LIVE: SpaceX Dragon CRS-1 (SpX-1) (EOM) Unberthing, Entry, Splashdown

[edkyle99]

I'm pretty sure Shuttle had occasional issues with one of its triple redundant flight computers... Was the same "sky is falling" mentality expressed about that?

True, but Shuttle had five GPCs, all of which were rad hardened - making total loss much less likely.

The following paper describes the improved GPCs and single-event upset events on their non-radiation hardened SRAM chips.  Note that software was included to constantly check for upset events, to do error checking and correction, etc.  I wonder if SpaceX does all of that.

http://klabs.org/DEI/Processor/shuttle/oneill_94.pdf

 - Ed Kyle

I wouldn't call that software.  That was done outside of the software's "knowledge".  It was a dedicated the system that constantly scrubbed memory.  Maybe system-level routines would be a better choice, although that's not entirely accurate either.  It ran much faster than software of its era could do.

O.K.  So does anyone know if SpaceX is not using an extra memory scrubber system like this, to minimize radiation upset event effects in non-rad-hardened hardware?  That's how Shuttle did it.

 - Ed Kyle 

[Robotbeat]

I'm pretty sure Shuttle had occasional issues with one of its triple redundant flight computers... Was the same "sky is falling" mentality expressed about that?

True, but Shuttle had five GPCs, all of which were rad hardened - making total loss much less likely.

The following paper describes the improved GPCs and single-event upset events on their non-radiation hardened SRAM chips.  Note that software was included to constantly check for upset events, to do error checking and correction, etc.  I wonder if SpaceX does all of that.

http://klabs.org/DEI/Processor/shuttle/oneill_94.pdf

 - Ed Kyle

I wouldn't call that software.  That was done outside of the software's "knowledge".  It was a dedicated the system that constantly scrubbed memory.  Maybe system-level routines would be a better choice, although that's not entirely accurate either.  It ran much faster than software of its era could do.

O.K.  So does anyone know if SpaceX is not using an extra memory scrubber system like this, to minimize radiation upset event effects in non-rad-hardened hardware?  That's how Shuttle did it.

 - Ed Kyle 

Such systems can be had off the shelf these days.

[MikeAtkinson]

Its a long time since I did this, but memory error detection and correction circuits had a write-back mode. If they detected an error and corrected it they wrote the corrected value back into the memory. So all you had to do is read all the memory locations sequentially to scrub any errors out of the RAM. IIRC DMA was often used for this, at least some early processors could interleave DMA accesses with full speed processor instruction and data access, so memory scrubbing could occur without slowing the processor down.

It should be fairly easy to add scrubbing logic to a memory access FPGA which was already doing the EDC. The main problem is that EDC takes time, and this adds latency to every memory access.

[Robotbeat]

Its a long time since I did this, but memory error detection and correction circuits had a write-back mode. If they detected an error and corrected it they wrote the corrected value back into the memory. So all you had to do is read all the memory locations sequentially to scrub any errors out of the RAM. IIRC DMA was often used for this, at least some early processors could interleave DMA accesses with full speed processor instruction and data access, so memory scrubbing could occur without slowing the processor down.

It should be fairly easy to add scrubbing logic to a memory access FPGA which was already doing the EDC. The main problem is that EDC takes time, and this adds latency to every memory access.

While these techniques do indeed go beyond simple ECC, it should be noted that it's very difficult to get a server mainboard without ECC these days. Cosmic rays affect electronics on the Earth's surface, too...

And you're right that ECC and other techniques usually do incur extra latency.

[mmeijeri]

O.K.  So does anyone know if SpaceX is not using an extra memory scrubber system like this, to minimize radiation upset event effects in non-rad-hardened hardware?  That's how Shuttle did it.

IIUIC, that is what rad-hard hardware does too. It's not a question of rad-hard vs scrubbers and redundancy. Redundancy is an important part of radiation hardening, both at the gate and system level. Rad-hard hardware isn't immune to radiation, it's just less susceptible to it. Depending on the technology used it may be immune to latchups and be able to withstand a greater total dose. But from what we've heard, latchup or total dose were not the problem here.

[Comga]

intersting tweet

If we keep landing this precisely, we're going to have to start issuing the recovery team titanium umbrellas. #Dragon

jb

Bumping to repeat the question:
Does anyone have a quantified measure of "this percisely" for CRS-1?
Still looking for a value for COTS-2+, also. 
Supposedly COTS-1 was within 0.8 km of the target, COTS-2+ had a much greater offset, and now this tweet.

[Pedantic Twit]

Bumping to repeat the question:
Does anyone have a quantified measure of "this percisely" for CRS-1?

Does this help? (new(?) footage including recovery)

http://www.youtube.com/watch?v=K0wZ1MOzFSU&feature=youtu.be

[corrodedNut]

If this wasn't a bullseye it was pretty close. Seems like the camera is looking right up into the chutes, you can even make out the drogues coming down seperately.

Looks like two ship flotilla. When COTS 2/3 came back, they said they had a telemetry vessel; I wondered where it was for this mission, and there it is. They've got it boiled down now, two ships, two ridgid inflatables. One vessel for recovery, the other for support and telemetry.

[corrodedNut]

Wait a minute, I take that back. There's at least three ships, here's the third:

I stand corrected, see below.

[rickl]

Was there an airplane this time?  COTS 2+ had a NASA P-3, as I recall.

[corrodedNut]

Was there an airplane this time?  COTS 2+ had a NASA P-3, as I recall.

As you can see from the second photo I posted, there was some sort of aircaft involved. Most likely a fixed-wing, based on the range involved. Who and what is unknown.

[rickl]

As you can see from the second photo I posted

{blush} 

Having grown up in the 60s watching Gemini and Apollo, I think it's very cool that they can recover Dragon with such a minimal recovery fleet, from an economic standpoint.  No aircraft carrier task force necessary.

[Comga]

Bumping to repeat the question:
Does anyone have a quantified measure of "this percisely" for CRS-1?

Does this help? (new(?) footage including recovery)

Well, that's not the quantitative answer I am looking for, but it is new footage to me and very welcome.  Many things I hadn't seen before. Thanks!

However, if anyone has a value for the offset distance , it would be appreciated.

[Thunderbird5]

Having grown up in the 60s watching Gemini and Apollo, I think it's very cool that they can recover Dragon with such a minimal recovery fleet, from an economic standpoint.  No aircraft carrier task force necessary.

Not quite apples for apples; I would imagine that the recovery support for a manned capsule would be significantly larger.

[rickl]

Having grown up in the 60s watching Gemini and Apollo, I think it's very cool that they can recover Dragon with such a minimal recovery fleet, from an economic standpoint.  No aircraft carrier task force necessary.

Not quite apples for apples; I would imagine that the recovery support for a manned capsule would be significantly larger.

Yes, I've thought of that, but I have a feeling that SpaceX will be doing propulsive landings on dry land before they fly the manned Dragon.

[schaban]

Wait a minute, I take that back. There's at least three ships, here's the third:

Looks to me like the top floor of the 1st ship, which hauled Dragon. Mast, flag and short, see-through, cabin seems to match...

[Silmfeanor]

Wait a minute, I take that back. There's at least three ships, here's the third:

Looks to me like the top floor of the 1st ship, which hauled Dragon. Mast, flag and short, see-through, cabin seems to match...

Agreed. On the right you can see the bracket that holds the crane when it is not used.
so, 2 ships.

[Avron]

Wait a minute, I take that back. There's at least three ships, here's the third:

I stand corrected, see below.

Do we know the name of this ship?

[corrodedNut]

Do we know the name of this ship?

The main recovery vessel is the American Islander, the other, smaller ship is probably this one:
http://www.amarinecorp.com/Vessels/Crew%20Boats/Gladys%20S%20Spec.pdf

[Avron]

Do we know the name of this ship?

The main recovery vessel is the American Islander, the other, smaller ship is probably this one:
http://www.amarinecorp.com/Vessels/Crew%20Boats/Gladys%20S%20Spec.pdf

Awesome..  ty.. remember to keep an eye on these two next mission..