Quoting Keith Packard <kei...@keithp.com>:

On Mon, 28 Feb 2011 15:05:37 -0800, I <kirk...@pdx.edu> wrote:

Quoting Nathan Bergey <nat...@psas.pdx.edu>:

> Yes, I think the main goal is to get people to think about GPS on rockets.
> Even if we don't make an attempt this should be fun to watch.
> More (selfishly) importantly I've been invited on the Evadot podcast
> tomorrow morning to be an "expert" pundit of sorts about the contest and
> what is hard about it, what's easy, etc.
> Anyone have thoughts about it? Is it a good idea? Is it easy?

It's not trivial, but it can be done. The hard part is tracking the
*rate* that the Doppler-affected satellite signal frequency changes.
If you build a GPS that can track (the frequency change) that fast, it
becomes susceptible to noise. This can be greatly alleviated by
tightly coupling the GPS and INS sensors (+ details).

> How long will
> it take for someone to win?

GPS units capable of high dynamics already exist. If someone that has
one knows a rocket guy, problem solved.

Tracking the max altitude doesn't require high dynamic response; rockets
tend to have low acceleration outside of the boost phase. There are lots
of commercial GPS units capable of accurately tracking rocket position
during coast, apogee and descent. The Trimble units are often used for
this altitude. In fact, the new TRA altitude record rules for flights
above 40k' require use of one of these in place of a barometric

I guess that's true... re-reading the challenge, the point is the 100,000 ft, not the GPS tracking.

OEM's have a tracking loop (google Costas loop) typically with a 1Hz
bandwidth (source: a GPS INS book I have. It's at home at the moment).
Some OEM units can change the loop bandwidth to something higher, but
with poorer accuracy and greater likelihood of loosing the satellite
lock. The result is this: if the GPS accelerates (relative to the
satellite) above a certain rate (some GPS OEMs advertise 4 g's) the
correlator cannot change the carrier tracking frequency fast enough to
keep up, so the signal lock is lost.

Right, you lose the signal during boost, but rapidly re-acquire it after
the motor burns out.

The GPS in TeleMetrum loses lock under boost, but usually re-locks
within a few seconds of motor burnout. It's problem is not the tracking
loop low-pass filter, but the Kalman filter post-processing the raw GPS
coordinates into the displayed values. That shows a huge lag, which
you'd expect if the model error co-variance values were very small.

I'd be happy enough to be able to change the Kalman filter values and
not mess with the tracking loop filter; I'm not trying to steer.

Huh? I'm pretty sure the tracking loop is causing the GPS to lose lock. Especially in your rockets (with your more-reasonable-than-PSAS mass fractions)! I'd agree that one could build a KF to track position during that time, but from what I've read on tracking correlators, IMHO the loop filter is never going to keep up with your 20+g rockets on boost.

So... if this is a height competition, I would think that the sponsor would provide the GPS... Mostly because folks like us can produce a GPS log that clearly indicates 100,000 ft from my Liberty 4 ;) In fact, I'd bet Dave can do it with 2 lines of PERL.

This is actually doable. Too bad we're broke...

psas-team mailing list

This list's membership is automatically generated from the memberships of the 
psas-airframe, psas-avionics, and psas-general mail lists. Visit 
http://lists.psas.pdx.edu to individually subscribe/unsubscribe yourself from 
these lists.

Reply via email to