On Jan 14, 9:51*am, bildan wrote:
On Jan 14, 5:22*am, dhaluza wrote:



Differentiating altitude from a 1Hz GPS is going to be fairly useless
for a vario. You will need a GPS that outputs true vertical velocity,
and preferably 1Hz. GPS velocity is very accurate because it is taken
directly from doppler shifts in the GPS signal, not from
differentiating position. It has been reported that low cost GPS
horizontal velocity has an accuracy of 5mm/s (0.01 knots). Vertical
velocity error will be greater because of satellite geometry, but
probably only double that. You don't need a WAAS receiver to get this
precision, but it probably helps somewhat since you need to know your
position precisely to calculate the relative satellite velocity. You
don't need fancy antennas, just a good view of the sky. Fortunately
it's the overhead satellites that provide vertical velocity
information, so as long as your bank angle does not exceed 45 degrees,
you should be able to track these constantly as long as the antenna is
properly positioned.

Unfortunately, vertical velocity is not included in any of the
standard NMEA sentences--probably because boats are not supposed to
have a vertical velocity. So most low-cost GPS units will not have
this. The Garmin GPS-18 does have a proprietary sentence with X, Y & Z
velocities, and it also comes in a 5 Hz version for ~$150.

TE compensation will be a bit tricky without airdata input, but it
could potentially be better since it is not subject to errors like
gusts and lags (not to mention leaks). I don't see why an INS based on
low-cost automotive grade sensors could not provide TE comp at least
as good as a pneumatic system. Any change in kinetic energy will
require an acceleration that can be measured by an accelerometer. But
you would need to remove acceletation due to gravity and rate of turn,
so this would also require solid-state gyros, and a lot of software to
integrate them.

On Jan 13, 2:38*pm, Darryl Ramm wrote:

On Jan 13, 10:05*am, "kirk.stant" wrote:

On Jan 13, 11:03*am, wrote:

On Jan 13, 8:36*am, Darryl Ramm wrote:

On Jan 13, 8:23*am, "kirk.stant" wrote:

Is anyone working on or studying using WAAS GPS data to make a stand-
alone vario? *How would TE be implemented in such a device? Ground
speed change during a pullup? *TAS based on circling winds?

Do the current PDA software programs (mSeeYou, Winpilot, etc.) when
used in GPS-only mode provide accurate vario data when hooked up to a
WAAS GPS?

Just curious - seems a 5 hz WAAS GPS could be the basis for a really
nice vario that wouldn't need any pitot-static imputs. *I use mSeeYou
and a Themi, non-WAAS, and the "vario" data seems close, but I use it
more for trend and average than instantaneous data.

Winter can't end soon enough!

Kirk
66

Yes but how does it tell the difference between an increase in wind
and a pull up? How will it handle STF calculations in a strong
headwind/wave where it may be really confused what is going on?

Darryl

Darryl's right (again).

I don't think there's really a satisfactory way to take wind out of
the equation. You could try to do it with track and drift trends, but
that's not very precise given how much the wind can vary with
altitude, position and over time.

9B- Hide quoted text -

- Show quoted text -

Cmon, let's think out of the box!

An INS/IMU can measure winds without air data. *How accurate (read
"expensive") would an inertial sensor need to be to provide useful
wind data (or just TAS vs GS)?

Or bite the bullet and T into the pitot/static system and provide the
data to the GPS vario - no worse than a TE connection for a
conventional vario.

Question is whether it would be cheaper and/or better than current
mechanical or electronic TE varios? *Better or worse at altitude,
faster/slower response, etc?

I'm thinking: *One little box with an LCD display. *TE vario, audio,
horizontal lift distribution in the current thermal in real time (for
centering), current average, past averages/acheived climbs (trend for
MC settings), logger, GPS output for other devices. *Run off ship
power and have a backup rechargeable battery.

Stick in a 68mm hole, hook up pitot/static and power, off you go.
Replaces your backup mechanical or electric, gives you a real backup
when your TE probe falls off! *Do it for less than $500 (yeah, right,
I know...) and you stick it in every glider out there (think of all
the crap instruments in club/commercial gliders in the US).

And please don't say "Cambridge 302" - TE probes are soooo 20th
century!

Kirk
66

I am not aware of any IMS/IMU that attempt to calculate wind without
air data input. Do you know of one? It is likely a significant
challenge for rate based systems (vs. position based with GPS). e.g.
"taking a turn" cannot help determine wind with an INS. All the poor
thing can try to do is integrate external accelerations on the
aircraft caused by changes in wind.

While trying to integrate up rate based sensors to determine wind is
likely impractical. Using rate based sensors to filter other rate base
inputs likely makes more sense. I believe accelerometer based
assistance is already used to help improve/filter gusts and other
effects on variometers (wether using TE probe or digitally adjusted TAS
+pitot). I believe the Cambridge 302 uses it's accelerometers for
this, but also have heard rumors that this was never really fully
developed in the software. I am not sure if other vario/computer
systems also do this. I certainly like how the vario in the 302
performs (and I'm using electronic TE compensation with mine).

Paul gives on example of the SeeYou Mobile thermal assistant not
working well with just GPS+WAAS input. In the past other people have
tried to use STF data through PDA software. *Finally Naviter had to
warn pilots this can't possibly work, they just don't have enough data
to calculate something useful.

Even if all the above were not show-stopping issues you'd have to look
at the noise spectrum of the altitude signal around a fraction to 1 Hz
to see how bad differentiating (for vertical velocity) and filtering
this is going to be. I just don't have that data handy. And you may
need a sophisticated antenna system to provide a good GPS satellite
sky view when turning tightly. A TE probe, which is just a couple of
holes or a slot cut in a few dollars worth of tube seems a lot easier
way to get basic data. As Dick Johnson kept reminding us, you don't
need a fancy tail mount TE probe a simple home made fuselage mounted
one works great. An electronic pressure sensor to incorporate into a
vario costs a few dollars. The software to make all this work really
well. Priceless. I can't wait to see what Dave Ellis does at
ClearNav...

Darryl

GPS, even WAAS enabled, is best considered to be a highly accurate but
interruptable data source. *Inertial reference units are best
considered a less accurate but non-interruptable data source. *GPS
signals can provide not only position data but, with multiple
antennas, can provide attitude data. *Combining the two with a Kalman
filter where GPS keeps the INS updated results in the best of both.

This combination outputs highly accurate Euler angles, 3D velocity and
position data. *In other words, your gadget would know where it was,
its pitch, roll and heading angles, and its velocity on each of three
axes - all to extremely high precision. *There's a lot that can be
done with these data.

Could this hypothetical gadget be used as an inertial TE vario?
Absolutely, as long as all TE calculations were done in the same
inertial reference frame. *Speed to fly would require air data,
however.

The benefits of a GPS/inertial system would include:

A vario with no gust sensitivity, high S/N ratio and instant
response,
Instantly updated, highly accurate vector wind data.
Accurate lift mapping for thermal centering assistance.

Bill Daniels

Spoken like a true control systems whiz.

Maybe this is what you had in mind, but wouldn't the best Kalman
filter combine GPS, inertial and air data into a single optimal
estimate for all the relevant rates and angles? The air data has lags
for sure, but you should be able to model the dynamics and use it to
take the drift out of integrated accelerometer signals.

The cool part (I think) would be that you'd be able to make a vario
that can distinguish between horizontal and vertical gusts - which
ought to better correlate to and enhance the "seat of the pants" feel
that a lot of us try to use in choosing and centering lift.

Do you think differential GPS has the resolution to give good attitude
information? I think there are laser ring gyros out there now that are
pretty inexpensive and I think the drift is pretty good for our
application, but I have no idea which is would be more accurate - GPS
would almost certainly be cheaper and easier on the batteries.

Andy