AA77 FDR Data, Explained

[bje]

I don't think, but I know nobody here can.
Your theories are so funny, they don't even make sense.

Funny, the only "theories" spam2) being bandied about are from you, Turbobubba. You have yet to show AA77 flew over the Pentagon or debunk the massive evidence against you.

 

[Turbofan]

Translation:

We are the NTSB and we're too lazy to confirm everything. We'll let the
FBI deal with it...

isn't it strange that PFT decoded RAD ALT, but the NTSB coul...woul...
did not?

How's that bold action workin' for ya Funk?

 

[jaydeehess]

[qimg]http://procision-auto.com/911/current.gif[/qimg]

OK students, what is the voltage across R1, R2, R3 if VCC is 50 VDC?

Votlage accross aqll resistors is 50 V. This is of course because you are showing a steady state DC supply voltage and therefore there is no ac component to the supply and any ac impedances simply do not enter into the analysis. secondly in this cct the elements are all strictly resistive which means there is no electron tunneling or punchthrough taking place.

Question #2: We'll pretend these resistors are transistors. What is the
voltage across R1 and R3 if R2 is:

a. partially open
b. open
c. partially shorted
d. shorted

??? the question makes little sense. Do you mean to replace the resistors with transistors and have the ends of the transistors equate to the drain and source of the FET? If so then the gate voltage will control the voltage drop accross the D-S.

However, put a large high frequency spike on the drain or source and the ac characteristics of the FET will take precedence and given that the Drain , source and Gate impedances will vary slightly between each transistor the effect will be different in each one. Furthermore if these transistors are packed together within microns of each other there exists the distinct probability that electrons will flow between them not even on the etched current trace. There will be an inherent capacitance between each lead, between each lead and adjacent transistors and between adjacent transistors themselves.

Once again you demonstrate that you are completely in the dark when it comes to how transients affect a cct let alone how they affect the densely packed circuitry in a complex IC such as an EEPROM.

You do realise, do you not, that the cct diagram one sees on a data sheet is electrically but not physically the same as what is etched inside the IC right?

There will be places in the IC where a current trace just ends before it goes onto the next element. Wanna guess why? In other instances the cct trace will form a zig-zag for no apparent, physical, reason. Wanna guess why?

I asked you before;
do you recall the mnemonic "ELI ICE"?
Sometimes it is said as "Eli the Ice man".

How about telling me why those little snap on pieces of plastic are seen on power or signal cords so often nowadays? Why does a transmitter often have an iron ring with a split in it around the coax leading to the antenna?
HINT: It all has to do with attempting to manage transients.

 

[lapman]

Translation:

We are the NTSB and we're too lazy to confirm everything. We'll let the
FBI deal with it...

isn't it strange that PFT decoded RAD ALT, but the NTSB coul...woul...
did not?

How's that bold action workin' for ya Funk?

Translation: WAAAAA! WAAAAA! You won't play with us so we'll make up stuff while we throw a temper tantrum instead of addressing our numerous errors. WAAAAA!

 

[DavidJames]

Turbo - Why don't you go back to fixing cars and come back when you have evidence to refute the physical evidence observed and collected at the Pentagon left by AA77, including the body parts and DNA. Note: Questions are not evidence. If you want to claim the physical evidence is incorrect/wrong/faked/planted, you will need to provide evidence of same. In addition, if you believe AA77 did not hit the Pentagon, include, upon your return, evidence of what happened to the plane, passengers and crew.

Thanks.

 

[jaydeehess]

Replace that 50 VDC cell with a transient/AC/ or your sock for all I care.

What are the answers now?

The transient would equally go down each path IF
the leads were all the same length and had the same impedance along their length and the elements in the cct were all identical and if the capacitance between elements, between leads and elements, and between leads and other leads is high enough (basically that they are widely separated), none of which would be a problem with discrete elements on a large cct board. So what would that have to do with a densely packed IC?

 

[Turbofan]

The transient would equally go down each path IF
the leads were all the same length and had the same impedance along their length and the elements in the cct were all identical and if the capacitance between elements, between leads and elements, and between leads and other leads is high enough (basically that they are widely separated), none of which would be a problem with discrete elements on a large cct board. So what would that have to do with a densely packed IC?

Are you saying Ohm's law doesn't apply to IC internals?

 

[R.Mackey]

Are you saying Ohm's law doesn't apply to IC internals?

That isn't what he's saying at all. With practically every post, you further indict yourself as ignorant of electrical engineering.

Let me try to explain what you keep leaving out graphically:


See those stray capacitances and resistances? These are always there, and their values always vary between chips, lines, even individual cells are slightly different. They vary with temperature, with chemical concentration, sometimes even with pressure.

As a result, the real model for a transistor is as follows:



And even that is a gross oversimplification. This is only true while voltages and currents are still "reasonable," and without any physical damage.

This is why you often get purely local effects. Not all cells are created equal. Their precise values depend on the architecture and composition of the substrata. Your arguments are all based on the most simplified, everything-in-spec, steady-state simplification. The situation we're talking about cannot be treated from such a naive viewpoint.

Picture credits: Gray, Hurst, Lewis, and Meyer, Analysis and Design of Analog Integrated Circuits, Fourth Edition, Wiley, 2001. Pictures appear as Figures 1.19 and 1.20 on pp. 31 and 33 respectively. I apologize for the poor photographic quality, but I was in a hurry.

 

[funk de fino]

Translation:

We are the NTSB and we're too lazy to confirm everything. We'll let the
FBI deal with it...

isn't it strange that PFT decoded RAD ALT, but the NTSB coul...woul...
did not?

How's that bold action workin' for ya Funk?

Did UT decode all the data on the FDR? All the paremeters in the appendix from the report?

Why would any data be not recorded properly? Its a simple question and one you have run away from time and time again.

Did you read the Time Correlation report that the FDR reports mentions? If not why not?

Got a name from the pft liars list yet with my credentials? That is who you claimed were your experts.

 

[Myriad]

No Myriad, I don't work with such basic circuits. It is simplicity which helps rookies understand the big picture.

Right. And once those rookies learn the big picture, they're supposed to start learning the real-world details. Like the fact that real components do not behave in the ideal ways that are assumed, for simplicity's sake, in basic lessons.

You were right about 'd', and you should have mentioned a,b, and c would see VCC.

But that's not exactly true, except in idealized approximations.

In the real world, a 50VDC power supply sourcing 15 amps would slump a few percent below its rated voltage. (Unless there was some very good reason for the engineers to want VCC to remain at exactly 50VDC at all times, OR they wasted lots of money and excessive weight allowance for an unnecessarily robust power supply.) So, R1 through R3 would not actually be at 50 volts.

In the real world, interconnections between components (especially detachable connectors) have resistance. Not much, but when the load resistance is only 10 ohms, a little bit is significant and you'd definitely be able to measure the effects. Hence the potential at R2 would be slightly less than at R1, and R3 slightly less than R2. Again, that's unless there was some really good reason for engineers to spend extra money to reduce or eliminate that extra resistance. (That's why audiophiles -- even the sane ones -- find heavy-gauge speaker wires and gold plated connectors to be worth the extra cost. But for an electric heater, who would bother?)

In the real world, tolerances for resistors are up to +/-10%. So the current through the resistors would almost certainly not be exactly 5 amps nor exactly equal to one another as the diagram shows. (If you were to design additional circuitry that depended on the same amount of current flowing through each resistor in order to operate correctly, it would fail.) One resistor will sink a little more current, and run a little hotter, than the others.

And all this is still for simple DC operation. For rapidly changing voltages, things act even less like the ideal school text case.

So here's a quiz question for you: your circuit with the transistors substituted for the resistors is operating, each transistor sinking about 5 amps (some a little more than others, of course), heating the room up nicely, when suddenly someone sticks a screwdriver in the wrong place causing the output of a nearby 500VDC power supply to arc into the circuit. The arc hits the connection to the collector of T1 (where the diagram shows R1). Which of the following COULD happen:

1. None of the transistors is damaged.
2. All of the transistors are destroyed by internal arcing due to overvoltage.
3. Only T1 is destroyed.
4. T1 and T2 are destroyed but T3 remains functional.
5. Only T2 and T3 are destroyed while T1 remains functional.
6. T1 and T3 remain functional but T2 is destroyed.
7. Any of the above could happen.

Respectfully,
Myriad

 

[Jonnyclueless]

Translation:

We are the NTSB and we're too lazy to confirm everything. We'll let the
FBI deal with it...

isn't it strange that PFT decoded RAD ALT, but the NTSB coul...woul...
did not?

How's that bold action workin' for ya Funk?

Lazy? Talk about a cop out. Care to back that up?

The experts who do this for a living, had reason to believe that the data could not be confirmed as reliable. They didn't just say "Oh we don't want to be bothered". They said they could not confirm it was reliable. That's not being lazy, that's being a professional and not a bunch of hacks who don't understand how FDRs work. I could decode a corrupt file, but that doesn't mean the data is correct. Why would they use data that could be incorrect? That's what you call being an expert? I guess by PFT standards it would be.

The fact is they had a crashed plane at the pentagon and 1000s of witnesses, and enough evidence to fill a dictionary to confirm this. They're professionals, not crackpots. If the NTSB started making the claims that PFT is making, then everyone there would be fired for incompetence.

Why do you think the issue is about being able to decode the data? And you wonder why you aren't taken seriously?

 

[Turbofan]

Right. And once those rookies learn the big picture, they're supposed to start learning the real-world details. Like the fact that real components do not behave in the ideal ways that are assumed, for simplicity's sake, in basic lessons.

Just like the rookie engineers that designed an FDR that can't handle
impact!

Remember, these units are certified to 3400 g's and could likely withstand
more.

What's your guess as to the impact force for AA77 if it were to hit the
Pentagon at 500 MPH?

In the real world, tolerances for resistors are up to +/-10%.

Depends on circuit class and grade of course...

And all this is still for simple DC operation. For rapidly changing voltages, things act even less like the ideal school text case.

Sure.

So here's a quiz question for you: your circuit with the transistors substituted for the resistors is operating, each transistor sinking about 5 amps (some a little more than others, of course), heating the room up nicely, when suddenly someone sticks a screwdriver in the wrong place causing the output of a nearby 500VDC power supply to arc into the circuit. The arc hits the connection to the collector of T1 (where the diagram shows R1). Which of the following COULD happen:

1. None of the transistors is damaged.
2. All of the transistors are destroyed by internal arcing due to overvoltage.
3. Only T1 is destroyed.
4. T1 and T2 are destroyed but T3 remains functional.
5. Only T2 and T3 are destroyed while T1 remains functional.
6. T1 and T3 remain functional but T2 is destroyed.
7. Any of the above could happen.

Respectfully,
Myriad

A lot to assume here, including the arc and where the screwdriver is inserted.
We're also assuming (not giving a parallel to the FDR) that our circuit has no
filters or RV protection, etc.

My answer would be 7 based on the minimal info given.

 

[beachnut]

Remember, these units are certified to 3400 g's and could likely withstand
more.

So are you sure 77 FDR was not under the 1700 G spec?

 

[Jonnyclueless]

Turbo, there's more to the crash than just the impact as well. There's everything that happens after the crash. Some of those things cause 110 story buildings to fall. It's not like plane impacts building, then no damage can happen after impact. Yes, they design them for reasonable crash scenarios, but they aren't indestructible as has been shown by many black boxes not being recoverable. Including one of the black boxes on flight 77.

The CVR is a perfect example. The data was not recoverable. Same accident. It sustained the same G forces. So if the FDR was rated for a million Gs, you still have an example o a black box on that plane designed for plane impacts that ended up with unreadable data.

 

[jaydeehess]

Are you saying Ohm's law doesn't apply to IC internals?

See Mackey's post? It outlines the full ac treatment for a transistor. A very similar illustration appears in my old electronic textbook. Does your's not? Those stray inductances and capacitances only have to be taken into account in cct design when the cct will be handling high freq signals. They can be ignored in normal operation if such high freq's are not part of the situation. If however one requires razor sharp square waves then some consideration for those impedances will have to be taken into account. However, a spike or sudden drop in voltage contains even more high freq components than a square wave. (Did you do Fourier analysis in your course?).

Why do you not use anything you learned in school past the first term?

Transients are NOT DC!

Transients that the input filtering were designed to handle are those which may reasonably thought to occur during normal operation. There is no cost effective way to gaurd against the ill effects of transients that result upon crashing. I suppose that a severe 'g' detector ( something greater than the airframe would handle anyway, say 25-30 g) could be incorporated that would switch the FDR to an internal power supply that would also be specifically designed to not fluctuate during these high 'g' encounters and that the FDR would , during this time also be separated from the data i/p bus. Thus completely isolated it would then automatically go through a power down procedure.


Of course the FDR would cost $100,000, and likely weigh 3 times what one weighs now and take up 4-5 times the volume.

 

[Turbofan]

You know what, rather than repeat myself and waste my time, I'll take these
comments to the EE forum where there are actual FDR designers.

We'll see what filtering is used and if they happened to consider the effects
of high impacts when they built these dinky little things.

 

[JimBenArm]

You know what, rather than repeat myself and waste my time, I'll take these
comments to the EE forum where there are actual FDR designers.

We'll see what filtering is used and if they happened to consider the effects
of high impacts when they built these dinky little things.

Will you listen any better to what they say?

 

[WildCat]

Will you listen any better to what they say?

Maybe Undertow will register to independently confirm the PfffT findings...

 

[jaydeehess]

You know what, rather than repeat myself and waste my time, I'll take these
comments to the EE forum where there are actual FDR designers.

We'll see what filtering is used and if they happened to consider the effects
of high impacts when they built these dinky little things.

So, when they use subject matter that goes beyond first term college electronics you will be able to follow them or will you rely soley of Ohm's Law?

Your out of context cherry picking of the comments thus far would not seem to indicate that you do actually follow the conversation.

Now that they know that it is an FDR and that up to 6 seconds of the last data recorded was corrupted do you believe that they will be converts to the PfT, "something nefarious is amiss", line of paranoid contentions?

If they do not agree with the PfT senario then will they simply join the ranks of anonymous internet non-experts?

Care to link to the rest of "all" the electronic forums, or are we going to be doing this one at a time?

 

[Myriad]

A lot to assume here, including the arc and where the screwdriver is inserted.
We're also assuming (not giving a parallel to the FDR) that our circuit has no
filters or RV protection, etc.

My answer would be 7 based on the minimal info given.

Well, I did specify that the arc hits the circuit at the collector connection of T1. But in any case, that is the correct answer.

Any, all, or none of the transistors could be damaged. If not all are damaged, then the one most likely to be damaged is T1 followed by T2, but there's no guarantee. The surge could take out just T3. When one transistor does fail, it provides a path of less resistance for the remaining transient charge, protecting the other transistors to some degree. So it's common to have some components fail, and others that are in parallel off the same bus remain operational.

That's how failures work. Same as when lightning striking a power line trips some breakers but not others. Same as when "some idiot" who should have been locked up as a danger to himself and others foolishly touches a live 250 amp common cable (could have been worse, at least it wasn't the hot) to the Ground socket of a stage light dimmer panel, and six out of 21 SCR dimmers (all in parallel) are toasted leaving the other 16 operational. Components that are identical in a schematic diagram are not identical in reality, and they don't necessarily all fail or all survive just because one does. It works that way at the micro scale too.

Respectfully,
Myriad