Wind Instrument Calibration

[Dockhead]

I have always used this method to calibrate the alignment of my wind instruments:


1. Wait for a really calm day.
2. Motor directly into the wind at max speed (9.5 knots on my boat).
3. Direction for "directly into the wind" is determined by stopping the boat and watching the windex while rotating the boat with the bow thruster (or motoring very slowly with the helm hard over).
4. Log indicated AWA.
5. Correct offset based on this. If AWA is +5 degrees then correction is -5 degrees.
6. Repeat to check that the correction worked.



Anyone have a different or better method?


More difficult is to correct apparent wind for heel angle and upwash. Short of using an H5000, does anyone have a way to do this? It looks like some software programs (Expedition, OpenCPN Tactics Plugin) can do this, but don't allow for pushing this data back to the network (discussed in another thread). But if we could at least KNOW this information, even if we don't get it on the network, that would still be very useful.

[ronstory]

With my CV7 step 1 was make sure it was mounted as straight as I could get it on the masthead.

For speed calibration, I compared to two other docmates whirly things while on a conference call. It was like "Let meet at the tree fort!" with walkie-talkies from my youth. ;^) They all were within a knot or so, but the CV7 seems to do better at really low wind speeds.

For AWA, I just noted angles coming out of tacks over time and the jotting down any difference in port vs starb angle. I did this on and off for 2-3 months of sailing and noted that my tack angles constantly varied 1-2 degrees higher towards starboard. I called that more good enough to not mess with calibration.

YMMV and good luck.

[Dockhead]

Cheers.

[ChrisseH]

Originally Posted by Dockhead

I have always used this method to calibrate the alignment of my wind instruments:


1. Wait for a really calm day.
2. Motor directly into the wind at max speed (9.5 knots on my boat).
3. Direction for "directly into the wind" is determined by stopping the boat and watching the windex while rotating the boat with the bow thruster (or motoring very slowly with the helm hard over).
4. Log indicated AWA.
5. Correct offset based on this. If AWA is +5 degrees then correction is -5 degrees.
6. Repeat to check that the correction worked.



Anyone have a different or better method?


More difficult is to correct apparent wind for heel angle and upwash. Short of using an H5000, does anyone have a way to do this? It looks like some software programs (Expedition, OpenCPN Tactics Plugin) can do this, but don't allow for pushing this data back to the network (discussed in another thread). But if we could at least KNOW this information, even if we don't get it on the network, that would still be very useful.

The way you are calculating and adjusting the AWA offset error is correct. If you have a very well calibrated speedlog and perfectly calm and slack conditions you could (at least in theory) check also whether your AWS reading is correct. The challenge is that you might still have a bit of airflow at your mast top despite that it is calm at sea level. You could basically compensate for this if you motor the same short route in both directions, but be aware of that even then (the tiny) wind might easily change.

In regard to the heeling compensation there is no simple way to do this. Either you test the heeling dependence in a well calibrated wind tunnel or then you use two similar sensors next to each other in a free air test and tilt one of them. In both alternatives the test should be done at different wind speeds, which is especially challenging in the free air test. The free air test alternative require a lot of data and careful validation and will always be less accurate than a proper wind tunnel test.

The commonly used heel compensation model for wind is only a rough model suitable for mechanical wind sensors. I am quite sure that B&G for instance has tested the heel dependence of their sensors in a wind tunnel.

[Dockhead]

Originally Posted by ChrisseH

The way you are calculating and adjusting the AWA offset error is correct. If you have a very well calibrated speedlog and perfectly calm and slack conditions you could (at least in theory) check also whether your AWS reading is correct. The challenge is that you might still have a bit of airflow at your mast top despite that it is calm at sea level. You could basically compensate for this if you motor the same short route in both directions, but be aware of that even then (the tiny) wind might easily change.

In regard to the heeling compensation there is no simple way to do this. Either you test the heeling dependence in a well calibrated wind tunnel or then you use two similar sensors next to each other in a free air test and tilt one of them. In both alternatives the test should be done at different wind speeds, which is especially challenging in the free air test. The free air test alternative require a lot of data and careful validation and will always be less accurate than a proper wind tunnel test.

The commonly used heel compensation model for wind is only a rough model suitable for mechanical wind sensors. I am quite sure that B&G for instance has tested the heel dependence of their sensors in a wind tunnel.

That makes sense. Unfortunately.


There is a racing version of the CV7, though. Surely they've figured out the heeling compensation for it? The manufacturer claims that the SHAPE of the sensor "partially compensates inclination".


And old test of the WSO100 in Practical Sailor said that this ultrasonic wind sensor was "virtually immune" to heeling. https://www.practical-sailor.com/mar...nd-sensor-test


So maybe this is less needed, than with mechanical transducers?

[ChrisseH]

Originally Posted by Dockhead

That makes sense. Unfortunately.


And old test of the WSO100 in Practical Sailor said that this ultrasonic wind sensor was "virtually immune" to heeling. https://www.practical-sailor.com/mar...nd-sensor-test

So maybe this is less needed, than with mechanical transducers?

Read through the Practical Sailor article. I would not count much on that test. Making wind tunnels is a whole science. I have seen both good and bad ones. Especially the home built small are challenging as the flow is not laminar because of the friction the walls are causing. And when you put objects in the flow they accelerate the flow locally. This means that the wind tunnel needs to be surprisingly big before it becomes useful. And as I said earlier, it is very important to make the heel test for a variety of air speeds.

[Paul L]

To calibrate the AWA with a single correction, why not motor on a completely calm day. If the AWA is not 0, then its reading is the correction.

[Dockhead]

Originally Posted by Paul L

To calibrate the AWA with a single correction, why not motor on a completely calm day. If the AWA is not 0, then its reading is the correction.

Well, that's exactly what I do to calibrate for offset -- as described in Post #1 in this thread. That's how I know my sensor is misaligned by 2 degrees. Another question however is how to correct this error in the network. I can do this calibration either in one of my Triton MFDs, or in one of my Zeus MFDs. Unfortunately setting this offset in one of the MFDs is not recognized in the others. This is maddening and one of my jobs next week is to figure out how to make it work across the network. If I can't get it to work I might have to climb the mast an correct the alignment mechanically. Or rewire the CV7 so that I get 0183 data from it which I can correct in the device which converts it to N2K (presently converted to N2K at the masthead with a WindyPlug).




But offset is only one element of calibration. There is also:


1. Linearity
2. Heel
3. Upwash


I can't calibrate any of that with the equipment I have now. Perhaps linearity is already good enough that I can ignore this. It think Heel is the big one, and perhaps it doesn't need calibration either to get reasonably usable true wind. We shall see.

[wingssail]

B&G recommends a calibration sequence which includes multiple readings and taking half of the error out when correcting each time. These are done sailing, not motoring in calm conditions.

More modern versions have a lot of software assist for this process.

Compensation for heel angle is software dependent.

To reduce upwash a tall vertical spar like shown below is helpful. (Note that the transducer is not vertical, it is tilted to compensate for the heel of the vessel.)

[Dockhead]

Originally Posted by wingssail

B&G recommends a calibration sequence which includes multiple readings and taking half of the error out when correcting each time. These are done sailing, not motoring in calm conditions.

More modern versions have a lot of software assist for this process.

Compensation for heel angle is software dependent.

To reduce upwash a tall vertical spar like shown below is helpful. (Note that the transducer is not vertical, it is tilted to compensate for the heel of the vessel.)

How do you do it, Fred? I know you're a successful racer, so you must care about true wind and so you must have a solution that works. I'm particularly glad you've come into this thread; I'll be grateful for your advice.

Yes, the classical approach would be to use one of those $4000 B&G mechanical ones with the long carbon spar, and calibrate it with the $2400 H5000 computer.

I might end up doing it that way myself, but that's not only a lot of money, but there are a couple of bridges I like to get under where an extra 1.5 meters at the masthead might make dicey (I have 23 meters as it is).

I'm using a CV7 ultrasonic wind instrument which is much less subject to heel error than mechanical ones. I'm hoping I will be able to get the rest of the network working well enough that I can continue using this.

How do you calibrate your instruments, and how good is your true wind data? You use an H5000 or H2000 or something like that? Do you use Expedition?

[wingssail]

Originally Posted by Dockhead

Well, that's exactly what I do to calibrate for offset -- as described in Post #1 in this thread. That's how I know my sensor is misaligned by 2 degrees. Another question however is how to correct this error in the network. I can do this calibration either in one of my Triton MFDs, or in one of my Zeus MFDs. Unfortunately setting this offset in one of the MFDs is not recognized in the others...

This is surprising. I know that Hercules (such as 5000) does this correction in the main computer and the corrected value is available on that channel at all MFD's or displays.

This has been this way in Hercules for many years. My B&G Hercules 390 had it, as did my Hercules 590 and my Hercules 2000 and I have calibrated Hercules 5000 systems on other boats as well. So maybe Zeus and Triton do not have this capability.

[Dockhead]

Originally Posted by ChrisseH

Read through the Practical Sailor article. I would not count much on that test. Making wind tunnels is a whole science. I have seen both good and bad ones. Especially the home built small are challenging as the flow is not laminar because of the friction the walls are causing. And when you put objects in the flow they accelerate the flow locally. This means that the wind tunnel needs to be surprisingly big before it becomes useful. And as I said earlier, it is very important to make the heel test for a variety of air speeds.

That makes sense; thank you.

[wingssail]

Originally Posted by Dockhead

How do you do it, Fred? I know you're a successful racer, so you must care about true wind and so you must have a solution that works. I'm particularly glad you've come into this thread; I'll be grateful for your advice.

Yes, the classical approach would be to use one of those $4000 B&G mechanical ones with the long carbon spar, and calibrate it with the $2400 H5000 computer.

I might end up doing it that way myself, but that's not only a lot of money, but there are a couple of bridges I like to get under where an extra 1.5 meters at the masthead might make dicey (I have 23 meters as it is).

I'm using a CV7 ultrasonic wind instrument which is much less subject to heel error than mechanical ones. I'm hoping I will be able to get the rest of the network working well enough that I can continue using this.

How do you calibrate your instruments, and how good is your true wind data? You use an H5000 or H2000 or something like that? Do you use Expedition?

I have an old Hercules 590 system installed and a newer Hercules 2000 system on the bench right now which I hope to install this summer. No way can I justify an entirely new system which would be several thousand dollars, (maybe 10's of thousands of dollars) but I get the most out of my existing system and I expect to do so with the 2000 system which I got used.

I follow the calibration procedure in the B&G manual which involves first boat speed calibration, next compass calibration, then the TWA calibration which compares the Magnetic Wind Direction (TWA) on either tack and correction factors entered into the Master Computer or Performance Computer if you have it.

I have a new vertical spar but it is not mounted yet.

How accurate it is? Not perfect. The H5000 systems I have used as navigator on other boats are much better.

However, I care less about the exact true wind direction than I do about wind shifts and the compass heading on each tack. Even if the true wind is off by a few degrees (as revealed when you tack and it changes) the changes of wind direction while on a tack are quite exact. If you see a lift or a knock, it is pretty damn close.

We are interested in wind shifts, and favored tack or gybe, both when sailing upwind and downwind. The instruments give us fine wind shift information which we use tactically. For the big picture we look at the track on the plotter and from that we can see which direction the wind is blowing. It is surprising how closely that comes to what we are seeing on TWA on the B&G.

[Dockhead]

Originally Posted by wingssail

. . . We are interested in wind shifts, and favored tack or gybe, both when sailing upwind and downwind. The instruments give us fine wind shift information which we use tactically. For the big picture we look at the track on the plotter and from that we can see which direction the wind is blowing. It is surprising how closely that comes to what we are seeing on TWA on the B&G.

OK, that makes sense.


One of the biggest (of my several) problems I have now is not being able to determine the favored tack when sailing upwind. In a long distance (multi-day) upwind leg, playing the wind shifts correctly is the whole difference between success and failure. And this is true even cruising keenly.


If my mark or waypoint is within 15 or even 20 degrees of the true wind direction, my instruments are useless -- I can only figure it out empircally, by making a series of useless tacks. It gets really old doing this over and over again on a long upwind leg even if my boat is easy to tack. I long for accurate information, even accurate within 10 degrees. I would kill for true wind direction accurate to 5 degrees.


Downwind troubles me less. I have a big assy on a "cheater pole" (from an old TP52) and don't sail that deep anyway. It's so much trouble to gybe that I will not do it for a short term wind shift.



I wouldn't necessarily have to have the accurate true wind data on my instruments (although that would be awfully nice). If I had it at least on my ship's computer, I could get it from there onto a tablet, and this would be enough for making tactical decisions. So perhaps the OpenCPN Tactics plug-in is going to be adequate for my purposes. We shall see. I just hope I don't end up buying an H5000 here in Europe, where it is 1.5x as expensive as it would have been in the States which I just left.

[wingssail]

Originally Posted by Dockhead

...Yes, the classical approach would be to use one of those $4000 B&G mechanical ones with the long carbon spar, and calibrate it with the $2400 H5000 computer...Do You use an H5000 or H2000 or something like that? Do you use Expedition?

We use OpenCPN for navigation but not for wind information or tactical information while sailing. For that we use the B&G displays directly and we also have a "Sail Racer" app on an android tablet which receives raw data from the B&G via blue tooth and provides most of the tactical information we use on deck while racing (and I use it while cruising, just to play around). It's a bit fiddly but it does not require a below decks PC like Expedition and costs a lot less.