Best performing Vario?

Seriously - 20 samples per second? Engineers put men on the moon with slide rules. You just don't need the kind of accuracy being touted these days.

The sensors used are not that well-behaved. Not very linear, noisy, etc. So you might actually need that 20Hz to produce a proper value at 1Hz. Many noisy measurements average
out as one pretty good measurement. This has even been proven, by letting 1000 people guess the number of balls in a fish bowl. The average was spot-on.

Also, fast sampling allows one to do all kinds of filtering in software which required analogue filters in the past.

What I have read on this forum has led me to think that the Butterfly is currently a decent vario in operation but nothing special. Can anyone with practical experience say different?

I surmise that the software engineers have not yet discovered what to do with all the data available to them.

Since my LX 9000 has the optional giro box what data would it not have that would potentially make the Butterfly better?

At 17:34 13 February 2013, waremark wrote:
What I have read on this forum has led me to think that the
Butterfly is
currently a decent vario in operation but nothing special. Can
anyone with
practical experience say different?

I surmise that the software engineers have not yet discovered
what to do
with all the data available to them.

Since my LX 9000 has the optional giro box what data would it
not have that
would potentially make the Butterfly better?

The very alluring Butterfly vario is using the sensor data but the
concern that has stopped me choosing it so far is the
importance for its performance of getting an accurate reading
from the magnetometer. It is inside the main sensor box which
has all the electrical and pneumatic connections and that makes
it awkward to mount in a way that meets the installation
manual's specifications for distance from electromagnetic field
interference - especially in a retractable motor glider. Two
friends have returned them because theye were unable to
achieve this.

I have the same worry about getting the similarly configured CN
vario for the same reason - although it doesn't matter at
present while the magnetometer is not used (AFAIK).

The LX 9000 AHRS unit has "3 axis Gyros with MEMS technology
and 3 axis digital +-6g accelerometer" according to the
website. If the separate LX 9000 compass module is 3-axis(??)
then including that along with the GPS in the 9000 and the
pneumatic sensors from the V5 vario would appear to make it
match the sensor mix of the Butterfly. Whether they could all
be integrated to work as a full intertial sensor unit I have no
idea - nor whether LX Nav have any plan to try to make it do
so. The fact that the magnetometer is a separate box should
make it easier to locate away from interference.

John Galloway

On Wednesday, February 13, 2013 11:33:57 PM UTC+2, John Galloway wrote:
At 17:34 13 February 2013, waremark wrote:

What I have read on this forum has led me to think that the

Butterfly is

currently a decent vario in operation but nothing special. Can

anyone with

practical experience say different?



I surmise that the software engineers have not yet discovered

what to do

with all the data available to them.



Since my LX 9000 has the optional giro box what data would it

not have that

would potentially make the Butterfly better?



The very alluring Butterfly vario is using the sensor data but the

concern that has stopped me choosing it so far is the

importance for its performance of getting an accurate reading

from the magnetometer. It is inside the main sensor box which

has all the electrical and pneumatic connections and that makes

it awkward to mount in a way that meets the installation

manual's specifications for distance from electromagnetic field

interference - especially in a retractable motor glider. Two

friends have returned them because theye were unable to

achieve this.



I have the same worry about getting the similarly configured CN

vario for the same reason - although it doesn't matter at

present while the magnetometer is not used (AFAIK).



The LX 9000 AHRS unit has "3 axis Gyros with MEMS technology

and 3 axis digital +-6g accelerometer" according to the

website. If the separate LX 9000 compass module is 3-axis(??)

then including that along with the GPS in the 9000 and the

pneumatic sensors from the V5 vario would appear to make it

match the sensor mix of the Butterfly. Whether they could all

be integrated to work as a full intertial sensor unit I have no

idea - nor whether LX Nav have any plan to try to make it do

so. The fact that the magnetometer is a separate box should

make it easier to locate away from interference.



John Galloway

I thought exactly same about placement of Butterfly IMU. Separate magnetometer would be more convenient, as you have to route GPS antenna, three pneumatic tubes and CAN-bus to the box. AFAIK succesful installation requires quite a distance from *everything*. BTW common radio speakers don't have any magnetic shielding around speaker magnet, for some reason. Strong magnets can disturb compass as far as 1,5 metres away...

If the butterfly -- or clearnav, with future software -- did read out 3 d wind 20 times a second, this would be a big advantage. Not only could you detect lift long before F = MA gets the glider moving upward, you could tell upward gusts from forward gusts and sideways gusts. Dynamic soaring might even become possible, or at least better energy extraction from gusts. Even the 302 has the necessary sensors, my impression is that we're all waiting on the software development.

John Cochrane

On Thursday, February 14, 2013 12:58:13 AM UTC+2, wrote:
If the butterfly -- or clearnav, with future software -- did read out 3 d wind 20 times a second, this would be a big advantage. Not only could you detect lift long before F = MA gets the glider moving upward, you could tell upward gusts from forward gusts and sideways gusts. Dynamic soaring might even become possible, or at least better energy extraction from gusts. Even the 302 has the necessary sensors, my impression is that we're all waiting on the software development.



John Cochrane

I believe the inertial system needs to sort out the F=MA part before solving 3D-wind. Gyros and acceleration sensors don't measure vertical wind, they measure glider vertical movement.

At 22:58 13 February 2013, wrote:
If the butterfly -- or clearnav, with future software -- did read
out 3 d
w=
ind 20 times a second, this would be a big advantage. Not
only could you
de=
tect lift long before F =3D MA gets the glider moving upward,
you could
tel=
l upward gusts from forward gusts and sideways gusts.
Dynamic soaring
might=
even become possible, or at least better energy extraction
from gusts.
Eve=
n the 302 has the necessary sensors, my impression is that
we're all
waitin=
g on the software development.=20

John Cochrane

Regarding gusts, the Butterfly website (in FAQ) already claims:

"A conventional variometer uses changes in air pressure (TE-
pressure, static- and total-pressure) to determine energy
changes the aircraft experiences. Butterfly Vario does the same.
Additionally it uses an inertial sensing platform that allows for
real-time determination of airmass-movement and realtime
determination of wind. With this technology a pilot can judge
the difference between gust induced energy changes and
thermal induced energy changes."

John Galloway

On Feb 14, 4:49*pm, John Galloway wrote:
At 22:58 13 February 2013, wrote:









If the butterfly -- or clearnav, with future software -- did read
out 3 d
w=
ind 20 times a second, this would be a big advantage. Not
only could you
de=
tect lift long before F =3D MA gets the glider moving upward,
you could
tel=
l upward gusts from forward gusts and sideways gusts.
Dynamic soaring
might=
even become possible, or at least better energy extraction
from gusts.
Eve=
n the 302 has the necessary sensors, my impression is that
we're all
waitin=
g on the software development.=20

John Cochrane

Regarding gusts, the Butterfly website (in FAQ) *already claims:

"A conventional variometer uses changes in air pressure (TE-
pressure, static- and total-pressure) to determine energy
changes the aircraft experiences. Butterfly Vario does the same.
Additionally it uses an inertial sensing platform that allows for
real-time determination of airmass-movement and realtime
determination of wind. With this technology a pilot can judge
the difference between gust induced energy changes and
thermal induced energy changes."

John Galloway

The Butterfly vario does not filter out gusts. The other sensors are
used only to drive the blue ball that shows air mass movement. The
pilot responsibility is to put the two together and decide what is a
gust and what is lift based on the two sources of information. It is
not a bad solution as one can learn the vario behavior. If algorithms
are used to filter out gusts a pilot might have a difficult time to
learn the vario behavior or understand what it is showing as an
algorithm may have many paths covering many different scenarios
behaving differently in different situations. Of course if algorithm
is perfect that would not be a problem, but this is not an easy
problem to solve. I assume it will be years before this happen mainly
to extensive testing/feedback process that is required to develop and
test these kinds of algorithms what works in theory might not be good
in practice. So for the near future approach that Butterfly took is a
good one, we need to see in practice how workable it will be.

After Argentina some competition pilots were quite disappointed
with the new variometers using acceleration to calculate lift. I
have not tested myself, just telling their opinion.

At 21:49 14 February 2013, John Galloway wrote:
At 22:58 13 February 2013, wrote:
If the butterfly -- or clearnav, with future software -- did
read
out 3 d
w=
ind 20 times a second, this would be a big advantage. Not
only could you
de=
tect lift long before F =3D MA gets the glider moving upward,
you could
tel=
l upward gusts from forward gusts and sideways gusts.
Dynamic soaring
might=
even become possible, or at least better energy extraction
from gusts.
Eve=
n the 302 has the necessary sensors, my impression is that
we're all
waitin=
g on the software development.=20

John Cochrane

Regarding gusts, the Butterfly website (in FAQ) already claims:

"A conventional variometer uses changes in air pressure (TE-
pressure, static- and total-pressure) to determine energy
changes the aircraft experiences. Butterfly Vario does the
same.
Additionally it uses an inertial sensing platform that allows for
real-time determination of airmass-movement and realtime
determination of wind. With this technology a pilot can judge
the difference between gust induced energy changes and
thermal induced energy changes."

John Galloway

On Thursday, February 14, 2013 3:19:21 PM UTC-7, Kimmo Hytoenen wrote:
After Argentina some competition pilots were quite disappointed

with the new variometers using acceleration to calculate lift. I

have not tested myself, just telling their opinion.



At 21:49 14 February 2013, John Galloway wrote:


If the butterfly -- or clearnav, with future software -- did

read

out 3 d

w=

ind 20 times a second, this would be a big advantage. Not

only could you

de=

tect lift long before F =3D MA gets the glider moving upward,

you could

tel=

l upward gusts from forward gusts and sideways gusts.

Dynamic soaring

might=

even become possible, or at least better energy extraction

from gusts.

Eve=

n the 302 has the necessary sensors, my impression is that

we're all

waitin=

g on the software development.=20



John Cochrane



Regarding gusts, the Butterfly website (in FAQ) already claims:



"A conventional variometer uses changes in air pressure (TE-

pressure, static- and total-pressure) to determine energy

changes the aircraft experiences. Butterfly Vario does the

same.

Additionally it uses an inertial sensing platform that allows for

real-time determination of airmass-movement and realtime

determination of wind. With this technology a pilot can judge

the difference between gust induced energy changes and

thermal induced energy changes."



John Galloway





Not sure, but I suspect they are mixing frames of reference. In theory, the way to do this is solve the TE equations in the inertial and air-data domains separately then compare them. A gust will show up strongly in the air-data but less so in the inertial data so a computer - or a pilot - can tell the difference.

A purely inertial vario will require a full IMU with GPS updating. These things are probably still too expensive for sailplanes but only just.