Sukhoi Su-24 flight data recorder teardown video

[free_electron]

Er, you did see the massive dent in the corner of this thing? There comes a certain point where you have to say that just about anything a human could inflict on this thing (hey, even with a crowbar) isn't going to change the final outcome...

yes i did see it. all the more reason to be very careful not to inflict more damage.
it looked very amateurish ... .

properly done  one would dissolve that white stuff, then open the insides properly. no drilling no hammering , no prying with screwdrivers.

did you see how he cut those boards apart ? that cutter is way too large. should have done with excelta's or lindstroms, not an electricians cutter made for 4mm square wire ...

[Gyro]

The fun part that made me laugh was at 3.42. Someone silently said in Russian (rough translation, IMO impossible to accurately translate): "those goats (morons) should be exposed to educational program".   

"These morons should go back to scohol"

To be more precise, he said:

Damn, should I [or we] have to learn them [how to do that]? These goats?

It is not clear who did he call goats: technicians that were trying to unscrew bolts (which is more likely), foreign observers in the lab room or reporters in the press room.

Also, some ppl think he said "kak vzlamyvat" (how to crack this open) rather than "kozlam etim" (these goats)

I LOVE this Forum, where else could you get such international insights 

[rch]

Did anyone else think the board may have been bent in assembly when the external cable was pulled too tight?

[SeanB]

Reason for flat head screws is because all USSR made stuff uses a special design of USSR made flat head screw, and it has a thinner slot than the Western flat head screw. That way they could tell instantly if a product was local make or a import, as the import would use a Phillips, Posi or JIS head screw. The flight recorder was probably designed to withstand a fire in the aircraft, not a fall out of the sky after being hit with a missile, which typically will be a heat seeker aiming right at it. I know of one pilot who landed the plane after the missile removed a large part of the rear.

The chips themselves can be read, even if cracked, though the cost is going to be higher and you will really need the original designs to see where to probe the individual cells, using a head similar to the probe fixture Dave got from Vincent as a teardown. Complete chips are easier, but even partial chips can be read, though the data will need to be decoded as to where it actually was on the chip itself and might not be a linear write through the cells. Does rathger look like standard Western made flash with a USSR made controller chip on the boards.

[mikeselectricstuff]

Surely the way to mount PCBs in something like that is NOT to use rigid pillars, which transmit shock to the PCBs but to have them floating in a soft filler material with plenty of space around, so any impact just moves them around instead of bending the PCB.

Was disappointed to not see a shot of the inside of the inner enclosure to show that the PCB damage was plausible, and whether there were any chip pieces in there, and if not, was there a big enough hole for them to have fallen out - from what we saw it could just as easily have been whacked with a hammer.   

[TerraHertz]

I do NOT envy the guy doing the disassembly, with so many people watching. Especially when it became difficult due to the screws seized from impact deformation.

Hands up who else can use tools skillfully enough, unless someone is watching - in which case you become an apparently incompetent stumblebum?

Also to have to resort to hammering and drilling, in that situation... awkward.
Heh, and they really didn't expect that. At 19:09, that looks like a concrete nail he's using to centerpunch the head of the screw, to drill it out. Well it works...

So amusing to see all the members of the press, bored to distraction by having to wait a few minutes while some guy drills out a few screws.

Weird they didn't have a small vacuum cleaner, to suck the insulation dust out of the recorder as they dug down inside it. You'd think they'd expect that. Since this is supposedly a room used for many international flight recorder accident investigations. Had to bring a standard vacuum cleaner in from outside.

Hmm... All three nested cases are cast alloy, and quite thick. With compressible insulation between them. That would probably be good enough for impact damping from an aircraft hitting the ground. But that hole in the outer case was made by a high velocity shrapnel from the missile, and it looks like it penetrated to and damaged the second casing as well. Hence the seized screws in that one's lid. I guess that explains the extremely high G forces that smashed the boards and chips. The direction of breakage is consistent with a high G impulse due to the case penetration impact. On the underside of the bottom PCB, the flash chips pulled away from the PCB by their own mass... I wonder what the actual G force is, required to do that?

It's a pity. The Russians know their plane never entered Turkish airspace, and the shootdown was a prepared ambush with no justification. But it would have been useful to be able to prove it via indisputable public process.

At least they still scored points for opening a black box in public, as opposed to the total secrecy of all such Western investigations in recent decades.

[firehacker]

Was disappointed to not see a shot of the inside of the inner enclosure to show that the PCB damage was plausible, and whether there were any chip pieces in there, and if not, was there a big enough hole for them to have fallen out

[mikeselectricstuff]

Was disappointed to not see a shot of the inside of the inner enclosure to show that the PCB damage was plausible, and whether there were any chip pieces in there, and if not, was there a big enough hole for them to have fallen out

Not clear enough - shows no signs of incursion, in fact surprisingly little damage on the visible part.

[mikeselectricstuff]

Do they not have bandsaws in Russia? That would have been be a low-trauma way to remove the casings.

[firewalker]

Maybe the case was impacted by the projectile that shot the plane down. That would have cause significantly more g forces than the crass.  Are those recorders designed to survive 20mm ammunition?

Alexander.

[firehacker]

Do they not have bandsaws in Russia? That would have been be a low-trauma way to remove the casings.

I bet there is no bandsaw in this lab. But let me translate what two dudes are talking at ninth minute:

9:02 — [angry guy] So, they began using hammers...
9:05 — Would be nice to have a saw.
9:07 — [angry guy] Huh?
9:07 — Would be nice to have a saw here.
9:09 — [angry guy] Probably he will be using an angular grinder soon.
9:11 — Really?
9:11 — [angry guy] They've got one prepared in the another room.

[daqq]

So lessons learned. Use multiple QFN/DFN packaged flash chips instead of a large TSSOP chip, to manage stress cause by impact, as well as increase redundancy.

I don't know if anything that's off-the-shelf or is not specially built for this can survive stuff that involves this large acceleration. Look at military or aerospace PCBs - very different construction techniques, IC packages...

[zoltan]

Just watched the video (up to 40') and I'd like to ask.

There are "services" for reverse engineering processors, flash, etc for reasonable low prices commercially available. Is reading a broken flash chip so much difficult? Considering that you are the manufacturer of the product, and know how it works, how hard it can be? Cost is not an issue.

[SeanB]

Just watched the video (up to 40') and I'd like to ask.

There are "services" for reverse engineering processors, flash, etc for reasonable low prices commercially available. Is reading a broken flash chip so much difficult? Considering that you are the manufacturer of the product, and know how it works, how hard it can be? Cost is not an issue.

Only chip made in USSR is the cerdip controller, the rest is generic Hynix or some other brand Flash, just like is used in your SSD drives. Easy to read if the chip is still intact, and even if it is cracked you can still read the data off the broken bits if really needed, after a decap and clean.

[zoltan]

Only chip made in USSR is the cerdip controller, the rest is generic Hynix or some other brand Flash, just like is used in your SSD drives. Easy to read if the chip is still intact, and even if it is cracked you can still read the data off the broken bits if really needed, after a decap and clean.

That's the point. The controller is SSSR made, so they know the storage algorithm. The flash chips are (probably) generic, and even if broken, or shattered (that would be pain in a**) I believe the data can be restored. Could the flash chip be read by optical means? If you have part of the chip and take a hi-res photo of it under different wavelengths? Of course after decapping and cleaning.

[daqq]

Could the flash chip be read by optical means?

To the best of my knowledge no. The data is stored as a charge that is either present or not - not producing any optical effects.

The controller is SSSR made,

Russian, not SSSR, USSR etc.

[wraper]

http://www.vesti.ru/doc.html?id=2700775#
Only 1 out of 16 ICs survived. They also told that will see if they can read the data out of the dies.
The hole penetrates up to the internal container.

[Hydrawerk]

I would expect a different construction, this looks like a design fail.

[firewalker]

The extend of the penetration (The noticed it and said F.U.B.A.R.), could "justify" the crudity of the operation.

Alexander.

[reagle]

A nice writeup by somebody familiar with Soviet technologies (sorry, it's in Russian)
https://www.drive2.ru/b/2615302
Gist : "55 minutes of public shame. From using German screwdrivers on Soviet sized screw slots, to COTS Flash chips and 4 layer boards supported by corners".
And then how this kind of stuff should be /was done in the past.
I like the quote:
"Despite anti-static smocks and gloves, the impression is of a company of drunkards who found a rusty thingie and are trying to extract precious metals from it"
 Mostly things this thread covered one way or another

[helius]

Thanks for the link, I'm fascinated by the picture of the ceramic hybrids assembly with green feedthroughs. So that was what shockproof electronics was like in the Soviet era?

[reagle]

I think those are space grade boards from 1992.


And then the board broken in pieces is from an exploded rocket of sorts

[helius]

Thanks, it makes sense that the feedthroughs would be for hard vacuum applications. The Bosch ABS assembly also looks very shock resistant:

The ceramic or CEM4 board is evenly supported by cement against a metal plate, and bond wires used for connections. It would need careful design to ensure that shorting the wires together wouldn't disrupt already stored data.

[SeanB]

Hand soldered, so not really shock proof, probably a RF front end for some satellite receiver or earth station, made with all Russian made parts. The Bosch boards are an example of what you do to make stuff vibration proof, not putting the thin boards where vibration and forces are applied only to the weakest point of the board and where the most flexing can occur.

Using a decent ceramic core with bonding wires as storage would be easy, you just do your assembly and final testing then carefully conformal coat with  flexible reenterable silicone conformal coat, so that all the wiring bonds are fixed firmly and do not move, though they might flex. The ceramic could be in a hybrid housing, extra protection and a lot more rugged than the PCB, and you can make the long walls thicker for added support, as you do not have many feedthroughs in this case with an on board controller. Then the lid is laser welded on after a final fill with encapsulant. That could fit in the same volume, but be a lot more rugged, especially if you use a simple silicone sleeve as both protection and mounting, spreading the mount over the entire container.

[Scrts]

And somebody believes that russians did not open that box before?

If they had information, that the bomber did not cross Turkey border, the box would probably be not damaged at all.