[N282S]

I'm curious how winds aloft are derived in both the GRT and Dynon. I have a GRT sport on the left side and a Dynon D10A on the right side. The GRT magnatometer is mounted high and aft of the baggage bulkhead and the GRT temp probe is mounted in the side cabin air intake. The Dynon Mag and temp probe are mounted out under the wing (away from engine and exhaust heat). As expected the GRT OAT reads a bit higher than the Dynon.

Both units display very different winds aloft. Sometimes as much as a 20 knots and 120 degree difference. I suspect the Dynon to be the most accurate. Anyone with some insight?

Thanks

 

[Bavafa]

As far as I know the direction of the wind is calculated based on heading and track. The speed is calculated based on TAS and GAS, this is at least in the case of GRT. You can calibrate the magantameter in GRT for limited errors that maybe as a result of installation location or other factors.

Your track is ought to be accurate since it is derived from GPS, so I would check your heading between GRT and Dynon in the air and see how closely they correspond and if one or both needs calibration.

You can also check your speed as well, IAS probably reads the same since it is of the same source (pitot/static) but need to double check your TAS. In GRT you can calibrate your TAS to compensate for any error there may be in pitot/static). I don?t know if you can do the same in Dynon or not, I suspect you can.

Hope this helps

 

[Ironflight]

Steve - you don't say how much of an error you're getting in your GRT TAS, but depending on the geometry of the winds, basic trigonometry is going to come in to play and a small TAS error can become a large wind error under the right conditions, so if your TAS is off by 10 knots, the winds are probably going to be pretty bogus. Wind calculations are pretty basic, assuming the input data is correct, and I have observed that both GRT and Dynon's algorithms work fine.

Relocating temp probes is not that hard, and when you consider all of the functions in an EFIS that depend on an accurate TAS, it is worth getting the capabilities that you've already paid for!

 

[pierre smith]

Calibration is key!

When I bought my -10, the winds aloft were completely out of the ballpark!

I re-calibrated the Dynon 100 on a compass rose that had been placed there by surveyors...and voila!!...instant accuracy, it had been so far off! Now, the 430 and the D-100 have nice conversations

Best,

 

[N282S]

Steve - you don't say how much of an error you're getting in your GRT TAS, but depending on the geometry of the winds, basic trigonometry is going to come in to play and a small TAS !

I'm flying today and will write down some readings. The heading is, however, only off a couple of degrees. I suspect the prob is the temp probe.

I do recall a time, long long ago, when you could figure your winds. What was that thing we used to use? An e6b or something??

 

[Dynon]

When you break out the E6B, you'll see that a few degrees of heading error can make 20 degrees of wind error if the winds are light. Wind calculations are nothing without really accurate heading and TAS.

 

[N282S]

inflight data

I went back to the compass rose and did another alignment of both the GRT and Dynon. I double checked the magnetometer locations of both the GRT and Dynon. Everything looks good - proper physical alignment, no ferrous metal, no nearby wires. Should work.

Here's the data I recorded at 6,000' msl:

. . . GRT : : Dynon

HDG 359 : : 002
TAS 122 : : 133
TEMP 22 : : 16
WINDS 205/15 : : 172/18

HDG 090 : : 098
TAS 110 : : 118
TEMP 21 : : 16
WINDS 212/17 : : 196/37

HDG 180 : : 180
TAS 110 : : 122
TEMP 21 : : 16
WINDS 243/17 : : 249/13

HDG 270 : : 266
TAS 109 : : 122
TEMP 21 : : 16
WINDS 203/17 : : 184/16

Differences in OAT could account for TAS difference between the two units. But, what is interesting is that both the Dynon and GRT report roughly the same errors in winds, ie. GRT is constant at north and east and west, but south reports roughly + 30 to 40 degree wind shift. Same with the Dynon but, a slight larger shift. Two things immediately come to mind that both units have in common with my airplane. One, they both receive data from the same Garmin 296 gps and two, the airplane is taildragger and both units were calibrated on the ground without raising the tail to an inflight attitude.

Should the tail be raised to an inflight position when aligning the units?

Comments or suggestions?

 

[Rainier Lamers]

I would assume that both Dynon as well as GRT would use a very similar, if not identical formula to what we use at MGL.
You do not give ALL of the required input values so I can't punch this into the formula to see what pops out.

You need:

a) Ground speed from GPS
b) TAS (you have given that).
c) Heading from compass (you have given that)
d) Ground track from GPS.

I have copied the complete formula text below for those that want to play a bit.

There is nothing difficult about this but it does require accurate inputs to give meaningful outputs. Errors tend to relate to TAS and heading as GPS tends to be quite accurate.
Magnetic heading relates to the direction your aircraft is pointing. This means the difference your aircraft is pointing vs ground track (corrected for variation of course) is your aircraft's crab angle.
Crab angle is important here. It also means that if you fly wing low (ball not in the center) the formula will not work right as the crab angle is not measured right. We also can't use this in a turn. Straight and level, ball in the middle.
If there is any sort of crab angle - wind is coming from the side. Difference between ground speed and TAS would indicate if wind is from behind or on the nose. If you have both differences, you can work out the wind triangle and thus the apparent wind speed.
Good numbers in - good numbers out.

Your listed measurements already show significant differences between the GRT and Dynon systems, quite enough to give you the results you see even if we do not know all of the factors involved.

TAS should be reasonably accurate if OAT is measured correct. Magnetic heading is where things usually go wrong. It can be difficult to calibrate a compass so it is accurate to say 1-2 degrees on all headings (which is what you want) AND for it to stay that way !

Hope this helps...

Rainier
CEO MGL Avionics


From "Aviation Formulary V1.45"

Wind Triangles
In all formulae, all angles are in radians. Convert back and forth as in the Great Circle section. [This is unnecessary on calculators which have a "degree mode" for trig functions. Most programming languages provide only "radian mode".]

angle_radians=(pi/180)*angle_degrees
angle_degrees=(180/pi)*angle_radians
A further conversion is required if using degrees/minutes/seconds:

angle_degrees=degrees+(minutes/60.)+(seconds/3600.)

degrees=int(angle_degrees)
minutes=int(60*(angle_degrees-degrees))
seconds=60*(60*(angle_degrees-degrees)-minutes))
[ You may have a built-in HH <-> HH:MM:SS conversion to do this efficiently]
Let CRS=course, HD=heading, WD=wind direction (from), TAS=True airpeed, GS=groundspeed, WS=windspeed.

Units of the speeds do not matter as long as they are all the same.

(1) Unknown Wind:

WS=sqrt( (TAS-GS)^2+ 4*TAS*GS*(sin((HD-CRS)/2))^2 )
WD=CRS + atan2(TAS*sin(HD-CRS), TAS*cos(HD-CRS)-GS) (**)
IF (WD<0) THEN WD=WD+2*pi
IF (WD>2*pi) THEN WD=WD-2*pi
( (**) assumes atan2(y,x), reverse arguments if your implementation
has atan2(x,y) )

(2) Find HD, GS

SWC=(WS/TAS)*sin(WD-CRS)
IF (abs(SWC)>1)
"course cannot be flown-- wind too strong"
ELSE
HD=CRS+asin(SWC)
if (HD<0) HD=HD+2*pi
if (HD>2*pi) HD=HD-2*pi
GS=TAS*sqrt(1-SWC^2)-WS*cos(WD-CRS)
if (GS < 0) "course cannot be flown-- wind too strong"
ENDIF
Note:
The purpose of the "if (HD<0) HD=HD+2*pi; if (HD>2*pi) HD=HD-2*pi" is to ensure the final heading ends up in the range (0, 2*pi). Another way to do this, with the MOD function available is:
HD=MOD(HD,2*pi)
(3) Find CRS, GS

GS=sqrt(WS^2 + TAS^2 - 2*WS*TAS*cos(HD-WD))
WCA=atan2(WS*sin(HD-WD),TAS-WS*cos(HD-WD)) (*)
CRS=MOD(HD+WCA,2*pi)
(*) WCA=asin((WS/GS)*sin(HD-WD)) works if the wind correction angle is less than 90 degrees, which will always be the case if WS < TAS. The listed formula works in the general case

 

[N282S]

Rainier,
Thanks for chiming in. Good to hear from a "third" party! I think it has become apparent that until I fix a couple of things, I am hoping for accurate data that's just not going to happen. First, I need to move the GRT OAT probe to a location out of the engine airstream. It is consistently 5-6 C higher than the Dynon which is mounted under the wing. Then double check both magnetometers are perfectly aligned with the aircraft. Another variable is the yaw angle. On a no winds day, my aircraft yaws just a teeny tiny amount to the left. Which would say there are winds when there aren't. Also, I did not center the ball when the four readings were taken. And finally, for comparison between the two units, I should have flown the gps cardinal directions rather than slave the autopilot to the four cardinal directions which would have given me a wind induced crab angle (track) from the gps direction and that would have produced a more accurate track for comparison on the two efis units. So, back to the drawing board.

Although the formula did make my old tired brain hurt, I appreciate it. It gives me a reason to dust off my old trig text book. I'm just glad you added the part "There's nothing difficult about this." This explains why you develop high end avionics systems and I do not.