Calibration

Calibration

Boat Speed
All the work in the world on target speeds and angles, or on polar tables goes out the window if the primary input – Boatspeed – is inaccurate, even by a few %.

True wind calculations and performance functions depend on this input being accurate. If boat speed is wrong, the whole system's data quality is jeopardised.

• Paddlewheels will suffer from friction and changes in physical characteristics over time. The Ultrasonic speed sensor will not suffer from the same problems.

A number of situations could point at the paddlewheel calibration being inaccurate:

• Unable to achieve target speeds at target angles
• Incorrect tide set and rate calculations

Calibrating

• The B&G three run method eliminates any changes in current during the calibration procedure.
• It is prudent to manually calibrate boatspeed using the trip log function whilst using the auto cal feature. By resetting the log at the beginning of each run, and recording the log distance run for each leg, we can deduce the error in the current speed calibration

Example:

• Current cal in Hz per Kt = 3.42
• Measured run distance = 0.73nm
• Run 1 = 0.71nm, Run 2 = 0.68nm, Run 3 = 0.70nm

It appears that the current is changing due to the difference in the first and third runs Log distance. It is safe to average run 1 and 3, and then combine with run two to get an overall average log distance:

• (Run1 + Run3)/2 = 0.705
• (Run2 + 0.705)/2 = 0.6925

Comparing the log distance to the actual distance reveals that the log is under-reading by 5.14%

(0.73 – 0.6925)/0.73 = 5.14%

Increasing cal Hz/Kt reduces boatspeed, and vice versa. Therefore by reducing the cal Hz/Kt for the above calibration by 5.14% will address the under-reading of the log.

• 3.42 - 5.14% = 3.26
  
  
• This operation should be used as a verification of the auto cal procedure. It is also very useful to complete a calibration procedure if any of the runs have to be aborted due to a necessary change of course.

Refining

• TWS, TWA, TWD, AWA, AWS, Leeway, Course, and target and polar speeds
• Once boatspeed has been verified as accurate, it should be monitored against other functions, and only adjusted if the wrong speed cannot be attributed to different sail trim, helmsman, rig settings, wind shear, crew weight, hull cleanness etc.

Wind
If wind is not calibrated accurately, then performance is next to impossible to assess.

 
If angles are inaccurate
a. The wrong polar speed will be chosen
b. The wrong target angle will be sailed to
c. The TWD will be incorrect

If speed is inaccurate
a. The wrong target speed will be chosen
b. The wrong polar speed will be given
c. The wrong sails may be selected

In essence, the calibration of wind is as vital as selecting the correct sail for the next upwind leg.

AWA
Installation offsets, mast twist, pressure, and flow effects around the mast affect wind angles. An offset is all that is required to correct for tack-to-tack errors for AWA. The perfect scenario for calibrating AWA is in non-tidal flat water, with stable conditions, i.e. no building sea breeze, a full crew, racing sails, a tuned rig, and a well calibrated boatspeed.

Using identical trim, crew weight distribution, and helmsman, complete half a dozen tacks with the helm concentrating purely on boatspeed. A target speed (not necessarily from the targets table) should be consistently achieved without changing trim, and without using any reference to AWA. The navigator or tactician should monitor the AWA from tack to tack, and average each tack. Any difference between Pt and Stbd tack can be corrected in the MHU offset.

If Stbd is greater than Pt then subtract half the difference from the MHU offset, and vise versa.

Once a good AWA is achieved, do not change it unless necessary (rig has been removed from boat, new wind sensor put in place). Any differences seen from here on in may be attributable to effects such as wind shear and current.

AWS
Do not alter the AWS calibration values. This is a wind tunnel based factory setting.

TWA
A TWA calibration matrix is utilised for corrections upwind, reaching, and downwind across the TWS ranges.

Two methods of calibrating TWA are available, either monitoring TWD from tack to tack or gybe to gybe, or using the ships' compass to verify the angles the yacht is tacking or gybing through.

1. If an error is seen in TWD, then the following rule applies
   If TWD is lifting you tack to tack then TWA is reading too wide
   If TWD is heading you tack to tack then TWA is reading too narrow
   
2. If according to the ships compass you are tacking through an angle different than the sum of the TWA’s on each tack (Port TWA + Stbd TWA) then the following rule applies
   
   If the tack angle < the sum of the TWA’s, the TWA is reading too wide
   If the tack angle > the sum of the TWA’s, the TWA is reading too narrow

If the TWA is reading wide, then half the error must be subtracted from the correction table. If the TWA is reading narrow, then add half the error to the TWA table.

A Worked Example:
Your TWD is telling you that you are being lifted by 10 degrees from tack to tack (or the sum of your TWA is 10 degrees less than the angle you are tacking through), and you know this is wrong. This is in 5 kts of TWS. When you go into the TWA correction table for 5 Kt upwind, you see that a value of –1.5 degrees is already entered. To correct for the error you are seeing, a further 5 degrees needs to be subtracted so a total correction of –6.5 degrees is applied.

* Tip for TWA calibration:

Typically, when graphed, a TWA table for upwind should have a curve like the following:

If it has only been possible to achieve reliable values for the TWS 5, 10 and 20 range and the rest of the table left blank then the table would look as follows: 

It is clear to see that the 15, 25, and 30 kt values can be accurately guessed at as opposed to being left with zero values.

Calibrating and refining TWA is an ongoing process. Differences in TWA can been seen from morning to afternoon races as sea breezes develop, changes in gradient conditions etc. TWA should be checked at the beginning of every race, any necessary changes made, and the conditions logged for future reference.

TWS
TWS errors are seen from sailing upwind to downwind. This is most noticeable on Mast Head boats, however all yachts are affected to some degree. This is due to the acceleration of the airflow over the top of the mast and around the sails from the high-pressure areas to the low-pressure areas. The introduction of the Vertical Mast Head unit has gone some way to solving this, however calibration still may be necessary.

As a rule of thumb, it is safe to put a correction of –10% into the table before first sailing. Monitoring the change in TWS from close hauled to flat running will enable further refinement of this calibration value.

* Note

From Feb 2003: Hydra and Hercules systems have the additional functionality of re-calculation of Apparent wind speed and angle, and faster rates of calculation. The Apparents will be calculated from the TWA and TWS, after they have been submitted to calibration. This in effect removes further errors from the Apparents that are only visible in the TWS and TWA. Please contact your nearest B&G dealer for further information on this.

Compass
Once again, as a primary input, accurate compass calibration is important. Of the range of sensors available from B&G the Halcyon Gyro Stabilised compass is the most accurate and responsive, however if this is not set up correctly then almost no benefit will be achieved.

Every compass on a yacht will suffer from local deviation effects, due to local magnetic interferences such as the engine, electrical cabling, speakers, computers - almost any ferrous or magnetic object. These effects will vary at different points of the compass, for example only a degree or two at a due north heading, and 5 or 6 degrees at an easterly heading.

If a compass is not correctly calibrated
1. The wrong course will be steered
2. TWD will be incorrect
3. VMC calculations will be incorrect
4. It will not be possible to calibrate TWA correctly
5. Tide set and rate will be incorrect
6. Dead reckoning will be inaccurate.

B&G has an algorithm built in that can sense these effects during an electronic ‘swing’. An offset can be input to the system to correct for slight heading errors due to installation.

Calibration
It is straight forward to ‘swing’ the compass. Ideal conditions for completing this is in flat water, well away from large magnetic interferences such as container terminals, radar/radio installations, and large underwater/overhead cabling.

All equipment on the yacht should be in its normal position for this operation. It is not unusual for yachts to have an anchor and chain in the bows during a ‘swing’ and removing the anchor and chain prior to racing. This will affect the quality of the compass data if the sensor is mounted within 10 to 20 feet of the bows (around the mast).

A slow and accurate rate of turn is essential, and should this be disturbed in any way, cancel the turn, and start again. For the sake of an extra 6 to 10 minutes for the turn, the swing is well worth repeating.

Once a PASS has been given for the swing, it is necessary to offset the compass to remove installation errors. It is not acceptable to just alter the compass offset to match a hand-bearing compass or a binnacle compass, as these compasses may have their own errors.

The only reliable method is to create transits on a chart with easily identifiable landmarks, and derive their exact magnetic bearings. If time is limited then it is possible to check the offset on just one bearing, however it is strongly recommended to complete a transit run on all four points of the compass. This will provide verification of the quality of the swing, and provide a more accurate offset to enter into the B&G system.

By creating a table of actual bearings, and logging the heading of the yacht on those transits, it is possible to generate an offset that will provide the least amount of error in the heading.

As can be seen below, if only one transit is used, the potential error in heading is up to 3 degrees. By graphing the offsets it can be seen that an offset of -1 degree is the most appropriate. This then reduces any potential heading error down to less than 2 degrees.