AA77 FDR Data, Explained

[wstutt]

... A lot can be determined by the actual voltage transitions (waveform) which can be reconstructed with the uncompressed data ...

Does this mean that you and the FDR specialist would be looking for irregularities where the waveform differs from the ideal theorerectical waveform that represents the uncompressed data? i.e. the actual electrical waveform that was fed to the FDR rather than the digital data decoded from it? Was that why you were asking the NTSB for the input to the FDR in your FOIA request?

Warren.

 

[BCR]

Does this mean that you and the FDR specialist would be looking for irregularities where the waveform differs from the ideal theorerectical waveform that represents the uncompressed data? i.e. the actual electrical waveform that was fed to the FDR rather than the digital data decoded from it? Was that why you were asking the NTSB for the input to the FDR in your FOIA request?

Warren.

Yep, something like that. The NTSB has told me it is possible to obtain the waveform from the raw file, but they of course refuse to provide it in that format. I have already worked with the serial bit stream in the waveform format from a previous crash incident, so I know the value of it in diagnosing such power loss issues. The problem is getting from one to the other.

I don't know if it can be done, but it is a nice technical problem to challenge my meager brain cells with.

 

[wstutt]

Yep, something like that. The NTSB has told me it is possible to obtain the waveform from the raw file, but they of course refuse to provide it in that format. I have already worked with the serial bit stream in the waveform format from a previous crash incident, so I know the value of it in diagnosing such power loss issues. The problem is getting from one to the other.

I don't know if it can be done, but it is a nice technical problem to challenge my meager brain cells with.

If the NTSB did produce a waveform from the raw file, wouldn't it just be an ideal theoretical waveform stripped of any irregularities that would indicate power issues?

Warren.

 

[BCR]

No. It would be the voltage transitions which represent the actual serial bit stream. Or at least that is my understanding from the NTSB. I will have to defer that question to GS, since he is the one who is requesting it. I'm in no way, shape or form an FDR specialist. I'm just a data guy.

 

[apathoid]

Careful guys, he's watching! Hi Bob!! Any luck raising the 1000$ yet? Why don't you get off your lazy butt and get a job already. Have you no shame?

lol... i saw they tried to equate door prox sensors to loss of power. Hey apathoid, are the gear and gear doors on the PSEU? Whooops! There goes that theory...

If only I could use google to troubleshoot complex aircraft systems. Ever hear of a schematic, or a wiring diagram manual Johnny? Maybe I already researched and found that the sensor excitation came from another power source, or through a relay which is only energized from from the R AC bus, or from the WEU. Why do you pretend to know what you are talking about? Maybe you could fool someone into hiring you as an avnx tech!

apacrap should have picked up on such blatant BS "hypothesizing" if he was worth his salt as an avionics tech. Especially noting the airspeed and G's needed to pull out of such dives (if anythnig, the open door conditions are more consistent with the flyover). Its not uncommon for Prox sensors to show open condition on doors in high speed flight. Especially on older aircraft. We had it happen several times on the 328jet and that thing was brand new!. And guess what... no FDR FAIL! Apathoid sure isnt the cracker-jack avionics tech they seem to think he is. Its clear why he also stays anonymous and refuses debate.

Yes, great point: a DO-328 Jet and a 757 have nearly idenitcal systems. I bow to thee superior aviation know-it-all.
Yes, great point: it always has to be a primary bus failure, and not a poorly rigged sensor/broken wire to cause door warning. I bow to thee superior aviation know-it-all.

(3) It receives its electrical power from the bus that provides the
maximum reliability for operation of the flight recorder without
jeopardizing service to essential or emergency loads;

Thanks for reiterating what I already said, genius.

Looking at a Boeing 757 Flight Manual, I guess some haven't noticed the dual B757 AC Transfer Busses, Hydraulic Driven Generator, redundant Transformer-Rectifier Units (TRU's), DC Bus Tie Breakers, APU BATT VOLTS, dual power sources listed for L-3 FA-2100 SSFDR, etc.:

I would quit while you're ahead...oops too late. You have absoluetely no idea what you have just written, have you?

Here's a free lesson. Actually, its something that would be learned in the first hour of the first week of Avionics 101: Power sources cannot be run in parallel. What does that mean? It means that a unit powered from the L AC bus cannot be powered by the R AC bus at the same time. Only when on a single source(APU, External, HMG) is when the busses can be "tied". Now here's another freebie: dual AC transfer busses mean L and R AC bus, as I've been saying. Here's yet another: TRU's need AC to run, and they provide power to units that need DC, genius. What does the FDAU run off? Here's another: A bus tie breaker is a relay, not a power source, oops. How about another: APU's arent normally run in flight, and they take about a minute to start, oops. As for the HMG, it starts automatically after both AC busses fail. Ooops again! Dont quit your day job genius! Any more brilliant insight?

By the same token, how can it erase anything if suddenly powered down? 4-6 seconds of "lost data" doesn't fit the system design, nor the data rate. The buffer does not hold 4-6 seconds worth of data. The system must write data within 500 mseconds of being clocked in. Electron tunneling cannot occur without the proper bias on the gate(s).

Erase data that hasnt been recorded because it's turned off? Are you daft? Or are you now saying the FDR can see into the future after a power loss? That's really interesting. Does it predict the stock market too? Where can I get one? And thanks for the electron tunneling, it just answered my next question...

eta: I see he said APU BATT VOLTS which is apparently avionics illiterate-speak for the APU battery, whose sole purpose is to start the APU. I thought he said APU Generator which is, in fact, a power source for the L AC bus.

Here is the link to the PfT analysis idiocy.

 

[beachnut]

The math challenged, failed physics p4t nut case idea brigade is taking up aircraft systems? Will stupid ideas flow freely from the cast of cult members spewing moronic poppycock; will they match or top delusional statements by Balsamo? When will Balsamo's next death threat come? Has Balsamo's enemies list grown? What happen to the big smoking gun FDR media explosion to expose p4t inside job no theory 11.2g super maneuver? Who cares?

Did the p4t delusion squad say the doors could not open? 77 is past Vmo; I have seen doors, access panels and aircraft skin opened, ripped off and delaminated by high Q.
Warren posts facts and his research. What will the cherry-picking quote-mining terrorist loyalist distort with his work? Will they claim Warren is debunking the world of logic and reason to save Balsamo's nut case ideas from begin vaporized by reality?

... since their theories/speculation of corrupt data have been thoroughly debunked by Warren ...

Balsamo is getting excited. No matter what Warren posts Balsamo can make it into a debunking of the evil JREF and those on his enemies list. What did Warren debunk? The data is not in the FDR so he can't debunk anything on the missing data or the fact data is missing.

Did Balsamo post the 0.5 second power up junk again? How does he explain the missing/lost/not recorded data on 77 and other FDRs. If not for the NTSB animation poor Balsamo would have nothing to do. Why have Lear's aliens not stepped up to help Balsamo with math and physics? Are all the pilots in pilots for truth failures in math and physics? 11.2g still posted? Yes.

I have not seen Warren tie his work to the failed delusional conclusions of p4t or CIT. They offer no theory to begin with.

It does not matter if Warren finds all the data is not corrupt for all 25 plus hours from the FDR.

I have written a program to check the Hamming codes and page parities in the AAL77 FDR file. There are only two pages where the Hamming code and page parity shows as being incorrect. It appears to me that this is because these pages were not completely recorded. You can read about it at my web site at warrenstutt dot com. Follow the AAL77 Hamming code and page parity checker link.

I see no evidence that any of the data was corrupted after being recorded.

He did find two pages not correct; corrupt? Does this effort help Balsamo sell DVDs which imply 77 did not impact the Pentagon?

Is Warren allowed to say the idea of 77 not impacting the Pentagon is an insane conclusion at p4tf?

Were all the flights in the FDR checked and only only two pages were in error?

Did Balsamo say Warren would be banned for posting his work at JREF? How paranoid; Balsamo is not good at reality. Does Balsamo know Warren's work does not support Balsamo's implied lies?

 

[wstutt]

<snip>

Were all the flights in the FDR checked and only only two pages were in error?

<snip>

Yes, the program checks all the flight data in the FDR including all the other flights stored in the FDR. The Hamming code and page parity matched for every page except for two.

Warren.

 

[beachnut]

Yes, the program checks all the flight data in the FDR including all the other flights stored in the FDR. The Hamming code and page parity matched for every page except for two.

Warren.

Thank you; sorry if I missed this information in your previous posts; thanks for saving me time.

Your posts are professional. If Balsamo told you to expect ad Hom and personal attacks you will be disappointed if you keep posting fact filled professional posts. If you find it insulting to have your posts labeled professional by an over the hill pilot and failed engineer, you will need to lower your standards by several orders of magnitude.

I bet Balsamo expects you to be banned for posting facts. His paranoid delusions have no bounds and he most likely confuses skepticism with speculation.

Facts and evidence are needed to shine light on the ignorance Balsamo spreads. Your efforts will help people understand the FDR, but will not change the terrorist flying Flight 77 into the Pentagon. Your outstanding professional posts are something to strive for IMO. Thanks for setting the standard high.

it will be my shout next time I get to the gold coast; one can dream

 

[jaydeehess]

Its not uncommon for Prox sensors to show open condition on doors in high speed flight. Especially on older aircraft. We had it happen several times on the 328jet and that thing was brand new!. And guess what... no FDR FAIL!

I would think that if such was a common occurance it would be considered a failure of the sensor system and require a redesign. One of the worst times for a door to open would be in high speed flight and an alarm condition showing a door open would be quite disconcerting to the flight crew. Later, if this occured on a regular basis, the crew would 'learn' to ignore the door open warning (the 'cry wolf' parable) which could be to the detrement of flight safety. In other words a sensor that is unreliable is worse than useless. Seems that the Doh'-jet had a few flaws.

As for TF's post, it does seem like he is saying that the right and left AC busses back each other up in an uninterruptible fashion. That would be quite unusual and, as Apathoid points out, is not the case here.

As a point in case, a town near by was having a lot of trouble with power outages and so a lot of residents bought generators so that they could keep their refrigerators running if the power went out. Most would fire up the genset and run an extension cord to the frig when needed but every once in a while someone would wire the genset with an extension cord that had two males ends on it and when power went out they would plug the set into a wall socket to supply power backwards through the breaker panel to the whole house. Of course this also sends power to the outside lines and can electrocute power co. workers. However usually when city power was restored the homeowner would call the power co. and complain. If they were lucky the main breaker would snap off before anything went up in smoke, and usually the genset would stall or have its own breaker snap off. A few unlucky stupid people had their genset bounce accross the lawn and catch fire, or worse, the house caught fire.

I wonder if TF can figure out why?

Furthermore TF seems to still not get it as far as what we are postulating here.
Hi TF, let me spell it out.
When power goes down on the buss supplying the DFAU and/or the DFDR then there will be no data being recorded, and if this occurs during flight due to a compressor stall on an engine then no data will be recorded during the time that buss is unpowered. If the DFAU and DFDR are then supplied with power from another source they will need time to reboot and no data will be recorded during that time either.
This has been shown to have occured in other flights and is bolstered by the fact that several systems suggest that power was going off on one buss just prior to the missing data time period.

However if one wants dearly to believe in a vast, complicated, complex , unneccessary, and wholly unreliable plot involving fake FDR's and decoy aircraft then examining the available data is hardly going to sway them.

 

[BCR]

However if one wants dearly to believe in a vast, complicated, complex , unneccessary, and wholly unreliable plot involving fake FDR's and decoy aircraft then examining the available data is hardly going to sway them.

Yes sir, I learned that the hard way.....

 

[apathoid]

I would think that if such was a common occurance it would be considered a failure of the sensor system and require a redesign. One of the worst times for a door to open would be in high speed flight and an alarm condition showing a door open would be quite disconcerting to the flight crew. Later, if this occured on a regular basis, the crew would 'learn' to ignore the door open warning (the 'cry wolf' parable) which could be to the detrement of flight safety. In other words a sensor that is unreliable is worse than useless. Seems that the Doh'-jet had a few flaws.

In my experience, I've never had much trouble with door sensors. I think I've replaced just one in 9 years. I used to rig them(adjust gaps to the targets) all the time when I worked in heavy maintenance, never really had a problem there either. Now the landing gear sensors actually crap out frequently, as they are out in the elements and the sensor targets quite often build up grease which could give a false gear unsafe light.

But, yeah, Johnny has a gift for hyperbole. I'm sure the DO Jet is just fine.

As for TF's post, it does seem like he is saying that the right and left AC busses back each other up in an uninterruptible fashion. That would be quite unusual and, as Apathoid points out, is not the case here.

Well newer model 767s(-400) and 777's have seamless power source switching. But AA77 wouldn't have had that. So, as you pointed out, the FDR would reboot on a power source switch, which would include loss of the primary power source(#1 IDG) causing the bus tie relay to energize, powering the L and R AC busses from the #2 IDG, which normally powers only the R AC bus. The power source switch take about 1-2 seconds on a 757, but that's about 1000-2000ms too much for the FDR .

When power goes down on the buss supplying the DFAU and/or the DFDR then there will be no data being recorded, and if this occurs during flight due to a compressor stall on an engine then no data will be recorded during the time that buss is unpowered. If the DFAU and DFDR are then supplied with power from another source they will need time to reboot and no data will be recorded during that time either.
This has been shown to have occured in other flights and is bolstered by the fact that several systems suggest that power was going off on one buss just prior to the missing data time period.

pft
But, but, but....the 757 has APU BATT VOLTS and redundant Transformer-Rectifiers duhhhhh! You sheeple!
/pft

However if one wants dearly to believe in a vast, complicated, complex , unneccessary, and wholly unreliable plot involving fake FDR's and decoy aircraft then examining the available data is hardly going to sway them.

Anyone still believing NPT at the Pentagon just wants to believe it(or in the case of PfT, needs to believe it). Honestly, what does it say about the Penta-dorks when even their own "Truth Movement" calls them delusional nutcases?

 

[jaydeehess]

In my experience, I've never had much trouble with door sensors. I think I've replaced just one in 9 years. I used to rig them(adjust gaps to the targets) all the time when I worked in heavy maintenance, never really had a problem there either. Now the landing gear sensors actually crap out frequently, as they are out in the elements and the sensor targets quite often build up grease which could give a false gear unsafe light.

I can accept that sensors in a harsh enviroment can fail more often than protected ones. Cappy Robby was however not even saying it was normal wear and tear. He was saying it was a common flight condition that would cause this error in the sensor.

But, yeah, Johnny has a gift for hyperbole. I'm sure the DO Jet is just fine.

I was not saying I would not fly on one, unless RB is in the cockpit,,,, scratch that,,, if he's on the plane, I won't be.

Well newer model 767s(-400) and 777's have seamless power source switching.

That would require frequency and amplitude matching, though I would believe that frequency would be the major concern. Or AC-DC-AC conversion with both AC sources available to the converter and the ability of the converter to absorb a slight delay in switching AC source (large capacitor on DC o/p would be a simple way to allow for that)

pft
But, but, but....the 757 has APU BATT VOLTS and redundant Transformer-Rectifiers duhhhhh! You sheeple!
/pft

This guy sells batteries?

 

[wstutt]

Thank you; sorry if I missed this information in your previous posts; thanks for saving me time.

Your posts are professional. If Balsamo told you to expect ad Hom and personal attacks you will be disappointed if you keep posting fact filled professional posts. If you find it insulting to have your posts labeled professional by an over the hill pilot and failed engineer, you will need to lower your standards by several orders of magnitude.

I bet Balsamo expects you to be banned for posting facts. His paranoid delusions have no bounds and he most likely confuses skepticism with speculation.

Facts and evidence are needed to shine light on the ignorance Balsamo spreads. Your efforts will help people understand the FDR, but will not change the terrorist flying Flight 77 into the Pentagon. Your outstanding professional posts are something to strive for IMO. Thanks for setting the standard high.

it will be my shout next time I get to the gold coast; one can dream

Thanks beachnut for your kind words about me.

Warren.

 

[George 152]

Thank you; sorry if I missed this information in your previous posts; thanks for saving me time.

Your posts are professional. If Balsamo told you to expect ad Hom and personal attacks you will be disappointed if you keep posting fact filled professional posts. If you find it insulting to have your posts labeled professional by an over the hill pilot and failed engineer, you will need to lower your standards by several orders of magnitude.

I bet Balsamo expects you to be banned for posting facts. His paranoid delusions have no bounds and he most likely confuses skepticism with speculation.

Facts and evidence are needed to shine light on the ignorance Balsamo spreads. Your efforts will help people understand the FDR, but will not change the terrorist flying Flight 77 into the Pentagon. Your outstanding professional posts are something to strive for IMO. Thanks for setting the standard high.

it will be my shout next time I get to the gold coast; one can dream

That'll mean you only have another 1200 miles for another beer or three

 

[BCR]

Oh by the way. I'm kinda stumped on some of wstutt's C# code. A bitwise operator, ^= has no complete translation in VB.NET.

iHammingCode ^= iBitHammingCode (C#) does not produce the same results as iHammingCode = iHammingCode Xor iBitHammingCode (VB.NET).

So, anyone who might know a work-around, let me know. I'll figure one out eventually, but I've spent several days in debugging and my brain hurts.

 

[wstutt]

Oh by the way. I'm kinda stumped on some of wstutt's C# code. A bitwise operator, ^= has no complete translation in VB.NET.

iHammingCode ^= iBitHammingCode (C#) does not produce the same results as iHammingCode = iHammingCode Xor iBitHammingCode (VB.NET).

So, anyone who might know a work-around, let me know. I'll figure one out eventually, but I've spent several days in debugging and my brain hurts.

I don't see why that shouldn't work. I've got the Visual Studio 2008 Express Edition of C# and VB.NET installed on my computer now, so either you could send me your code and I can look through it or I could port my code to VB.NET and put it on my web site for the benefit of yourself and others.

Warren.

 

[BCR]

I don't see why that shouldn't work. I've got the Visual Studio 2008 Express Edition of C# and VB.NET installed on my computer now, so either you could send me your code and I can look through it or I could port my code to VB.NET and put it on my web site for the benefit of yourself and others.

Warren.

Now that would just take all of the fun out of it. You are correct, the VB.NET translation should work. I doesn't (the integer values are off by just a few for iHammingCode variable. All of the other variables generate the exact same result for any give point in the loop. It is the finding out why that is the exciting part, not the solution

 

[BCR]

Actually, the problem seems to be in the prior code segment.

bitHammingCode = (bitNum1 == (1 << shiftNum)) ? 1023 - shiftNum++ : bitNum1;

The VB 'conversion' is ...

bitHammingCode = If((bitNum1 = (1 << shiftNum)), 1023 - System.Math.Max(System.Threading.Interlocked.Increment(shiftNum),shiftNum - 1), bitNum1)

The first iteration for the C# code is 1023, while the VB.NET translation returns 1022 (all other variables return the same values).

Update:

Oh my. Now that is why I like to take a couple days and then come back to look at something again. The solution turned out to be a simple one. Instead of using the one line of code equivilent, I broke it into two lines of code.

iBitHammingCode = If((iBitNum1 = (1 << iShiftNum)), 1023 - iShiftNum, iBitNum1)
If (iBitNum1 = (1 << iShiftNum)) Then iShiftNum += 1

Works fine now

Ain't science fun!

 

[wstutt]

<snip>

Works fine now

Ain't science fun!

I'm glad you got it working

Warren.

 

[apathoid]

Only tangentally related to the thread, but thought it would be an interesting contribution:

Despite PfffT's claims to the contrary, it appears that FDRs can't actually work without power. Who would have imagined?! Boeing will feature an FDR with an independent power source for the 787 program. So it appears that our concerns are being addressed.

The EAFR system features network-connectivity that enables reconfigurable flight data acquisition. It is also the first flight recorder to have a recorder independent power supply which allows operation in the event of a power failure, GE said.