Good point. Like Noam Chomsky, who is apparently a hero to some twoofers, said they would have to be insane to try something like that on the off chance of it actually working.
-
Security incident: ISF was recently accessed by intruders. Please change your password, and change it anywhere else you used it. Read more
AA77 FDR Data, Explained
- Thread starter Anti-sophist
- Start date
-
- Tags
- aa77 flight data recorder
Good point. Like Noam Chomsky, who is apparently a hero to some twoofers, said they would have to be insane to try something like that on the off chance of it actually working.
The only values stored as DME are x.0, x.2, x.5, x.8 (or x.7 depending on who decoded it and how their algorithm rounded it)
The fact is the plane can be 1.75 NM from DCA and the DME can be stored as 1.5 DME in the FDR; THIS IS A FACT YOU CANNOT PROVE ME WRONG; ASK A MATH TEACHER TO HELP YOU UNDERSTAND THIS SIMPLE FACT! I AM USING YOUR 0.1 NM ACCURACY
Remember the accuracy in practice of DME for 77 would be about 0.23 NM, this is reality, the 0.1 figure is a spec for a DME set not proven to be in 77! The new DME is more accurate, but 77 first flight was when?
Prove 4 seconds could not be in a buffer outside the FDR secure chip in 77 FDR system. Show me!
MacGyverS2000 you may want to ask him to stop using your name in what appears to be support of his theoris.
This link in particular makes it appear that you are supporting him regarding the light pole matter
http://www.internationalskeptics.com/forums/showpost.php?p=4177149&postcount=345
The NTSB decoded the file. They have all the software and hardware
needed to forensically study the data. x.8 would be incorrect for the
NTSB decode as clearly shown by looking at the file.
You want me to believe your cries over the Manufacturer and NTSB decode?
If the equipment has the ability, accuracy and capability to record, 1.5 DME
and 1.7 DME, why would the data be stored as 1.5 DME if the aircraft is
1.7 nautical miles from the beacon?
Take it up with the manufacturer (Rockwell-Collins), or the NTSB. I'm not
buying your theory. The MFG specs are on the site.
L3 Products are certified to EUROCAE specs. Each certification for the FA2100
is listed on their web site.
If you have a problem with the data performance details, take it up with
EUROCAE.
The FDR is certified and by design cannot have data floating in a buffer for
four seconds. Once again, this is by design. Take it up with L3, or EUROCAE,
not me.
The proof is in the spec. There is no need to debate this point.
Yes, the p4t decode rounds different and has 1.8 DME, you know x.8, and the NTSB has 1.7 DME rounded from what value; can you guess?
4617 feet from the Pentagon at 1.724 NM from DCA! Which can be stored as 1.5 DME. Wowzer, this is 6 seconds to go from :43, making impact at :48. And this makes you unable to understand math, DME, and reality.
Why! 77 is at 1.724 NM, the DME senses 1.624 due to the accuracy you give me (BTW the real DME for the 77 type system has a close in accuracy of 0.23, but for you woo I give you 0.1 DME for the guy with the conclusion that is a LIE). This 1.624 NM is stored by the FDR as 1.5 DME because it can only store 1.5 or 1.7 (which is really 1.75; proved by the fact the round up of p4t expert FDR decoders, and the round down with the NTSB, notice the 0.25 NM resolution is recorded as x.5 and x.0, but x.25 is stored as x.2 and 1.75 is stored as 1.7. Darn this math is confusing for p4t. Lucky for me my mommy made me go to engineering school! And then the United States Air Force Asked invited me to get my Masters in engineering to work on "Star Wars" with President Reagan.
Did I get close! The reason I think this is a good idea of what happen with 77, the fact DME is not updated every 4 seconds (the FDR does DME updates haphazardly, but this value is brand new and at 1.5 with the real errors involved, this matches a RADAR return see in blue. The time is off 4 or 5 seconds from 77's time, but it matches where 77 would be at the time 1.5 DME was placed in the FDR.
See!
The yellow line from the bottom right is a 1.724 NM line from DCA, I am saying that 1.724 NM is stored as 1.5 DME when you have an accuracy of 0.1 nm. But as I say the DME for 77 was actually only accurate close in to 0.23 NM, but I can live with 0.1 NM for your whinny rant about this, you don't understand the math in the first place.
Review: For my illustration, 77 is at the actual position of 1.724 NM from DCA.
With the accuracy of the DME receiver from Turbofan we can have 1.624 NM sensed by the DME receiver.
Because the FDR can only store .0, .25, .5 or .75, the 1.624 value sensed at 1.724 NM from DCA is stored as 1.5 DME!
Questions?
So it is possible for 77 to be 6 seconds from impact when 1.5 DME is stored. Math is required.
Turbofan, have you used DME to fly Heavy Jets into airports around the world? I have and 0.1 DME accuracy is possible, but not practical. I have showed you what AIM said for DME accuracy at the time, you fail to listen! In the real world what I just went over for DME is very possible and the most probable senario. BTW, the FDR was found in the Pentagon making your conclusion a lie. This math course is a waste of time you don't get it.
Someone can explain it better than I can. Maybe everyone can.
You have no practical experience in understanding the spec, and you can't prove 77 FDR had to conform to a specific spec, you have not presented a case for the FDR installed in 77 having to conform to anything. You lack substance. You mainly repeat junk from p4t who never look up this stuff they have to be called and then finally they post something.
The DME receiver in 77, you have not proven the accuracy of 0.1 NM was for the receiver in 77, or it met the specification at the time. BUT, I can use your accuracy, and I did use your accuracy to prove 77 can be 6 seconds away when 1.5 DME is stored! DARN, anyone with a grade school education can figure this out, but me, I had to go to engineering school and get a Masters! What a waste.
So recap, I used your 0.1 NM accuracy and found 77 can be 6 seconds out!
The blue dot is a radar return showing where 77 was at about 43, think the time is off 4 seconds for that radar! Oops, confirmation of 77 being where I think it is! Wowzer.
Funny, you said the working copy of the NTSB animation confirms the impact time! LOL
Good work
Last edited:
The NTSB is the file that is reviewed for forensic analysis. That is the file
in question. That is the government released file that everyone has access to!
Like I said before, the equipment has the capability and accuracy to record
1.7 nm (and it did), so why would it record 1.5 DME if the plane was 1.724 from
the beacon?
Also recall the aircraft traveled an additional two seconds after the 1.5 DME
point at a speed of 530+ MPH. Therefore the aircraft would be much closer
than 1.5 DME at the next update.
Study the NTSB released data and you will see this. Everyone can access
this file. There is no need to debate this point AGAIN.
See reply above.
You are using R02, and/or INS data to match DME. DME is more accurate
as it uses RF signals from an external beacon located at the airport, as opposed
to accelerometers onboard.
Once again you are using outputs from distinct instrumentation, or readouts
with different tolerances of error to match the NTSB data which uses a much
more accurate DME system.
At 530 MPH (as per FDR data), there is NO way the aircraft could be 1.5 nm from impact
based on the alleged flight path.
The rest of your post is opinion.
If you have a problem with the equipemt specs, go to the manufacturer site
and fight with them. Also ask them how they acquired the certifications if they
equipment could not perform to spec.
WRONG! The DME stored as 1.7 DME is really 1.75 DME. The value is rounded, this is why p4t read out is 1.8, the p4t round differently. But nice try, since you seem to ignore only four possible tenth values are stored in the DME, x.0, x.2, x.5, x.7, and in the p4t data x.8. Sorry, I use both data based cause I enjoy seeing all the errors in the terrorsit loyalist FDR expert p4t decode. So you stil have no idea what is going on. Good for you. Stand by your false information and deny math, accuraccy and resolution! Great.
No, the FDR only has one more second, so the plane can't fly form more than 0.5 second or the data from the next second would be stored in the FDR. Darn, you are not living by your own 500 ms rule. Are you saying the 500 ms does not count when the FDR hits the Pentagon? What?
OOPS! The DME say 3.2 DME in the p4t decode! What does this mean?
Turbofan's expert p4tr terrorist apologist have 3.2 DME stored at :43! What does this mean? HELLO
OOPS! There is no data in the p4t decode for :44! WHAT!!!!!!!!!!! The expert FDR decode team at p4t terrorist loyalist headquarters has no data in the :44 second.
Does this mean p4t are missing 500 ms of data? Why is the data damaged for p4t decoders but not the NTSB! Who has the most errors, who has the only errors in data? The p4t!
What else can I say! NO, there is not 2 seconds to impact after :43, there are about 6 seconds. NO, Turbofan can not say 77 flew for two more seconds after :43, there is no data for 2 seconds and according to Turbofan's rule of 500 ms, there has to be data in the secure chip, so the aircraft did not fly for 2 more seconds in Turbofan's world! Due to his own RULE! 500 ms! Told him he has no clue what it really means.
When 77 records the DME value of 1.5 DME, the aircraft is really at 1.724 NM from DCA and 6 seconds from impact! This is confirmed within one second by RADAR!
Sorry you have no idea what the implications of your own made up 0.1 NM accuracy means and a storage resolution of 0.25 DME means. Your lack of math skills is apparent, get some help form a math teacher.
Again for the math impaired! I am sorry, you show zero sign of understanding the 0.1 NM accuracy and the 0.25 resolution issues. Why?
When 77 is at 1.724 NM from DCA!
The DME receiver can sense a value of 1.624 because you told me it can when using your made up resolution of 0.1 NM for DME! Even though I tell you a practical value is 0.23 NM, I will use your value to prove 77 can be 6 seconds to impact at :43; oops, this means impact is at :49! Darn, I can't even add.
When the FDR sees the value of 1.624, it is stored as 1.5 DME! But 77 is really at 1.724 DME! 6 seconds to impact!
Do not say the stupid stuff you keep posting until you have a sound rebuttal. Us some math and explain the DME resolution and the accuracy!
So using math and Turbofans numbers I have shown 77 can be 6 seconds away from impact when 1.5 DME was stored in the FDR! This is backed up with RADAR data!
Hard evidence and analysis show p4t and Turbofan took my 1.5 DME needling and glom to 1.5 DME as absolute but failed.
The yellow line is 1.724 NM from DCA, and it can be stored as 1.5 DME based on accuracy of DME and the storage resolution of the FDR which you, Turbofan, do not understand! It takes math to understand this; got math?
The blue pin is a RADAR fix that is off in time 4 seconds, which corresponds to 77 DME position at the same time 1.5 DME is stored in the FDR. I got very close to the radar position using error analysis based on valid parameters.
The best part, I could be off by some, or a lot, but 77 still impacted the Pentagon on 9/11.
This is way above your head, so stop talking and posting bs and pure hearsay and opinions based on your fantasy of 77 not hitting the Pentagon. You need some real analysis which you lack the knowledge and experience to complete.
You will not post more talk and ignore accuracy, and resolution despite the fact I used your numbers to defeat your lies and fantasy. You have become a self debunking p4t false information specialist. Good luck
Certainly.
First, it will be easier to see the similarity once I clear up an apparent confusion. AA 587 did lose both engines -- physically. They were torn from the aircraft's wings shortly before impact.
However, this had nothing to do with the FDR performance. The FDR record ends over ten seconds before the engines were lost. Also, the CVR, running off the same power bus and (I believe, please check me on this) co-located with the FDR, ran all the way until impact, an estimated 13.6 seconds after the FDR record stops. The final valid data from the FDR show both engines running and responsive, and the engines can be heard running on the CVR after the FDR conks out.
So what happened? Why did the FDR quit? If it's tied to the engines and, presumably, disruption of power, how did it know ten seconds ahead of time that the engines would fail..?
Why did the FDR quit while the CVR didn't? Both have similar power systems. But there is one big difference -- the FDR on AA 587 was solid-state, whereas the CVR (a Fairchild A-100A) was analogue. This makes the CVR less flexible, but it isn't susceptible to failure modes associated with flash memory, trading those for mechanical failure modes.
We will never know precisely what happened, of course, just like AA 77. But again, the similarities are striking. It also leaves us with similar uncertainties. For instance, at what instant did the FDR actually stop recording? Just because the data record ends at a given point, that doesn't mean the FDR stopped at that precise instant, because it's working on stale data. But how stale?
Another, to my eye more telling, similarity is in the grey-scale. AA 587's FDR didn't just quit. It acted funny before it went silent. The funny data appears to be less than a second in length, but we can't be sure. That garbage data could have been written in an instant, or stretched over many seconds. It could have been written cleanly, but then garbled afterwards. It could have never been "written" intentionally at all.
But think about this. How could this happen? The DFDAU can't be held entirely to blame. It can send in-range data, screw up and send out-of-range data, send nothing at all, or send sheer nonsense. The FDR, if it is working properly, won't write garbage in the first two cases, and shouldn't write anything at all in the other two. Furthermore, in the case of AA 77, the DFDAU and FDR are the same box...
So we know, as a result of examining AA 587, that FDRs can produce strange results. We know they sometimes don't record everything we think they should. We thus have no reason to assume, as you have, that the AA 77 data record is "impossible."
It is true that AA 587 had different crash dynamics. But you simply aren't going to find anything completely similar to AA 77. Even if you did, there's no guarantee they'd produce exactly the same results. Crashes are highly chaotic events. There are many possible failures. The failure in AA 77's FDR is actually extremely mild.
This answer is perfectly reasonable, but you have to understand that "the sensor signal being generated" happens at the DFDAU, not the sensor. The sensor generates analogue data. It has to be digitized. Sometimes it gets filtered too.
A requirement for half a second from DFDAU to storage on the FDR flash is pretty easy to meet. But it's only a part of the total latency from sensor to storage.
Forget about the PFT decode, that is not the file under investigation. The
file we are concerned with is the NTSB data. You cannot compare them
because the software PFT used is not the same as the NTSB software.
If PFT released their file as FOIA to support the official story, then you would
have a point. PFT is not a government agency, they are a research team.
Start analyzing the NTSB data, that is the problem. Any further discussion
about comparing the NTSB file and R02 will be ignored.
No, the last DME was recorded at :43
The next frame would have been recorded at :45, but that is the "impact time". Do you not count the time to move into the alleged "wall"?
What does :43 have to do with 500 ms?
How can you have six seconds from "impact" if the aircraft is moving at over
780 feet per second on the 'official path', and 1.5 DME at :43?
No it's not. You are just throwing out numbers. If the plane was 1.724 nm,
DME would have recorded 1.7 period! It did not, it recorded 1.5 DME.
Also keep in mind PFT used + 0.1nm and - 0.1nm to account for
tolerances for 1.5 DME. The aircraft is still too high to hit poles according to
the NTSB data. This has been covered thoroughly.
Take it up with Rockwell-Collins. No need to debate this point AGAIN.
You are wasting everyone's time.
So using math and Turbofans numbers I have shown 77 can be 6 seconds away from impact when 1.5 DME was stored in the FDR! This is backed up with RADAR data!
Only when you skew the numbers and fake the tolerance of the avionics
equipment. That's real honest
What is your source? I will have to look into this further.
From what I recall about the system, the CVR does not tie into the DFDAU.
AA 587 had a DFDAU failure as shown by the Data Interrupt Markers which
were recorded by the FDR. THis shows that the FDR was functional and
responding, as well as recording DFDAU failure codes. The insertion of these
markers would indicate the time of DFDAU failure.
Sorry, I cannot agree that we have a similar case in the sense that data
was 'erased' upon impact. From the quick study that I see, the FDR continued
to record failure codes as the DFDAU went offline. I will however look
futher into the scenario.
I would have to look at the data from AA 587 to know exacty what you
are referring to.
We are looking for cases of FDR's which have erased data after it has been
stored in CPM. That is my contest. Whether the data is corrupt, or not, is
not my concern. I'd be happy to see corrupt data on AA77's FDR after
'impact', however we have nothing. That is a concern.
Also equally concerning is the proximity of the aircraft to the poles vs. the
altitude which doesn't make sense.
That may be true for 'dumb sensors' such as a limit switch, or thermo-couple,
however DME and RAD ALT for instance have ARINC outputs which send
ARINC ready data to the system.
The time for data to hit the system is negligible. As supported by this
message I received from the manufacturer. Typical and Max. transmission
intervals 35 msec and 42 msec respectively:
TRACKING NUMBER: A00000844600-00001484871
Hello Sir,
Most of our documentation is proprietary information but I am
happy to provide you with the following information which gives
you the parameters that I believe you are looking for.
Have a wonderful day,
Best Regards,
Teresa Adams
Product Support Specialist
Melbourne Fl. comnet 731-7009
Local 321-768-7009
LRA-900:
The formatting of the data is pretty much industry standard (Arinc 429).
Minimum : 32 ms (Milliseconds)
Typical : 35 ms
Maximum: 42 ms
Jonnyclueless
Philosopher
- Joined
- Jun 18, 2007
- Messages
- 5,546
So does the manufacturer support your claims that flight 77 did not hit the Pentagon based on your findings?
My source is the NTSB. Quoting from the Final Report:
NTSB said:
The "white streak" is hypothesized to be a fuel effect, marking the instant one or both engines began to separate. Let me clarify my previous post -- analysis of the CVR, listening to the engine whine, suggests the engines were still running until about 4 seconds before impact, and the distance they travelled from the main wreckage is consistent with this. However, the "white streak" appeared about 6 seconds prior according to a geometric argument. Nonetheless, whether the FDR record stopped 4 seconds or 10 seconds before engine separation, the unambiguous fact is that the record stopped significantly before.
I never implied it did.
Functional, responding, yet behaving erratically all by itself. Again, there is no known combination of DFDAU or other upstream faults that results in garbage being written. And, again, AA 77's fault could also have been a DFDAU fault, since DFDAU and FDR were one and the same.
This is a Complex Cause Fallacy. Data "erasure" is but one of several possible explanations for the AA77 phenomenology. It also may have occurred here. We simply cannot tell.
But, again, why did some fail and others not? Why did the FDR record stop completely several seconds before engine separation, and 13.6 seconds before impact?
Again, Complex Cause Fallacy. Both erasure and corruption can potentially give you the same results. We don't know how the FDR's controller behaved after things went nonlinear. We don't know how much data was queued up somewhere. The problem is exacerbated by the paucity of sync markers and the compression algorithm -- the amount of corruption needed to turn several seconds of data from parseable to gobbledygook is quite low. There are numerous explanations, and you have yet to retire any of them. AA 587 also demonstrates, as you asked for, that FDR data stoppages prior to impact are not unique to AA 77.
Off topic, and incorrect besides.
"ARINC-ready" does not mean unfiltered, and does not mean they come pre-packetized. They do not.
There is no reason to believe this is true. Again, the "transmission" time you are looking at is the time between DFDAU and FDR, not sensor and FDR. It is the latter quantity that concerns us.
Last edited:
According to the study, and toll booth video, etc. the FDR stopped at 9:16:01 accompanied by a loud noise caught on the CVR. This was no
doubt the plane breaking apart supported by the white streak which was
picked up on video at :06
This is a case of inflight failure for certain, not a loss of data upon impact.
The study also says they were able to determine the missing seconds for
AA 587. Why then were they not able to figure this out for AA 77 "if" such
a thing existed?
I thought maybe you were comparing the status of the FDR and
CVR not realizing they do not receive data from the same sources.
Can you clarify this fault code on AA77?
Because of the in-flight damage caused by the rudder inputs from the Pilot.
The loud noise on the CVR followed by the white streak is a clear indication
of catasrophic failure before a fuselage impact.
AA77 did not experience any in-flight failure. There was no reason for the
data acquisition to stop at any time.
According to the NTSB data which is plain to see, there's no debating the
fact that the parameters mentioned show the plane too high and too close
to hit the light poles. Certainly not off topic.
It means the signal has been converted and put into data format needed
for the DFDAU to understand.
There is a reason to believe this because it came from the manufacturer
directly.
42 msec is the time for the sensor to generate a signal and send it to the
DFDAU.
The path and processing from DFDAU to FDR being an IDARS is negligible.
We are back to finding a logical reason for data erasure from CPM.
This is the part that nobody has yet to produce as reasonable to support
your theories.
Once again, if a data frame is laid down to CPM every second (256 words * 12 bits = 3072 bits), show how a transient can erase only a few seconds
worth of data AFTER it has been put to flash memory.
Last edited:
So the p4t decode is full of errors. I understand. The p4t decode shows 3.2 DME and 1.5 DME for the last second. Which do we choose.
You were spewing some garbage as 77 has two more seconds to fly; but you are wrong again. Sorry you have the wrong information, :46 is the last time stamp on the NTSB data, and :43 is the last second in the p4t decode. You need to look at your information before making more errors.
The 1.5 DME was stored but 77 is at 1.724 DME, using your accuracy number for DME! Busted, b u.
WRONG, if the DME senses 1.624 when 77 is really at 1.724 DME, 1.5 DME is stored and you have no clue! Busted!
I used your numbers! You missed the fact I used your number to show at :43 77 is 6 second from impact at :49 busting your false ideas BIG time BUSTED. Sorry, but which part of using your 0.1 NM do YOU not understand? You are fraud by telling lies and ignore the fact I used your numbers. Sorry to expose your fraud by trying to talk junk instead of seeing the facts.
I used your accuracy number and showed 77 is 1.724 DME from DCA, over 4600 feet from the Pentagon at :43! And this is backed up by RADAR!
If you understood accuracy, and resolution you would not be a p4t no evidence supporter.
The NTSB information refutes the stupid p4t idea of being too high, and witnesses prove you wrong. Busted.
But you did not take into account the RESOLUTION the FDR stores DME at, it is 0.25 NM resolution!
This means a reading based on your accuracy of 1.624 can be 1.724 actual NM from DCA.
Then the 1.624 is stored as 1.5 DME.
The p4t experts are not able to simple math! You have proven it again as your numbers show 77 can be 4600 feet from the Pentagon at :43 and that means 77 is only 370 feet above impact with 6 seconds to go! NOT TOO HIGH TO IMPACT THE PENTAGON. BUSTED again!
You do not understand your own accuracy or the resolution issue of DME storage in the FDR. This is why your ideas are flawed.
jaydeehess
Penultimate Amazing
That answers one question.
What is the size of an erase block?
What is the size of a write block?
IIRC one frame of data is 256 words representing one second of data transfer.
IF the data must get from sensor to a location in the flash memory in 0.5 seconds then the data must be written to memory in blocks sized no larger than half a frame. If the CPM waits until an entire frame , 256 words, is in the buffer before sending that data to flash the first data in that frame arrived at the CPM 1 second ago. By your interpretation of the specs this is 100% longer than it must take.
Once again then, how big is an erase block on this chip? (also is this flexible and if so then what is the size of an erase block wrt to the installation in the FDR in question?)
How much data is written to memory during a write command? (also is this flexible and if so then what is it set to in the FDR in question)
No. You need to read the whole study, not just scan it.
The aircraft was not breaking apart when the FDR record ends. The "loud noise" is hypothesized as failure of the vertical stabilizer, brought on by excessive rudder inputs. The NTSB also provides photographs of the attachment points, and verifies that the stabilizer broke away cleanly.
This in turn led to a loss of control, primarily pitch-roll caused by the lack of vertical stabilizer and coupling between the engines (at full power) and the aircraft's new center of pressure. That led to excessive aerodynamic stress, and that caused the engines to separate, leading to the "white streak." The aircraft was largely intact all the way in, and all of this happend after the FDR record stops. And the CVR doesn't stop at all.
The inflight failure, according to your understanding of the FDR system, should not have stopped the data record where it does. Nonetheless, that is what happened. You're wrong here, why can't you be wrong about AA 77?
The study also says they were able to determine the missing seconds for
AA 587. Why then were they not able to figure this out for AA 77 "if" such
a thing existed?
Because the CVR survived, with reference points that could be checked between it and the FDR. In AA 77, the CVR did not survive. Simple.
Not at all. However, both run off the emergency power bus. Yet one quit and the other didn't.
Non sequitur. My point is that the AA 77 anomaly might have been a failure of the DFDAU function. That function, in AA 77, was carried out by the FDR itself.
Catastrophic failure that did not disrupt emergency power, did not halt the CVR, and was confined to the vertical stabilizer and rudder assembly until after the FDR record ends. So verified by the FDR data itself.
This is not necessarily true. It is possible, albeit unproven and I doubt it myself, that maneuver or lighter impacts with obstacles could have upset AA 77's FDR. But there is similarly no easy answer for AA 587's FDR to stop where it did. Again, you fail to understand the event chronology.
Nope. If the FDR is indeed missing the last few seconds, the aircraft can hit all the light poles very easily. You know this, and it is off topic to the FDR discussion.
The DFDAU interprets and packages all signals the FDR records. The transmission time you asked for in the e-mail above refers to the time between DFDAU and FDR. I accept this time is < 0.5 seconds as reasonable. That is not the latency we need.
Totally wrong. The FDR or DFDAU manufacturer does not get to levy this spec on the sensor manufacturer, or any of the various avionics boxes in between. In the case of AA 587, the report discusses at length the problems caused by a filtering algorithm between sensor and DFDAU. Again, the spec you quoted was from "signal generation" and recording at the FDR. The only signals the FDR sees are "generated" by the DFDAU. Don't misquote your own e-mail.
I have already produced journal papers that provide several candidate mechanisms. And, again, this is but one possible explanation.
I'm not sure about the erase block, however the write block is a min. of
3072 bits.
I'll ask my sources about your question, or maybe MacGyver will know.
The data is time division multi-plexed. There are certain sensors needing
up to eight writes per second, some only once per second, and some even
less. The frame must be built whether or not these parameters require
polling, and "placeholder" bits are inserted to maintain frame integrity. This was covered back several pages back and you may find discussions
with Anti-Sophist and Undertow appealing for this topic.
This is something MacGyver might be able to answer. I'll ask my sources
as well. Back later with that info.
Another fine point, the write block is not a minimum 3072 bits. You've neglected the (variable) packing ratio of compression applied to FDR data. That figure only applies if compression is turned off, which is only expected if the FDR fails a replay check after writing data. It's in the AA 77 FDR report.
Also, we're describing the subframe size. We do not know if the subframes fit into a bigger write block mandated by the flash memory. My guess is that there is no such restriction, but it's an interesting possibility.
Also, we're describing the subframe size. We do not know if the subframes fit into a bigger write block mandated by the flash memory. My guess is that there is no such restriction, but it's an interesting possibility.
Last edited:
Likewise.
white streak at 09:16:06
5 seconds after FDR stopped
Not 09:16:08
FDR stopped at 3 seconds AFTER tail seperation with NEW noises of loud bangs and roaring noises due to engine speration. lateral accel pinned.
Page 133 of the PDF, NTSB states: "FDR and CVR data showed that engine
separation occurred during the out-of-control airplane motion that followed
the separation of the vertical stabilizer."
I wish I could agree with you, but the NTSB reconstruction and videos
tell otherwise:
http://www.ntsb.gov/Events/2001/AA587/board_mtg_anim.htm
You may want to retract that statement after re-reading the study and
watching the video reconstruction of the flight including text highlights of
the events. You may want to read the Flight Path Study, specifically Time
Correlation done by the NTSB
Please re-read the study and watch the flight study. It's all there.
No, it is NOT. It is the time the sensor creates the signal, converts it and
transmits to DFDAU in ARINC 429 format. IE: Sensor to input side of DFDAU.
Here is MY conversation with Rockwell in it's entirety:
We are debating those points with MacGyver now to credit, or discredit them
Last edited:
In response to Jay's question, the min. spec is 256, 12 bit words per second
into CPM.
That may not be the min. write block however...that is sometthing MacGyver
may know, or Undertow. I will get back with that info upon their replies.