accuracy, I should point out that the full 'navigation message' requires 25 cycles X 30 seconds, because sub-frames 4 & 5, which contain the almanac data and other stuff which isn't needed all the time, are sent over 25 cycles. And since this thread has to do with accuracy, with a GPS up to 50 meters is reasonable, so if I am sailing at 5 knots, in 6 seconds I have traveled 1852 meters * 5 miles/hour * 1/600 hour = 15 meters OK, even if the GPS is accurate to 5 meters (which is sometimes is, but you and it don't know when that is unless you and it are standing on a benchmark) that still gives an accuracy of +/- 33% So for moving objects, I think the full 30 seconds would be appropriate. That gives 75 meters traveled and 5 meters accuracy = +/- 7% Hmmm, that Cheeto isn't looking toooooo bad in the accuracy game ;D For a stationary object, one might hope that repeated satellite fixes would increase accuracy, but I don't think so - the receiver doesn't have any way to know which is a more accurate position and which isn't, so I would expect only a modest 'learning' curve, if in fact the receiver trys to 'learn' and doesn't just display the current value. Cheers, Bob s/v X SAIL R 8
- Status