AA77 FDR Data, Explained

[Anti-sophist]

Don't forget VLSI concepts:
- Wiring and Interconnect: Elmore Delay, capacitance (fringing, interwire, cross-talk), low-k dielectrics, reduced swing, resistance, electromigration, inductance
- Gates: nmos, pmos, transient response, propagation delay, sizing, fan-in/out, subthreshold leakage
- combinational logic: logical effort, ratioed logic, pseudo-nmos, DCVSL, pass-transistors
- dynamic logic: charge sharing, backgate coupling, domino logic, differential domino logic
- sequential logic: latches, (positive/negative) edge triggered registers, C2MOS, TSPC, Schmitt Trigger, VCO

Etc, etc...just for starters

Sigh these are digital systems, and you haven't even built a NAND get yet, let alone a multiplexer or an ALU.... Better get to work. This document is gonna be 1000 pages before it's done.

 

[Skibum]

This document is gonna be 1000 pages before it's done.

Unfortunatley, it will be 999 and 15/16ths pages past the attention span of the CTs.

 

[gumboot]

Unfortunatley, it will be 999 and 15/16ths pages past the attention span of the CTs.

And that's only if the cover has a nice picture on it...

-Gumboot

 

[UnderTow]

You and JDX's dopey analysis is based on a misintrepretation of the CSV file.. that dopey intrepretation DOES assume bitrate and phase, inadvertently, by consequence of the way you are reading the file.

Ah see. Now we are getting somewhere. Please point out which dopey analysis you are referring to?

Also, please confirm or dismiss whether you believe every single crash investigation which uses FDR read outs also contains these error which you have designed. Because your paper could be describing any FDR and Tabular plot from any airplane right? Do you believe all investigations which use downloaded data from an FDR and make a table and graph out of it are flawed?

My initial intent was to properly analyze the data and debunk the variety of dopey conspiracy theories that abounded.

You never got to the 1st part, instead you waxxed on with your own assumptions and theories of errors. And the 2nd part is also incomlete becuase you have yet to specify which dopey theory you have debunked, other then to theorize how ALL fdr readouts beyond the raw file are flawed hopelessly and forensicly worthless.

You should dress up your paper and send it to the Boeing, L3 et all becuase I'm sure they could use the help in fixing these erorrs you have created.

Can you show one example from any FDR analysis of the massive errors you have given in your paper? After all, you "gathered up all the publically available flight-data-recorder information" right.(?)

I don't think your an idiot and I never said that. I think your paper is mush.

I'm glad we have been able to have this talk, but your intent seems to be more about debunking dopey stuff to gain Randi high-fives instead of actually examing the data. Regardless, after me and other "poor software engineers" get done doing another read-out from the raw file, I'll be sure to include you in the peer-review.

 

[Anti-sophist]

Also, please confirm or dismiss whether you believe every single crash investigation which uses FDR read outs also contains these error which you have designed.

It's been answered. Repeatedly. Just to clarify, I don't think this investigation even suffered from these errors. My claims apply solely to the CSV file, which is not what the investigation is based on. The real invesitgation used the real data. You aren't. These errors apply solely to your amateur investigation based on incomplete data.

Because your paper could be describing any FDR and Tabular plot from any airplane right?

Any derived CSV file in the format of the tabular CSV file from flight 77, yes. Luckily, real engineers use the real data.

Do you believe all investigations which use downloaded data from an FDR and make a table and graph out of it are flawed?

Of course not. I believe that basing forensic analysis on data meant to be plotted, however, is flawed. Luckily no actual engineer would ever do that. They'd download the data in its raw form, and use that.

And the 2nd part is also incomlete becuase you have yet to specify which dopey theory you have debunked

All forsenic analysis of the FDR data based on the CVS file, that I've seen form CTers, has flaws that are at least, in part, explained by this paper. I admit that this paper does not completely debunk all parts of all analysis (ie, it does not explain instrument error from an altimeter).

other then to theorize how ALL fdr readouts beyond the raw file are flawed hopelessly and forensicly worthless.

Lie.

You should dress up your paper and send it to the Boeing, L3 et all becuase I'm sure they could use the help in fixing these erorrs you have created.

Same, tired, strawman that you repeat every single post and I've answered 100 times. When are you going to stop repeating this tripe? When are you going to even comprehend what I'm saying?

Can you show one example from any FDR analysis of the massive errors you have given in your paper?

Real FDR analysis doesn't have these errors. Why? Actual engineers can correct this error given the FDR data. You can't for two reasons. The first is that you don't have the full FDR data. I'll let you guess at the second.


Are you going to stop claiming that I am saying FDRs are flawed? Nope.
Are you going to stop claiming that I have found errors in actual FDR data? Nope.
Are you ever going to understand that the CVS file isn't the raw FDR data? Nope.
Are you ever going to understand these errors exist solely because the CVS file is not raw FDR data? Nope.
Are you ever going to understand that engineers aren't using the CVS file to do the actual analysis? Nope.
Are you ever going to understand the purpose of the CVS file wasn't to do actual analysis? Nope.
Are you ever going to explain the footnotes that clearly show I am right? Nope.

What are you going to do? Keep repeating the same stupid strawmen, forever? Showing you are not only completely incapable of understanding the science in the paper, but it's intent, as well? Continue to keep claiming that I am saying things that I am clearly not saying? Forever?

As always, I am open to correction.

 

[UnderTow]

Therefore you are saying that the NTSB engineer who created this CSV file did not know what she was doing and did not correct for these errors. Therefore any other report this engineer has done is also flawed becuase they took the raw file, converted it to engineering units, made a table out of it, and created an assumption like this:

The SSFDR data indicated that at 1031:37 DFDR elapsed time, while the airplane
was descending through a pressure altitude of 5,984 feet, maintaining a magnetic heading
of 139 degrees, and flying at an indicated airspeed of 191.47 knots, roll attitude values
consistent with a left wing down motion were observed. Engine data at this time indicated
N1 values to be 30.3% and 31.4% for the left and right engine, respectively.

Which is represented in one Row in a Table like this:

10:31:37 5984 191.47 139.53

This is not AA77.

Again, a Table is based on Real data. They use the Table to investigate the event.

I believe that basing forensic analysis on data meant to be plotted, however, is flawed. Luckily no actual engineer would ever do that. They'd download the data in its raw form, and use that.

"Data meant to be plotted". You mean that taking a raw FDR data file, converting it to human readable form (table or graph) an using that table or graph is "flawed"?

 

[Anti-sophist]

Therefore you are saying that the NTSB engineer who created this CSV file did not know what she was doing and did not correct for these errors.

Lie.

This is not AA77.

Source.

Again, a Table is based on Real data. They use the Table to investigate the event.

Not the CSV file from AA77, they don't. This is so wrong it's funny.

"Data meant to be plotted". You mean that taking a raw FDR data file, converting it to human readable form (table or graph) an using that table or graph is "flawed"?

No, the CSV file is more than "converted to human readable" form. It has been processed beyond that. A tabluar version of the FDR, just "converted to human readable form" does not look like that CSV file from AA77. Try again. This has been explained to you, at least twice, and will not be repeated a third time.


PS. Still waiting for your explainations of the footnotes.

 

[UnderTow]

What are you right about?
Which CT analysis are you referring to?

How does this NTSB fdr-sourced table differ from any other table in any other instance?

The NTSB investigation used the real data.
THE CSV file made by the NTSB used the real data.
The Graph made by the NTSB used the real data.


Unfortunately, you can no longer edit your orginial post, so I don't know how your going to correct that.

 

[Anti-sophist]

What are you right about?

Huh? Everything?

Which CT analysis are you referring to?

Already been answered.

How does this NTSB fdr-sourced table differ from any other table in any other instance?

Already been answered.

The NTSB investigation used the real data.
THE CSV file made by the NTSB used the real data.
The Graph made by the NTSB used the real data.

All true. The CSV file, however, is not EQUAL to the real data. It is not just the REAL data presented in a different form. It is _based_ on the real data, but some information has been lost in the processing.

This has been explained, to you, repeatedly.

Unfortunately, you can no longer edit your orginial post, so I don't know how your going to correct that.

Correct what? Why would I edit my post? All I need to do is copy/paste the same answers to the same stupid questions, over and over and over.

PS. Still waiting for your explaination of the footnote.

 

[UnderTow]

Source?? HAha..
"gathered all publicly available information"
I guess it should be in one of your files then.

Quote:a Table is based on Real data. They use the Table to investigate the event.

"Not the CSV file from AA77, they don't."
How can you prove this? What make this NTSB CSV Table different from any other Table?

 

[Anti-sophist]

Source?? HAha..
"gathered all publicly available information"
I guess it should be in one of your files then.

You claimed it was not from AA77. I didn't download every single piece of information ever in the history of the universe.

Source it, or go away. Are you scared of giving me the full data? Why?

What make this NTSB CSV Table different from any other Table?

This has been answered at least twice, and will not be repeated. The short answer is that the CSV file _cannot_ be raw FDR data presented in tabular form. The CSV file is _not_ what raw, tabular, FDR data looks like. In order for raw FDR data to be put into the CSV format, it would need to be processed. The details on why this is so have been addressed, with you, repeatedly, in this thread.

PS. I must have missed your explaination of the footnote... did deal with it yet?

 

[UnderTow]

Huh? Everything?

Only in your mind

Already been answered.

Lie

Already been answered.

Lie

wow, that's nice an easy just to type 'lie' and forget about it

All true. The CSV file, however, is not EQUAL to the real data. It is not just the REAL data presented in a different form. It is _based_ on the real data, but some information has been lost in the processing.

Oh right. In other words at the 46'th second, you would theorize that the Speed parameter could be from 2 seconds ago (or in the future) and not in sync with any other parameter within the same second. That the NTSB doesn't know how to use a frame descriptor to compute the raw parameter, that L3 doesn't know the 573/717 Requirments in order to keep all this data in sync, and the Boeing doesn't know who to program a 429 bus to calculate for the delay from the instrustment to the FDAU.

Becuase you have designed all these systems by yourself already and if it wasn't for the "poor software engineer" messing everything up, then all these parameters in tables wouldn't be so messed up.

 

[Matthew Best]

Reading that last UnderTow post suddenly brought the phrase "Look at this collapse" to mind for some reason.

 

[DavidJames]

Reading that last UnderTow post suddenly brought the phrase "Look at this collapse" to mind for some reason.

It seems clear to me UT doesn't understand what's "below" the csv file. He's missed AS point completely, whether it's an honest mistake or not, he doesn't get it.

 

[mcMike]

Anti-Sophist could you please explain.

I am trying to understand more deeply this FDR-issue. Here's couple of things that confuses me.

1. Why is the flight-path in NTSB-animation in wrong position in final approach? I checked the CSV-file and didn't find any "GPS-coords" as I expected so I assume that the flight-path is reconstructed with external radar-data or via inertia calculation? Or whats the deal?

2. What is the thing with the heading issue? About last 20 seconds are indicating heading to 68-70 degrees. The last 10 seconds 69.6-70 degrees. The official flightpole-path is about 61 degrees. The "eyewitness-path" is over 80-degrees. So the heading in FDR-data seems to be in the middle of those. Am I missing something or how to interrept the heading? With 70-degrees it seems to be impossible to hit the poles.

I used GoogleEarth and overlaid image with both paths into it and drew some lines in Photoshop to determine the headings. Path is from 911-strike, the image name flight_path.jpg (can't post links)

Thanks.

 

[Anti-sophist]

Oh right. In other words at the 46'th second, you would theorize that the Speed parameter could be from 2 seconds ago (or in the future) and not in sync with any other parameter within the same second.

Fairly true, except for the ridiculous bolded part. Again, you'd need to define "in sync". All measurements have some error. Saying two things are "in-sync" implies within some error tolerance, they are sync'd.

That the NTSB doesn't know how to use a frame descriptor to compute the raw parameter, that L3 doesn't know the 573/717 Requirments in order to keep all this data in sync, and the Boeing doesn't know who to program a 429 bus to calculate for the delay from the instrustment to the FDAU.

Of course they know how. The CSV file doesn't contain this information. The real engineers know how, and do, when they do real analysis.

Again, this is for plotting data. Listing every single data point and it's associated measured time is unnecessary to produce plots at the resolution required. The plot doesn't require the full precision, so the CSV file isn't in the full precision.

Becuase you have designed all these systems by yourself already and if it wasn't for the "poor software engineer" messing everything up, then all these parameters in tables wouldn't be so messed up.

Wrong, again. It has to do with intent. Generating data within acceptable error tolerances to do the plots attached to the NTSB report is not the same as generating data to base forensic analysis of the final 2 seconds of flight 77.

Those are different problems and require different amounts of error tolerance. The CSV was designed to solve one, not both.

PS. The mysterious footnote that confirms everything I've said remains.. unexplained. I should start counting how many times I've asked you to explain it.

 

[Anti-sophist]

I am trying to understand more deeply this FDR-issue. Here's couple of things that confuses me.

1. Why is the flight-path in NTSB-animation in wrong position in final approach? I checked the CSV-file and didn't find any "GPS-coords" as I expected so I assume that the flight-path is reconstructed with external radar-data or via inertia calculation? Or whats the deal?

2. What is the thing with the heading issue? About last 20 seconds are indicating heading to 68-70 degrees. The last 10 seconds 69.6-70 degrees. The official flightpole-path is about 61 degrees. The "eyewitness-path" is over 80-degrees. So the heading in FDR-data seems to be in the middle of those. Am I missing something or how to interrept the heading? With 70-degrees it seems to be impossible to hit the poles.

First, let me reiterate an important point. This paper and most of the issues discussed in it was created before this "new" analysis from JDX came into being. I haven't looked, in depth, at the heading issues, like I have the altitutude, acceleration, and roll angles.

My initial reaction is that to look at this, in detail, is a waste of my time, because none of the CTers seem to have any intellectual honesty, to begin with. JDX and UndeTow continously misrepresent my position and argue with claims I am not making. They've failed every single attempt at even comprehending my position, let alone attacking it.

That being said, I'm willing to look at the heading thing in more detail, but my tolerance for this strawman and intellectually dishonest gibberish is growing short.


I have two questions, first, that I'd like you to answer:

68-70 degrees.
69.6-70 degrees.
The official flightpole-path is about 61 degrees.
The "eyewitness-path" is over 80-degrees.

Are these all relative to north? True north or magnetic north?

 

[mcMike]

I have two questions, first, that I'd like you to answer:
Are these all relative to north? True north or magnetic north?

Good point. Didn't realise that at all.
The heading from FDR is seems to be magnetic (MAG HEADING CAPT (DEG)).
Unfortunately I don't know which is the "North" in GoogleEarth. Anyone?

What is the difference between these two? In degrees?

And what about the explanation for flight-path inaccuracy in animation?

IMHO, looking this into detail is definetelly not waste of your time. And just for the record. I am not "attacking you". We're on the same side. I'm just looking for an explanation for myself since I want to understand the issue here.

ps. I sent the imageshack filelink via PM to you.

 

[Anti-sophist]

Good point. Didn't realise that at all.
The heading from FDR is seems to be magnetic (MAG HEADING CAPT (DEG)).
Unfortunately I don't know which is the "North" in GoogleEarth. Anyone?

I am almost positive google is "true" north.

What is the difference between these two? In degrees?

In DC, according to http://www.ngdc.noaa.gov/seg/geomag...n;jsessionid=4C91FDDF4A0D1C1FE6F0EA567DAF2F53 it is 11 degrees (EDIT: to the west, which makes it line up almost perfectly).

That's 2 minutes of searching, so don't take my word for it.

And what about the explanation for flight-path inaccuracy in animation?
ps. I sent the imageshack filelink via PM to you.

I'm at work. I'll get in more depth when I'm home. We also need to correct heading for wind to get the bearing.

 

[mcMike]

I am almost positive google is "true" north.
it is 11 degrees.
I'm at work. I'll get in more depth when I'm home. We also need to correct heading for wind to get the bearing.

Thanks. So difference between true and magnetic north alone and maybe windspeed / direction can explain the heading. (if the differences are in correct direction )

You really should take a look of this in detail because that could debunk the "eyewitness-path"-theory at once. If we can trust the heading from FDR it is highly unlikely that the plane flew the "wrong-path" in 20-degree yaw.

I believe there is +-1 degree accuracy in the image I posted to you. I tried to be very precise with layer-aligment and rotation.