DeWiggler™ is an automated calibration tool
The most important mission of an instrument system is correctly reporting wind direction (see Importance of true wind). If lifts or headers occur every time you tack or gibe, there is a problem with your instrument calibration. DeWiggler is the solution to this problem.
We call Wiggle the change in wind direction before to after a tack. DeWiggler is an application and process for determining how much wiggle occurs when tacking and jibing, then recommending adjustments to the instrument system’s calibrations.
Wind direction is derived from apparent wind, boatspeed and heading. Wiggle is reduced by correctly calibrating these three inputs.
There is also another important environment variable that greatly affects performance; current (see Sailing in current). We call Current Wiggle the change in current calculation with heading. Current is derived from boatspeed, heading and GPS course/speed over ground (COG/SOG).
Notice that boatspeed and heading are common to both wind and current wiggle. Therefore, boatspeed and heading should be the first sensors to calibrate. Boatspeed is fairly straightforward; timed runs over a known distance. However compass calibration is more involved.
How DeWiggler works
- DeWiggler is now included in OckamSoft 5. If you don’t already have OS5, you will need to install it first.
- Set up an email client (e.g. Outlook) on the same machine so DeWiggler can send results for analysis.
- Record a Speed & Heading test, then as many tacks and jibes in as great a wind range as possible.
- Email the log files to Ockam for processing.
- When the analysis comes back, drag the recommendations file to DeWiggler. Here is a sample report.
- Review your cal changes and set the cals into the screws.
- Do another set of tests for confirmation, and repeat the process if necessary.
Robert Gordenker (Time Machine, a champion J35 out of Ann Arbor) has been DeWiggled. In his blog, he nicely describes the Care and Feeding of Racing Instruments.
Here is his feedback on DeWiggler
The numbers we are getting now are much more stable than anything we have ever seen before. We are sailing in a Windward/Leeward regatta this weekend and then in the North American Championships (also W/L) next weekend. For the first time, I feel that I can really rely on the numbers to help us approach maximum performance. For this weekend, this is of special importance, since we are the ONLY J/35 out there. At the NA’s we have the best performance measurement yet invented, another well sailed boat right BEHIND us.
Before we hoisted the sails on Wednesday we motored at about 2 knots directly into the wind. I set up the ‘ShearOmometer‘ (a line with cassette tape streamers every 6 feet) from the masthead. I deliberately made the top streamer long so that it would rub against the backstay (Mike’s idea). Using this and the windex I had my ‘instrument guy’ punch the SET button on the DeWiggler ‘Offset’ screen and write down the numbers he got for offset. We gathered 18 numbers and then threw out the lowest 2 and the highest 2, took the average of the remaining 14 numbers and used that as the WindAngle Offset. The number ended up at -10.009, obviously I can’t drill a hole straight (the hole in the wand).
We then hoisted the sails (#3 and full main) and started to work to windward. The Shear-o-meter had told us that on this night there was little to no wind shear, so I was really interested to see if our TWA would be the same on both boards. IT WAS within 2 degrees.
Looking at the stripchart of the True Wind Direction, I could not identify the tacking points (no wiggle). When we turned to set the kite, same thing…. no wiggle and the sailing angles were equal on both sides. Lot’s of wind holes on the run… Wind was mostly 15-17 knots with 1 minutes periods of 8-10…. the holes always came just as the big wave sets came through, so no surfing.
Kevin Grainger (Gumption, a J105 out of Rye NY) had his wiggle reduced from 5° to less than 1° after two sailing days.
Testing boatspeed and heading
The analysis gives the amount of current the test was sailed in and recommendations for boatspeed calibration, compass lubber offset, your compass deviation rank percentile and a table of compass deviation. The deviation table can be inserted into the T1 processor for a complete compass correction.
Compasses comprise approximately half the error encountered with wind direction and current set & drift calculation. There is no ready way to determine how much the compass is contributing without testing it.
- First results of the Speed-Heading test showed that the median peak-to-peak deviation was 6°, with a worst case of 13.6°, and a best case of 1.2°. Compass deviation contributes 100% into wind direction wiggle and current calculation.
- Compass misalignment (‘lubber offset’) averages 3° with a worst case of 7.7° and a best case of 0.3°. Misalignment has no effect on wind direction wiggle, but affects current calculation.
For current calculation, total compass error (deviation plus offset) produces an error in current calculation of roughly 0.1 knot for per degree of error. In a half knot current, this will throw the current direction calculation off by about 10° per degree of compass error.
- Performing the automatic compensation (“Re-spinning”) of the compass and testing again (about half tried this) results in an average improvement of about 2°.
- For the two worst cases (peak-to-peak deviation in excess of 10°), moving the compass was recommended. Retesting showed an average improvement of 220% or 8°.
- There was one case where the compass was replaced with a more modern one. There was no effect on accuracy, indicating that the problem was the compass environment and not the hardware or testing method.
- Of all the DeWiggler tests, the speed & heading test is the easiest to do, taking about an hour, with no crew or sail handling required. It will address a major contributor to instrument error with very little investment in time.
- If your compass tests out with more than 6° peak-to-peak deviation, the compass should be moved to a location with less magnetic interference.
- If the lubber offset is more than about 1.5°, the compass mount should be adjusted to improve the current calculation.
Wind Direction Wiggle
DeWiggler statistics for wind direction are harder to compare due to the extra dimension of true wind speed. Because of this, tests rarely can be compared apples to apples. The increased effort required makes it harder to get retests done, so before and after comparisons are rare (at least in the first year of operation).
However, by picking two cases, this can be said: DeWiggler tends to reduce wind direction wiggle by about 50% in the first pass.