|
|
<r9ns@xxxxxxxxxxx> wrote in message
news:1124471646.892674.87620@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
> George Dishman wrote:
>> <r9ns@xxxxxxxxxxx> wrote in message
>> news:1124043045.730875.166230@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
>> > George Dishman wrote:
>> >> <r9ns@xxxxxxxxxxx> wrote in message
>> >> news:1123965839.715622.84990@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
>> >> >
>> >> >> >> >>
>> >> >> >> > All of this necessary information is on my web page. The
>> >> >> >> > point
>> >> >> >> > is
>> >> >> >> > that 12ns laser pulses were received by the photodiode when
>> >> >> >> > the
>> >> >> >> > photodiode was blocked at the photodiode 30 feet away before
>> >> >> >> > the
>> >> >> >> > expected time of arrival about 30ns after being emitted but
>> >> >> >> > these
>> >> >> >> > pulses were not received if the pulse was blocked at the
>> >> >> >> > photodiode
>> >> >> >> > during the time of emission. In both cases the specified and
>> >> >> >> > unspecified conditions of the apparatus were the same so that
>> >> >> >> > any
>> >> >> >> > differences had to be due to the difference in time when the
>> >> >> >> > pulse
>> >> >> >> > was
>> >> >> >> > blocked and when it was not blocked.
>> >> >> >
>> >> >> >> >> > I told you I could not read the raw data on the web with
>> >> >> >> >> > your
>> >> >> >> >> > program
>> >> >> >> >> > or with mine. If your argument is valid with this data why
>> >> >> >> >> > can't
>> >> >> >> >> > you
>> >> >> >> >> > find some other data to make the same argument?
>> >> >> >> >>
>> >> >> > My argument is that you only use data which cannot be read by a
>> >> >> > program which can read all of the other NSSDC data but not this
>> >> >> > specific one.
>> >> >>
>> >
>> >
>> >
>> > Here is the page I wrote in February:
>> >>
>> >> http://www.georgedishman.f2s.com/Ralph/index.htm
>> >>
>> >> The first link has the Excel sheet which includes the
>> >> code. Use "Tools", "Macro", "Visual Basic Editor" to
>> >> access it. If it doesn't appear, Excel may show the
>> >> wrong page by default, choose "Sheet1 (Control)" from
>> >> the project panel on the left. This assumes you have
>> >> Office 2003. In case you still cannot get into it, the
>> >> code itself is here:
>> >>
>> >> http://www.georgedishman.f2s.com/Ralph/Code_For_Ralph.txt
>> >>
>> >> As you can see it is trivial and written to be obvious
>> >> that there is no trick data included.
>> >>
>> > Sorry I see the program but it is not a c++ program
>>
>> "The first link has the Excel sheet ..."
>>
>> It is an Excel sheet. You opened it before but said
>> you couldn't get into the code, you could only see
>> the results and didn't trust them. I have repeated
>> the instructions for accessing the code above.
>>
>> > and you dont
>> > provide a complier
>>
>> None needed, is is a Visual Basic script as I said.
>
> I did this and my computer said a corrupt file and could not read
> this.
I just downloaded it myself and it is fine.
I assume you unzipped the file. Your own web
pages include Word documents so I assume you
have Microsoft Office. What version?
>> Just press the button to run it or display the Debug
>> toolbar and Excel includes the interpreter that lets
>> you single step through the code looking at the sheet
>> to see what it does as it goes.
>>
>> > etc. and the data file was corrupted or not in
>> > readable format
>>
>> What data file? The second link is the VB code as a
>> plain text file. Internet Explorer will display it
>> just by following the link. Just looking at it will
>> show there is no data in it, that was the point!
>>
> The point is that I cant read your files.
You can't read a text file? It should be readable
in your browser without even downloading it.
>> >> > I gave you the web site of my c++ code and of the
>> >> > c++ compiler it is to be used with. You have not done the same.
>> >>
>> >> See above.
>> >
>> > Its not the same and I cant access it.
>> >
>> >> > But
>> >> > again it is strange that you have found no other set of data that
>> >> > shows
>> >> > the same result as your one example. Perhaps you have tested what I
>> >> > think is your oversimplified argument on other data and have found
>> >> > it
>> >> > to give conflicting results.
>> >>
>> >> No, there are others but...
>> >
>> > No graphs are necessary.
>>
>> Yes they are.
> No the raw data numbers show which number is the minimal doppler
> shift in each case.
>
>>First you have to look at the graph
>> to identify the periods of one-way versus three-way
>> operation, second you have to identify the rapid
>> sweep periods and third you need to note the period
>> of settling time for the transponder. That lets you
>> identify the periods of valid data.
>>
>> > Just show the received frequencies at two
>> > successive sites and their times and what the sites are.
>>
>> The method requires determining the time of zero
>> crossing by taking a first derivative (difference
>> column), eliminating bias by selecting a sample
>> period which is symetrical about the crossing
>> and then taking a second difference and doing a
>> least squares fit to find the result. You do all
>> that independently for each of the two receive
>> sites to get two times and then compare with the
>> site longitudes.
>>
>
> Maybe your unnecessary elaboration of details keeps you from seeing
> the basic error in your approach.
Maybe ignoring important details is why you
keep getting errors. If you don't use some
method to isolate the one-way data from the
three-way and get rid of the settling effects
they will mess up your results.
> Maybe you are assuming that the craft
> is in the equatorial plane or not far from it and that is an
> unwarranted assumption.
It would be, but I don't.
>> >> > But there may have been some standard delays before the
>> >> > transponder
>> >> > was used at the earthsite where two way communications were being
>> >> > used
>> >> > as opposed to where one way comunications were used.
>> >>
>> >> There is often a period at the start of a contact where
>> >> the system worked in one-way mode before changing to
>> >> three-way but these are easily distinguished as I said
>> >> but that isn't a delay of the signal. There is no way
>> >> to introduce a delay since the transponder is simply a
>> >> PLL used to lock the downlink frequency to a multiple
>> >> of the uplink. We are not talking about data here.
>> >
>> > I understand how it might have been used the question is how it was
>> > used.
>>
>> It has two ways of being used, "On" and "Off". Either
>> you lock the downlink signal or you don't, but there
>> is no question of software delays because it is a hardware PLL that works
>> in real time.
>
> I am not absolutely sure there might not be a delay circuit
If you knew how it worked, you would not say that.
Any delay will destabilise the loop. The only delay
is in the low-pass filter and is the inverse of the
bandwidth, about a second. Since the frequency
readings are averaged over a minute, it is negligible.
> and it is
> possible also that when the coding is going on the craft is not
> relaying the carrier but emitting its own carrier so that there is a
> change in the frequency being emitted etc.that contributes to the data
> that is observed.
The switches between the two modes are obvious when
you plot the graph for each contact. I plotted them
for you before:
http://www.georgedishman.f2s.com/Ralph/1987.htm
>> The delays that do exist in terms of locking on can
>> be seen when you plot the data which is why it is
>> essential that you plot graphs as I said above.
>>
>> That said, you still don't seem to grasp the fact
>> that the transponder is a way of ensuring an accurate
>> frequency for the downlink and there are no specific
>> timings being measured which could be delayed anyway.
>>
> You don't seem to grasp the fact that the craft could turn off
> the transponder when decoding is going on
You don't seem to grasp (a) it can't, that is
commanded from the ground and (b) if it was
commanded off the rate of change of frequency
would halve. It is easily visible in the example
graphs
http://www.georgedishman.f2s.com/Ralph/1987.htm
The green line at the bottom indicates off then
on and back to off, one-way then three-way then
one-way again.
> and so the next transponded
> in phase relayed uplink to downlink occurs later,or that the craft one
> way communication is substituted for the transponder relay of the
> uplink to downlink or something else involving a delay circuit.
This is a real time effect so what happens is that
is has no alternative but to revert to the on-board
oscillator, one-way mode. You can see it happening.
>> >> > And such delays
>> >> > may have caused in part the time difference between the minimal
>> >> > Doppler
>> >> > at the two earthsites that you attribute to different positions of
>> >> > the
>> >> > craft.
>> >
>> >> > The point is that the changes in frequency and the changes in
>> >> > strength of signal were both methods used to determine the position
>> >> > and velocity ten years earlier
>> >>
>> >> No they weren't.
>> >
>> > Yes they were
>> >
>> >>The methods were integration of
>> >> the frequency and the dedicated ranging system.
>> >> Pointing is adjusted to maximise signal to noise
>> >> ratio.
>> >
>> > And used in the integration.
>>
>> No, what is integrated is the measured speed (to get
>> the change in distance) and the speed is found from
>> the frequency, not the signal strength.
>
> I should have said used with the integration.
You would still have been wrong, it is not used
at all.
>> >> > and that when their false assumptions of
>> >> > the time delay led to false assumption of the position and velocity,
>> >> > these methods could be used to constantly correct the craft position
>> >> > and velocity.
>> >> > The basis for the correction was that the predicted velocity and
>> >> > positions were not the same as those implied by successive observed
>> >> > signal strength and doppler shifts.
>> >>
>> >> The basis was that there was a difference between
>> >> integrated Doppler and the ranging measurement.
>> >
>> > This is essentially what I said in a much clearer way. I am not
>> > trying to impress morons with jargon but to say what happened as
>> > clearly and simply as possible.
>> > I am not trying to bring up red herrings all the time.
>>
>> Then listen to what I am saying because it is
>> not "essentially what I said", perhaps I still
>> wasn't clear. There are two methods supported
>> by the craft that they can use for navigation
>> and the JPL software normally makes use both
>> in fitting the most likely trajectory. The first
>> is the measured downlink frequency which gives
>> the speed through the Doppler effect. The second
>> is a direct measurement of the distance ('range')
>> where they modulate the carrier with a pseudo-
>> random pattern and measure the time delay for
>> that to be returned.
>>
>> They do not use signal strength as a pimary
>> source of navigational information.
>>
> But they do use it and they need to use it to correct the mistakes
The 'mistakes' are in range only, they _cannot_
use signal strength to correct range data. The
difference in signal strength caused by a few
metres error at tens of AU is ludircrously small.
> made by these other measurements which assume the wrong light speed
> delay and to confirm what subsequent doppler and range show. I have
> been told this by those working at the antenna sites.
They can use pointing data for near-Earth craft
where slight errors in altitude have a significant
impact on orbital period, and they adjust the
pointing to maximise the signal strength to account
for atmospheric refraction effects on the deep space
craft, but they cannot get range from the antenna.
>> >> The
>> >> error was of the order of 12m IIRC (it's in the paper)
>> >> whereas your claimed error would be millions of km.
>> >>
>> >
>> > Not true. The error is increasing to twice this and my claim is just
>> > this.
>>
>> Maybe my memory is playing tricks but I thought
>> you said it was tens of Hertz in a contact period.
>> Their figure of 3Hz in 8 years is equivalent to
>> about 25000km so the frequency error you claimed
>> would be an enormous distance. Maybe we are talking
>> at cross purposes somehow.
>>
>> > The point is that false speed of light delay assumption required
>> > constant corrections and you are covering up this basic point why?
>>
>> The corrections were comparable to the noise on
>> the signal so entirely expected.
>>
> The corrections I showed you and you acknowledged are much larger.
> They may be due to gr effects but this is not clear at all especially
> the 90 Hz you refer to or to the .004Hz you refer to.
Be careful, those are not comparable numbers. One is
an error in the amplitude of the diurnal sine wave
while the other is the rms of a gaussian spread.
> There is no time
> delay so the curvature of spacetime does not apply
That is nonsense, the 'curvature of spacetime' is
another name for gravity and has nothing to do with
any transponder delay.
> and it is not clear
> how changes in the infinitessimally faster reception of the carrier
> frequency received on the craft minute by minute as the craft gets
> further from the sun and jupiter will cause changes in the doppler each
> minute so that increasing errors on the order of 10Hz are observed. And
> maybe changes in the transponder phase matching compensates for this.
The transponder is phase locked, not phase matched.
Phase matching refers to selecting pairs of components.
Come on Ralph, we went over the concepts of PLL
operation by email for weeks some time ago.
>> >> > Ten years of such corrections could have been done but I gather
>> >> > some
>> >> > trial and error approach to fitting a trajectory was used instead
>> >> > ten
>> >> > years or twenty years later. Perhaps the former data and method were
>> >> > thrown out because the cumulative error was unbelievably great due
>> >> > to
>> >> > the false speed of light delay assumptions.
>> >>
>> >> The cumulative error was a few tens of metres,
>> >>tens of nanoseconds in a round trip of hours
>> >
>> > There is no round trip of hours and the error is every second.
>>
>> No, that was the cumulative error.
>
> derived from the error every second.
Per day on one craft, cumulative on another
IIRC. Bottom line is that they concluded the
two methods were consistent for those craft
that had both systems running.
>> >>> But you are saying that my analysis predicted frequencies that
>> >> > were further from the observed than the conventional prediction.
>> >> > True,
>> >> > > but
>> >> > this was because my analysis required the false initial positions
>> >> > and
>> >> > velocities based on the conventional light speed delay assumptions.
>> >>
>> >> Then the correct approach is to alter them by
>> >> trial and error until you eliminate the error.
>> >> Your solution is then no longer dependent on
>> >> those values, they would just be the seeds for
>> >> your iterative solution. That is what I meant
>> >> by saying _you_ needed to apply a correction
>> >> to _your_ solution.
>> >>
>> >
>> > That is what you need to do if you have nothing better to do and
>> > want to ignore all of the evidence that the anomaly no matter what is
>> > tried, is much larger than claimed.
>>
>> For goodness sake Ralph, the published error is
>> the remainder after applying _conventional_ theory,
>> I am talking about _yours_ ! Try to get a grip on
>> this.
>>
> You get a grip. I am talking about the conventional model
I know, but you are criticising comments I made
about _your_ model as if I had been talking about
the conventional model. That's why your criticisms
are irrelevant to what I said. Let's drop it, what
I said was many posts back and you're not going to
get out of this way of thinking now.
> and its
> error. My model and the error difference between my model prediction
> and the observed Doppler is a red herring because it assumes the wrong
> initial position and velocity and so must be larger.
> Your attempt to use this to show that the nearly instantaneous
> hypothesis is disproven is disingenuous.
>
>> > In fact the correct approach is to
>> > go back and look at all of the corrections that were made from the day
>> > of launch and see that if the right speed of light delay assumptions
>> > were made there would have been no need for constant corrections.
>>
>> No that isn't the correct approach. In fact it is
>> a completely unworkable method because all the
>> flight parameters would be changed by using your
>> suggestion.
>
>
>
> You don't mean flight parameters, you mean position and velocity
> estimates based on wrong speed of light delay assumptions ...
No, I mean the derived ephemeris parameters.
> And what I am proposing is that if the correct speed of light dealy
> assumptions were used with the same received carrier frequency doppler
> shift data that there would be no need to make repeated corrections.
And I am saying that if the JPL ephemeris was
derived using the wrong assumption, it is useless.
The only way to get rid of that error is to start
again using the raw data and derive a new ephemeris
using your assumptions about light time delay. That's
what I just told you:
>>The only thing that can be done is to
>> start with the raw data files which you have and
>> fit a brand new trajectory to them.
>
>
>> > Of course this is not available although the records were obtained.
>> > Why?
>> > What is your explanation for this?
>>
>> Because they don't apply specific corrections the
>> way you seem to think they do. The approach is to
>> produce a brand new fit each time gradually
>> discarding older data because they are aware that
>> various forces like the solar radiation, inter-
>> planetary dust and so on cause the craft to wander
>> from its previous course very slightly.
>>
> So they do apply specific corrections the way I think
> they do and the way you think they do.
No they substitute a new ephemeris which fits the
current arc best. If they applied a correction to
the previous values, it would have the effect of
trying to fit a single trajectory to the whole
arc but they know there are external influences
that would result in poorer performance if they
did that.
> The wandering is slight minute by minute but it is cumulatively
> quite large. This wandering would not require unjustifiably large
> estimates of solar radiation and planetary dust effects if the correct
> assumptions about the speed of light were made
Prove it, do the work. I know you can't because
the form of the difference cannot be nulled out
that way but that sort of negative can't be
proved so the ball is in your court.
>> > The point is that no matter what nasa does with the data and the wrong
>> > speed of light delay assumptions they get errors which imply the
>> > planets should have fallen into the sun long ago(even though Anderson
>> > claims it is smaller than the data shows if you take into account the
>> > miniscule curvature of space time and throw out "bad" data)
>> > Since this is so obviously the case it is foolish of you to pretend
>> > otherwise and play little mind games.
>>
>> No mind games, I am just trying to educate you
>> in the basics of navigation techniques and how a transponder works
>>
> I understand how it works but you don't understand how it could be
> used in this context or you are pretending that you don't..
I know how to see the effect of switching
it off in the graphs and I have shown you
on the web page. It could be used that
way be it is obvious that it wasn't.
>> > You say the earth is not in orbit
>> > but is only rotating in the equatorial plane and that the rotation
>> > axis is perpendicular to the rotation axis and that the craft is also
>> > in the equatorial plane.
>> > Of course you are not literally saying this,
>>
>> Thank goodness you added that, it's a load of crap!!!!
>>
>> You have repeatedly accused me of not taking the
>> projection of the velocities into account but that
>> is what the projection does. We have to project
>> onto the plane in which the site is rotating and
>> of course that is perpendicular to the axis hence
>> parallel to the equatorial plane.
>>
> But the line to the craft is not in the equatorial plane
> or as near to it as you assume making the false speed
> of light delay assumptions.
I don't assume it is anywhere near the equatorial
plane. In fact I know it was close to the ecliptic
but even that is unnecessary because I use the
zero crossing method which eliminates the dependence.
I have told you that repeatedly.
>> > but your claim that the
>> > craftsite lines from two different earthsites 60degrees apart to
>> > wherever the craft is are equivalent to a line from the equatorial
>> > plane at the longitude of the site to the same but unknown craft
>> > position at some unknown time during the period of exposure of each
>> > site to the craft.
>> > And that this equivalence is explained by the fact that Acos(0)=A.
>> > Such mind games
>>
>> It is basic Geometry Ralph, project a circle and you
>> get an ellipse. Take the x component and it is a sine
>> wave whose amplitude depends on the angle used in the
>> projection but the phase of the resulting Doppler is
>> unaffected.
>>
> But we are not projecting a circle so as to get an ellipse
Yes you are, you are projecting the circular motion
of the sites around the axis of the Earth onto a
plane containing the craft (not the equatorial plane!)
> or a
> straight line, we are projecting a line representing the orbital
> velocity at a site and a line representing the spin velocity at
> the same site. These lines are projected through unknown
> angles onto the unknown craft to site lines at these sites.
> As the earth spins, the spin velocity at each earth site will
> at some time at each site be perpendicular to the unknown
> earthsite to craft line at each of these sites but the total
> velocity line at these times at these sites may not be
> perpendicular.
Right, that's why we extract only the radial
component by using the zero-crossing. It is
a crucial part of the method.
>> At each site what you say is correct, and the method
>> then tells you the celestial longitude of the craft.
>> You rproblem arises when you compare them because
>> the location of the craft necessary to explain the
>> Doppler at one site is 26 degrees away from the
>> location needed to explain the readings at the other
>> site. That combination is what disproves your theory.
>
> This 26 degrees is due to your assumption that the
> the orbital velocity has no effect.
Nope, the orbital velocity does have an effect
but we know it is less than 1 degree. That leaves
over 25 degrees unaccounted for.
>> > Also that the projection of the orbital and rotational earthsite
>> > motions onto the lines from these earthsites will be minimal when the
>> > vector sum of the orbital and rotational earthsite motions produce a
>> > vector perpendicular to one of these lines.
>> > Clearly we cannot substitute for these two lines and two
>> > projections,
>>
>> > these other lines and projections on them you want to use
>> > to make your cockamamie argument.
>> >
>> >
>> >
>> >>
>> He said and I have said that there was no possibility
>> of simultaneous transmission from one site when receiving
>> at the other with a 12 hour round trip. He also said that
>> there were instances of reception without matching
>> communication from the receive site and that is true.
>>Beyond that
>
> No need to go beyond that. He repeated exactly this as part of a
> diatribe
> saying nearly instantaneous communication with the craft was
> therefore impossible and that the communication had to have been
> relayed from another site at a much earlier time. As far as I am
> concerned this was a lie repeated again and again.
There is an example of that in the data we are
looking at so it is true, not a lie.
>>If you want to asking people of lying, cite
>> the post Ralph, put up or shut up.
>
> The posts were many and many repetitions of the same lie that you
> describe and
> anyone can see them all by looking at the archives.
What I said above is true, not always but sometimes.
>> >> >> > The
>> >> >> > point is that Anderson's claim of an anomaly is correct but his
>> >> >> > claims
>> >> >> > as to the size of the anomaly are contradicted by the data which
>> >> >> > says
>> >> >> > the anomaly is much greater than claimed.
>> >> >>
>> >> >> No it isn't, his value is accurate acording to
>> >> >> conventional theory. You have never calculated
>> >> >> the anomaly using conventional theory, only
>> >> >> your own version.
>> >> >
>> >> > No I calculated it according to both.
>> >>
>> >
>> >>You have never done a GR analysis,
>> > .
>> >
>> > This is a red herring.
>>
>> No, it is fundamental. There are significant
>> relativistic aspects that you have not taken
>> into account which will give errors larger
>> than the anomaly.
>>
>> > The magnitude of the difference between gr and
>> > Newtonian analysis on the received frequencies is on the order of
>> > .0001Hz
>>
>> No it isn't, I tried a comparison because you
>> used the wrong Doppler equation and it was
>> roughly 90Hz over a single contact.
>>
>> > and the difference between the observed and predicted
>> > frequencies is on the order of tens of Hz..
>>
>> Exactly.
>>
>
>
>> The difference between the GR version and the
>> measured results is of the order of 0.004Hz
>> from Craig's paper but that is the width of
>> the gaussian spread due to noise, not a
>> systematic error. For that see Anderson et al.
>>
> This is all unclear. I suspect that the relativistic
> effect is a red herring and that the changes minute
> by minute in the Doppler caused by this are negligible.
>
>
>> The fact remains that you have not done a GR
>> calculation so saying "I calculated it according
>> to both." is simply untrue.
>
>
> Both may have referred to something else but my analysis
> did assume it was negligible. But you are right. This is all unclear.
What is clear is very simple, you have not done the
relativistic calculation so you are only guessing.
Both Anderson et al and Mrkwardt have done the
calculations in full and they get comparable results
so your suspicion is incorrect.
Relativistic effects are undoubtedly small but given
the sensitivity of this study (milliHertz on a
gigaHertz carrier) even these show up. Do a quick
back-of-envelope check Ralph, what is the difference
between the Newtonian moving source, moving observer
and the relativistic Doppler for 2.2GHz and a speed
of 42km/s ? It will only take you a few minutes but
you might then appreciate the significance.
> I suspect that the relativistic effect is a red herring and that the
> changes minute
> by minute in the Doppler caused by a more rapid reception of
> the carrier by the craft, as the craft gets further from jupiter and
> the sun,
> is compensated for by the phase adjustment in the craft transponder.
Please go back and read how a PLL works before saying
things like that Ralph.
> The delay in carrier reception associated with light bending due to
> the sun
> only occurs when reception is during the day so does not explain the
> systematically increasing effect shown by the data.
Nor would the _fixed_ phase difference produced
by a transponder. A fixed phase shift doesn't
change the frequency.
>> <snip fusion>
>> > Your insistent opposition to my theory with red herrings and vague
>> > illogical arguments that have been shown to be just that
>>
>> Ralph, I can't help it if the geometry needed for
>>you to understand this is too complicated.
>>The only thing I can do is to ask you to treat it
>>like a homework exercise,
>
> Don't be so foolishly pretentious. I have explained
> to you what is wrong with your assumptions ...
And I have explained how the use of zero-crossing
eliminates your spurious objections. I do not
make the assumptions you think I do, instead I
developed a method which removed the effect of the
craft being out of the equatorial plane. I have
done all I can to help you follow how it works.
George
|
|