Calibrating the instruments mainly means calibrating the values returned by the NMEA station. The four basic values we will need to go any further are:

Those values are the basis of all the other values, which are calculated.

Calibrating the Boat Speed
Two things that need to be achieved:
  1. Make sure to have the same speed on both tacks
  2. Make sure the value of the speed is the right one

The Boat Speed is the most likely to have to be calibrated, as it depends on the boat the transducer has been installed on (shape of the hull, flow of the water around it, etc), unlike the apparent wind angle and speed which is the same for everyone (not completely true, the rig shape has a real impact on those factors...).
To calibrate the starboard and port reading, you need to do some logging, idealy on some sausage course, where starboard and port performances are likely to be identical. Logging those data will allow you to see what the difference is between data that should be identical... Finding the difference between starboard and port, calculating the average, and then apply a factor allowing the display to go from starboard reading to average, and port reading to average, is not a problem after that. This obviously implies that your NMEA station allows you such a calibration, even if you have only one transducer...
Note:  It is interesting to consider that this difference between port and starboad tack mainly exists upwind, as the difference of pressure and water speed is related to the heel of the boat, which is very different up and downwind. Idealy, the factor to apply to the data measured on port and starboard tack should also consider the wind angle (and maybe also the heel angle, but this data is not always available, unlike the wind angle, which is always there...).

Example: Calibrating starboard and port BSP on a B&G Hydra 2000 station on a J105.

3D Here is the 3D representation of the data logged during an upwind leg, superposed with the caluculated polars. It's visible that data logged on port tack have higher values than starboard ones. This is what we want to fix.
On a 2D representation, let's represent the data, the average starboard speed, the average port speed, and the average speed.


As we see on the diagram, the goal of this part of the calibration will be:
  • On Port tack (red background) to go from 6.78 to 6.49 knots.
  • On Starboard tack (green background) to go from 6.20 to 6.49 knots.
The original value for the BSP Calibration was 3.78 for both tacks. This value represents the number of Hertz per knots. To increase the speed, decrease this value, and vice-versa.
For the port calibration :

 6.78 / 6.49 = 1.04468  		      
then the new value becomes
 3.78 * 1.04468 = 3.95

For the starboard calibration :
 6.20 / 6.49 = 0.9553  		      
then the new value becomes
 3.78 * 0.9553 = 3.61
After calibration Now, port and starboard tacks should indicate the same value for the boat speed.

The next step, now both tacks indicate the same value, is to know if this value is the right one.
A good clue is the computed value of the True Wind Direction. As the Boat Speed is a major parameter in the True Wind computation, if the True Wind Direction (computed after True Wind Angle and Heading) changes after rounding a mark, or more simply after changing the course, it means that some value used for this computation is wrong. The values are BSP, AWA, AWS. The most likely to be wrong is BSP.

The information read from the outside world are Boat Speed, Apparent Wind Angle, and Apparent Wind Speed. The way to calculate True Wind values is the following one:

  TWS = [AWS2 + BSP2 - 2*BSP*AWS*cos(AWA)]1/2
  TWA = arctg[AWS*sin(AWA) / (AWS*cos(AWA) - BSP)]
As we see, BSP is a parameter of both formula. An error in the BSP value will result in error in TWA and TWS values.

After setting the new values for the calibration(s), further logging should be processed to validate the modifications.
You know you've reached the right calibration when:

Before calibration, the wind is shifting... After calibration, the wind is steady
Before calibration, wind is shifting every time the boat heading is changing, and the true wind angle is not the same on both tacks. After calibration, steady wind through the tacks, and identical true wind angle on both tacks.

Welcome to the real world...

Idealy, this boat speed calibration has to be done with no current, and no leeway, no drifting of any kind, which is almost impossible (specially in San Francisco!).
This means that before beginning this kind of calibration after the data logged during the sailing, you have to figure out what the current was, what was the part of the leeway in the drifting, what factor to apply on the leeway calculation, depending on the waves height and shape, all this kind of things difficult to model!
Have a look to the dedicated sections for details...

2003, Oliv Cool Stuff Soft Strikes Again !
Any comment ?