SpaceX COTS Demo 2/3 Updates (THREAD 2)

[rdale]

happy - March is the latest date they've mentioned. Unless you have some tidbits that they don't, I think for now that's the month they are shooting for.

[sdsds]

No, not a factor.

[...] Phobos-Grunt [...]

Thanks for confirmation of that.  The description in the article is a bit fuzzy; it rather sounds like SpaceX is planning a complete software "regression test," i.e. making sure new changes haven't inadvertently destablized parts of the software that worked before.  (This is a "good thing" to do, but can be time consuming.)  Can you confirm that interpretation?

[mr. mark]

A lot of good information on testing and delay in this article

"NASA sources said the company ran into problems with the planned rendezvous profile needed to guide the Dragon capsule to the space station. NASA dispatched a veteran flight director and trajectory analysts to Hawthorne to help SpaceX get to the bottom of the issue. Sources also said SpaceX engineers had encountered an electromagnetic interference issue with one or more components in the Dragon capsule.

A SpaceX spokeswoman confirmed that an EMI issued had been discovered during testing, but she characterized it as relatively minor. Likewise, Lindenmoyer downplayed the technical issues, saying "it's just good practice to wring out your hardware, your software, your operations to make sure you're in the best possible shape for a good successful mission."

http://www.spaceflightnow.com/falcon9/003/120120delay/

[Robotbeat]

A lot of good information on testing and delay in this article

"NASA sources said the company ran into problems with the planned rendezvous profile needed to guide the Dragon capsule to the space station. NASA dispatched a veteran flight director and trajectory analysts to Hawthorne to help SpaceX get to the bottom of the issue. Sources also said SpaceX engineers had encountered an electromagnetic interference issue with one or more components in the Dragon capsule.

A SpaceX spokeswoman confirmed that an EMI issued had been discovered during testing, but she characterized it as relatively minor. Likewise, Lindenmoyer downplayed the technical issues, saying "it's just good practice to wring out your hardware, your software, your operations to make sure you're in the best possible shape for a good successful mission."

http://www.spaceflightnow.com/falcon9/003/120120delay/

Thank you!

[ugordan]

A lot of good information on testing and delay in this article

That's a copy paste of the story from CBS News already linked to by rdale, even says so at the top of the article. Pay attention people...

[Space Pete]

So, there are basically five issues that SpaceX are currently working with Dragon:

. Integrated systems testing/closing out all "open ticket" items
. Flight software testing
. Flight procedures refinement
. Rendezvous profile refinement
. EMI resolution

Two months to solve all that? Hmm...

[ChefPat]

So, there are basically five issues that SpaceX are currently working with Dragon:

. Integrated systems testing/closing out all "open ticket" items
. Flight software testing
. Flight procedures refinement
. Rendezvous profile refinement
. EMI resolution

Two months to solve all that? Hmm...

I recognize whats happening here & it's not necessarily all about the hardware/software. It's also about "complex event processing."
Each step on the path from bending metal through splash down is complex on it's own & when added up is one of the most difficult endeavors possible.
While I want to see a Dragon berthed at the ISS yesterday, I'll be happy to wait until they're sure they have everything right before they launch.

[MP99]

Bill Harwood has a nice recap - http://www.cbsnews.com/network/news/space/home/spacenews/files/c51cc9aea9478427b4c6e472e596b685-377.html#unique-entry-id-377

"And they recognized they really hadn't tested the software to the same extent that they did for that previous flight and their engineers reported that to (SpaceX management) and they decided well, you know, they're going to take the time that's needed to make sure they're comfortable that flight software's ready to go."

Compare that with the reported development of Phobos-Grunt. You might almost wonder whether that fed a little into making the "SpaceX management" reconsider.

cheers, Martin

No, not a factor.

Thanks for that.

The "recognition" of the need for extra tests (more likely the recognition they were behind schedule for testing that was always intended) sounds good.

It would help if Elon would recognise their continued delivery against promises, and be a bit less gung-ho about future timescales.

cheers, Martin

[Norm38]

I hate EMI problems.  All specs can be met and an engineer can be absolutely convinced that there won't be interference, that there is no real design impact.  But if the reading is 2dB over the line, it's over the line and has to be hunted down.  And squashing EMI is like punching a water bed.  Right now someone at SpaceX is clamping ferrite beads on cables, one wire at a time...

[docmordrid]

^^---- ain't that the truth! It's like a game of Whack-a-Mole, with invisible moles. Grrrr....

[Ben the Space Brit]

Based on Norm and Doc's comments, I'm thinking that the next few weeks will be pretty much the Dragon C2/3 ground engineering team's definition of 'bad day' over and over again.

Sorry to hear it guys; really and sincerely. 

[Robotbeat]

I hate EMI problems.  All specs can be met and an engineer can be absolutely convinced that there won't be interference, that there is no real design impact.  But if the reading is 2dB over the line, it's over the line and has to be hunted down.  And squashing EMI is like punching a water bed.  Right now someone at SpaceX is clamping ferrite beads on cables, one wire at a time...

...which is why the move to optical cables where it's feasible makes a lot of sense, IMHO... There's extra cost, but it reduces the risk of EMI (over the length of the cables, at least). The potential for weight reduction or bandwidth increase is just a side benefit. With modern Ethernet-based avionics solutions, that should be easier than it was in the past.

[pippin]

...which is why the move to optical cables where it's feasible makes a lot of sense, IMHO...

Not necessarily. I've seen cases where EMC issues actually got worse due to the use of optical cables.
The cables help, but then it's rather simple to shield an electrical cable, too.
The problem with optical is the connectors which radiate a lot across a wise frequency band and which are hard to shield.
From a passive POV optical as an overall system is usually easier (it's less susceptible to radiation) which is good for Spaceflight application where you have external radiation but on the active side (radiating itself) it's worse and means you probably need additional margin for all your electronics components or you have additional design challenges like having to move connectors from your mainboards and things like that.

And then there is the _big_ issue of vibration, stress and harness placement, optical fibers tend to break much easier then copper.

[Nomadd]

I hate EMI problems.  All specs can be met and an engineer can be absolutely convinced that there won't be interference, that there is no real design impact.  But if the reading is 2dB over the line, it's over the line and has to be hunted down.  And squashing EMI is like punching a water bed.  Right now someone at SpaceX is clamping ferrite beads on cables, one wire at a time...

I had the unpleasent experience of working with Lockheed on some radios once. They decided that the average EM emission listed in the specs meant maximum and insisted that every unit meet that number. I used a few words mom wouldn't approve of.

 We generally have some margin between design for and acceptable. Say, design for 20db margin but accept 12db. Keeps tiny deviations from design from causing grief.

[iamlucky13]

So, there are basically five issues that SpaceX are currently working with Dragon:

. Integrated systems testing/closing out all "open ticket" items
. Flight software testing
. Flight procedures refinement
. Rendezvous profile refinement
. EMI resolution

Two months to solve all that? Hmm...

Possibly.

Most of the work should be able to happen in parallel. The guys working on EMI aren't depending on trajectory refinements to do their tasks and vice versus, for example.

But it's not out of the question that any of those can stretch out beyond 2 months. There's a lot of new systems on this launch, and more requirements to meet than on COTS 1.

[apace]

Most of the work should be able to happen in parallel. The guys working on EMI aren't depending on trajectory refinements to do their tasks and vice versus, for example.

About the EMI problem, I have my questions if they can solve this at the Cape. Do they have the shielded environment to do the testings or they need to transfer dragon back to their main facility?

[Robotbeat]

Most of the work should be able to happen in parallel. The guys working on EMI aren't depending on trajectory refinements to do their tasks and vice versus, for example.

About the EMI problem, I have my questions if they can solve this at the Cape. Do they have the shielded environment to do the testings or they need to transfer dragon back to their main facility?

Well, they did discover this at the Cape, so they can do SOME testing there...

[Prober]

Maybe combing Demo 2/3 was a bad idea?    Had Demo 2 launched by now these problem might have been fixed?   The focus on saving costs, and time, sometimes bites you and you find out in the end you didn’t save anything.

[dcporter]

Maybe combing Demo 2/3 was a bad idea?    Had Demo 2 launched by now these problem might have been fixed?   The focus on saving costs, and time, sometimes bites you and you find out in the end you didn’t save anything.

Yeah, maybe. Would love to hear other perspectives on this?

[wolfpack]

Maybe combing Demo 2/3 was a bad idea?    Had Demo 2 launched by now these problem might have been fixed?   The focus on saving costs, and time, sometimes bites you and you find out in the end you didn’t save anything.

Probably not. COTS 2 alone would be station prox ops, so I'm guessing a lot of these issues would need to be fixed anyway. Otherwise, why not just remanifest as COTS 2 only and launch on Feb 7? Clearly they are not doing that.

Combining Orbcomm was a bad idea, and since been undone.