Space-X Dragon

On Mon, 12 Feb 2007 02:06:35 -0600, in a place far, far away, Pat
Flannery made the phosphor on my monitor glow in
such a way as to indicate that:



Ian Stirling wrote:

X33 was not a engineering failure.
It was a managment and political one.

Of course if we'd gone with one of the less radical contenders and it
also hadn't panned out as far as performance goes; then everyone would
have been saying the contract assignment had been fixed, and that only
if they had gone with the miraculous Lockheed design, all would have
been well.
You can still see that effect running around with the Faget shuttle and
Bono SSTO concepts.

Which is why the notion of selecting a single contractor and concept
that early in the design process was a disaster, and one that was
obvious in prospect. If they didn't have the money to do a flyoff,
they shouldn't have done any. That program was the single biggest
disaster for our prospects in space in the 1990s.

Henry Spencer wrote:
In article ,
Pat Flannery wrote:
(In fact, because of this Apollo had some glaring problems - like the
three axis gimbal system in the inertial measurement unit.)

I wonder if they'll use laser ring gyros on the new one?

Probably not -- those are now yesterday's technology. (For one thing,
despite what you might think, they still have moving parts.) Fiber-optic
gyros are increasingly replacing them, and hemispherical-resonator gyros
are another major competitor.

Also, it's no longer necessary to push the gyro technology hard, because
automatic star trackers can provide almost continuous attitude updating.
(Compare to Apollo, which got updates only every few hours when the crew
did a manual star sighting.)

For at least earth orbit, they should be able to update their navigation
computers via GPS.

With smart signal processing, you can get some limited use of GPS up to
much higher altitudes. Not the sort of essentially-instantaneous full
position solution that you get on the surface or in LEO, but data that
puts constraints on position and can be used, over time, to correct for
drift in on-board estimates. Might not be practical out at lunar
distances, though.

It's not actually that bad.
The fade margin of current GPSs is some 15dB or better, with the stock
antennas.
This is at 20000Km from the antenna.
The moon is 20 times as far as this, so the signal is 1/400th the
strength.
Assuming that 5dB is lost to being at the edge of the beam pattern (you
can only see satellites near the horizon of earth, that are shining
you with the overspill.

Assuming we want a 3dB fade margin - compared to a civilian reciever.

This is 7dB, or around 5 times the required signal, neglecting distance.
Or 80 times the antenna gain of the standard omni.
Or an antenna with a beam of around .25 radians.
The GPS signal has a wavelength of around 25cm, which means a dish
of around 1.2m.

This isn't bad at all, especially as it hasn't even considered any
possible signal processing gain that might be gotten by a smarter
reciever.

Assuming that the relative range to each satellite can be gotten to
within 1m as is easily possible with commercial gear, neglecting
ionospheric effects. (not really range...)

This gives you a radial position easily within 30m or so, most of the
time.

Range (altitude) is much harder, as you can't see the satellites you
really want - which are obscured by earth.

The baseline is much worse - you're limited to that segment of the orbit
which you can recieve the GPS satellite over, perhaps 10000Km, compared
to 30000Km for the radial position.
I'd imagine that this is some hundred metres.

A very low power transmission of time from earth to you will nail this
down to under a metre.

This is all however instantaneous positioning, not averaging over a
period of several hours.

On 16 Feb, 22:02, Ian Stirling wrote:
Henry Spencer wrote:
In article ,
Pat Flannery wrote:
(In fact, because of this Apollo had some glaring problems - like the
three axis gimbal system in the inertial measurement unit.)

I wonder if they'll use laser ring gyros on the new one?

Probably not -- those are now yesterday's technology. (For one thing,
despite what you might think, they still have moving parts.) Fiber-optic
gyros are increasingly replacing them, and hemispherical-resonator gyros
are another major competitor.

Also, it's no longer necessary to push the gyro technology hard, because
automatic star trackers can provide almost continuous attitude updating.
(Compare to Apollo, which got updates only every few hours when the crew
did a manual star sighting.)

For at least earth orbit, they should be able to update their navigation
computers via GPS.

With smart signal processing, you can get some limited use of GPS up to
much higher altitudes. Not the sort of essentially-instantaneous full
position solution that you get on the surface or in LEO, but data that
puts constraints on position and can be used, over time, to correct for
drift in on-board estimates. Might not be practical out at lunar
distances, though.

It's not actually that bad.
The fade margin of current GPSs is some 15dB or better, with the stock
antennas.
This is at 20000Km from the antenna.
The moon is 20 times as far as this, so the signal is 1/400th the
strength.
Assuming that 5dB is lost to being at the edge of the beam pattern (you
can only see satellites near the horizon of earth, that are shining
you with the overspill.

Assuming we want a 3dB fade margin - compared to a civilian reciever.

This is 7dB, or around 5 times the required signal, neglecting distance.
Or 80 times the antenna gain of the standard omni.
Or an antenna with a beam of around .25 radians.
The GPS signal has a wavelength of around 25cm, which means a dish
of around 1.2m.

This isn't bad at all, especially as it hasn't even considered any
possible signal processing gain that might be gotten by a smarter
reciever.

Assuming that the relative range to each satellite can be gotten to
within 1m as is easily possible with commercial gear, neglecting
ionospheric effects. (not really range...)

This gives you a radial position easily within 30m or so, most of the
time.

Range (altitude) is much harder, as you can't see the satellites you
really want - which are obscured by earth.

The baseline is much worse - you're limited to that segment of the orbit
which you can recieve the GPS satellite over, perhaps 10000Km, compared
to 30000Km for the radial position.
I'd imagine that this is some hundred metres.

A very low power transmission of time from earth to you will nail this
down to under a metre.

This is all however instantaneous positioning, not averaging over a
period of several hours.

Don't GPS satellites have directional antenna's pointing at Earth?
They have so little power it doesn't make sense to waste it aiming at
the moon.

Even if you could pick up the software, the satellites are all in the
same direction, This is a problem that will be encountered in urban
canyons on Earth with GPS/Galileo dual systems . Even if you can see
three satellites, they are too close together to give an accurate
position.

Finally, the positioning software would need to be rewritten. Current
systems start with the assumption that the user is inside the GPS
sphere.

Alex Terrell wrote:
On 16 Feb, 22:02, Ian Stirling wrote:
snip
The baseline is much worse - you're limited to that segment of the orbit
which you can recieve the GPS satellite over, perhaps 10000Km, compared
to 30000Km for the radial position.
I'd imagine that this is some hundred metres.

A very low power transmission of time from earth to you will nail this
down to under a metre.

This is all however instantaneous positioning, not averaging over a
period of several hours.

Don't GPS satellites have directional antenna's pointing at Earth?
They have so little power it doesn't make sense to waste it aiming at
the moon.

Yes. Sort-of.
I was extrapolating the radiation pattern out a bit, from the fact that
IME, the signal bars on my GPS do not dip appreciably with satellites on
the horizon.
I suppose I should look up the radiation pattern, and calculate if you
will typically get 3 or 4 satellites needed for a minimal lock.

Even if you could pick up the software, the satellites are all in the
same direction, This is a problem that will be encountered in urban
canyons on Earth with GPS/Galileo dual systems . Even if you can see
three satellites, they are too close together to give an accurate
position.

This is geometrical dilution of position.
It hurts you badly for 'altitude' - not so much for position.

'altitude' could be greatly improved by beaming a time reference from
earth.

Finally, the positioning software would need to be rewritten. Current
systems start with the assumption that the user is inside the GPS
sphere.

Well - yes.
It's not a really big change though.

In article , Ian
Stirling wrote:

Even if you could pick up the software, the satellites are all in the
same direction, This is a problem that will be encountered in urban
canyons on Earth with GPS/Galileo dual systems . Even if you can see
three satellites, they are too close together to give an accurate
position.

This is geometrical dilution of position.
It hurts you badly for 'altitude' - not so much for position.

'altitude' could be greatly improved by beaming a time reference from
earth.

The GPS satellites are already beaming a time reference to your
spacecraft. Another coming from Earth wouldn't help much.

What would help is sending a time reference from your spacecraft to a
receiver on Earth. This is logically equivalent to a two-way ranging
measurement.

--
David M. Palmer (formerly @clark.net, @ematic.com)

David M. Palmer wrote:
In article , Ian
Stirling wrote:

Even if you could pick up the software, the satellites are all in the
same direction, This is a problem that will be encountered in urban
canyons on Earth with GPS/Galileo dual systems . Even if you can see
three satellites, they are too close together to give an accurate
position.

This is geometrical dilution of position.
It hurts you badly for 'altitude' - not so much for position.

'altitude' could be greatly improved by beaming a time reference from
earth.

The GPS satellites are already beaming a time reference to your
spacecraft. Another coming from Earth wouldn't help much.

What would help is sending a time reference from your spacecraft to a
receiver on Earth. This is logically equivalent to a two-way ranging
measurement.

Actually - it helps a fair bit, as it increases the baseline (the
satellites are all tens of thousands of kilometers behind the
transmission point. But, telling the spacecraft the two-way delay is of
course better.

On 18 Feb, 22:17, Ian Stirling wrote:
Alex Terrell wrote:
On 16 Feb, 22:02, Ian Stirling wrote:
snip
The baseline is much worse - you're limited to that segment of the orbit
which you can recieve the GPS satellite over, perhaps 10000Km, compared
to 30000Km for the radial position.
I'd imagine that this is some hundred metres.

A very low power transmission of time from earth to you will nail this
down to under a metre.

This is all however instantaneous positioning, not averaging over a
period of several hours.

Don't GPS satellites have directional antenna's pointing at Earth?
They have so little power it doesn't make sense to waste it aiming at
the moon.

Yes. Sort-of.
I was extrapolating the radiation pattern out a bit, from the fact that
IME, the signal bars on my GPS do not dip appreciably with satellites on
the horizon.
I suppose I should look up the radiation pattern, and calculate if you
will typically get 3 or 4 satellites needed for a minimal lock.

Even if you could pick up the software, the satellites are all in the
same direction, This is a problem that will be encountered in urban
canyons on Earth with GPS/Galileo dual systems . Even if you can see
three satellites, they are too close together to give an accurate
position.

This is geometrical dilution of position.
It hurts you badly for 'altitude' - not so much for position.

'altitude' could be greatly improved by beaming a time reference from
earth.

Haven't you got this the wrong way round? With one time signal, the
other satellites position you somewhere on the surface of a sphere of
known diameter. However, if all spheres are centered in the same
vicinity, my HDOP will be rather large.