Author Topic: Standard electronics in space  (Read 20794 times)

Offline Hotblack Desiato

  • Full Member
  • ***
  • Posts: 369
  • Austria
  • Liked: 74
  • Likes Given: 52
Standard electronics in space
« on: 01/13/2015 08:49 pm »
http://forum.nasaspaceflight.com/index.php?topic=36552.msg1315454#msg1315454

in this thread, I asked about using standard electronics with a radiation shielding on satellites, and mentioned regular notebooks on the ISS. the discussion quickly started revolving around the notebooks, that they don't last long...

so, I'm starting this thread here, because that discussion clearly doesn't belong into the other thread (spaceX entering the commercial satellite market) with a few questions:

1. how long does consumer-electronic last? so far I saw some notebooks on ISS-videos and Nikon full frame cameras. or are those devices specially designed to work (or last longer) in space?

2. what causes electronics to fail in space, whether directly exposed to space, like on satellites, on mars-rovers or perhaps on philae, or in a shielded environment like on the ISS?

3. how about repairing those devices? especially on missions to other planets, a failing laptop can't be easily replaced, but they might have a few spare parts for those devices. is that more complicated in zero-g? (I'm thinking about small screws floating around).

4. which parameters (radiation-shielding, temperature) are required to allow standard electronics to work for several years?
« Last Edit: 01/14/2015 11:50 am by Chris Bergin »

Offline robertross

  • Canadian Member
  • Senior Member
  • *****
  • Posts: 17939
  • Westphal, Nova Scotia
  • Liked: 659
  • Likes Given: 7692
Re: standard electronics in space
« Reply #1 on: 01/13/2015 09:48 pm »
Hard drives are routinely swapped on laptops
Cameras are now typically off-the-shelf, with only the battery chargers upgraded to meet ISS rquirements of power & safety.
Many (if not all) of the camera image sensors on orbit have hot pixels; clear evidence of radiation damage.
Radiation can also cause bit flips in data retention & transmission, causing 'upset' events.
Circuit traces are also typically much wider on rad-hardened systems.

Radiation damage to chips inside consumer/commercial electronics can be minor or major, some salvagable, some not. If you look at the postings of the Mars rovers, you can see they can re-route circuit paths for memory chip failures, but that is typically not possible on consumer products, unless you have software designed to ignore large damaged segments.

Typically any repair task is considered only when there are limited to no spares left on orbit. It's much easier to simply swap a unit out; hence why they fly so many spares on orbit on these re-supply missions. I visit to the ISS thread of any expedition will clearly show you that.


Offline Hotblack Desiato

  • Full Member
  • ***
  • Posts: 369
  • Austria
  • Liked: 74
  • Likes Given: 52
Re: standard electronics in space
« Reply #2 on: 01/13/2015 11:35 pm »
thanks for that answer, especially the explanation of how rad-hardening is usually done :)

okay, swapping harddrives isn't a task. although it's quite a pita to copy the backup onto the drive (okay, it just takes some time, maybe somebody from mission control can do that via teamviewer or a similar system).

but since SSDs are lighter, they might use them instead of regular hdds. bringing up another problem. the same situation that causes hot pixels on cameras (which can be erased from pictures quite easily) can also cause damages on the surface of flash-chips. and that can cause everything from slight pixel-errors in video files to severe malfunctions with the operating system.

regarding your last paragraph: I guess that's quite a big problem, if the crew is approaching mars or some other planet/dwarf planet/asteroid, without much spare part supply. it's "easy" for ISS, just request a new device, delivered by the next soyuz, progress, htv, cygnus or dragon capsule. for example: on mars, supply is up to 2 years away, and depending on which approach is finally used, they might even consider not to send that spare part.

Offline robertross

  • Canadian Member
  • Senior Member
  • *****
  • Posts: 17939
  • Westphal, Nova Scotia
  • Liked: 659
  • Likes Given: 7692
Re: standard electronics in space
« Reply #3 on: 01/14/2015 12:08 am »
thanks for that answer, especially the explanation of how rad-hardening is usually done :)

okay, swapping harddrives isn't a task. although it's quite a pita to copy the backup onto the drive (okay, it just takes some time, maybe somebody from mission control can do that via teamviewer or a similar system).

but since SSDs are lighter, they might use them instead of regular hdds. bringing up another problem. the same situation that causes hot pixels on cameras (which can be erased from pictures quite easily) can also cause damages on the surface of flash-chips. and that can cause everything from slight pixel-errors in video files to severe malfunctions with the operating system.

regarding your last paragraph: I guess that's quite a big problem, if the crew is approaching mars or some other planet/dwarf planet/asteroid, without much spare part supply. it's "easy" for ISS, just request a new device, delivered by the next soyuz, progress, htv, cygnus or dragon capsule. for example: on mars, supply is up to 2 years away, and depending on which approach is finally used, they might even consider not to send that spare part.


For Mars, I'm sure most, if not all the hardware will be rad-hard, for the reason you state: ISS is just that much closer.
It also allows the ground teams to fix any hardware issues with the next generation of the item sent up, whereas for Mars: you're stuck with the spares you have. If you have a bad batch, you have nothing. There was a recent sci-fi program on TV noting deficient avionics cards on the spacecraft causing trouble. The US (and other) military components have been found with counterfit components in their planes - it can happen.

SSD drives are probably more vulnerable to upsets (IIRC), whereas a HDD uses a magnetic field which 'polarizes' an area of magnetic particles (spot).

Another note on rad-hard chips. All the ones I've seen have a gold-plated metal cap which is soldered/welded over the chip mounting area, and the chip base is usually ceramic as opposed to plastic (although that is likely for higher temperature ratings).

Electronic components for the military & space are typically screened by various means. The more reliable you expect or require something to be, the more intensive the screening.; they need to route out premature failure and determine the mean time before failure (MTBF). 100% visual examination (by microscope), 100% burn-in, batch sampling of various testing procedures (sometimes over tolerance testing), baking, cold effects, pin and package cleaning.

Offline robertross

  • Canadian Member
  • Senior Member
  • *****
  • Posts: 17939
  • Westphal, Nova Scotia
  • Liked: 659
  • Likes Given: 7692
Re: standard electronics in space
« Reply #4 on: 01/14/2015 12:12 am »
here's a quick source document I found on Texas Instruments Space Qualified components. Has a fairly good write-up. Check the beginning & end sections

http://www.ti.com/lit/sg/sgzt006d/sgzt006d.pdf

Offline ThereIWas3

  • Full Member
  • ****
  • Posts: 948
  • Liked: 500
  • Likes Given: 338
Re: standard electronics in space
« Reply #5 on: 01/14/2015 12:25 am »
SSD/flash memory also has a limited life regardless of radiation.   This is due to how they work; each block of memory can only be written to a few thousand times.

Offline mtakala24

Re: standard electronics in space
« Reply #6 on: 01/14/2015 01:19 am »
I was under the impression that all COTS hardware is almost certainly equipped with firmware modifications to keep fans running (no convective cooling) or in case of system without fans, small fans installed permanently.

Also, I though the digital cameras had redundant firmware chips, so one could continue to use the camera even if the reprogrammable chip had fried because of radiation. I don't know if that applied to some previous camera models, though.

Offline Hotblack Desiato

  • Full Member
  • ***
  • Posts: 369
  • Austria
  • Liked: 74
  • Likes Given: 52
Re: standard electronics in space
« Reply #7 on: 01/14/2015 07:02 am »
SSD/flash memory also has a limited life regardless of radiation.   This is due to how they work; each block of memory can only be written to a few thousand times.

that applies for consumer-chips with Multi Level Cells (MLC), which can save 2 or even 3 states simultanously. Single Level Cells (SLC) are capable of 100.000 to several million write-erase-cycles.

I was under the impression that all COTS hardware is almost certainly equipped with firmware modifications to keep fans running (no convective cooling) or in case of system without fans, small fans installed permanently.

Also, I though the digital cameras had redundant firmware chips, so one could continue to use the camera even if the reprogrammable chip had fried because of radiation. I don't know if that applied to some previous camera models, though.

such firmwaremodifications should be easy. just alter a few configs. multiple firmware chips however, that might be really complicated and hard to achieve. as far as I know, sandisk once produced write-once memory chips. they might be immune. the downside is, that a firmware update isn't possible at all.

here's a quick source document I found on Texas Instruments Space Qualified components. Has a fairly good write-up. Check the beginning & end sections

http://www.ti.com/lit/sg/sgzt006d/sgzt006d.pdf

thanks, I will take a look at this.

Offline AS-503

  • Full Member
  • ****
  • Posts: 494
  • Orion Fab Team
  • Colorado USA
  • Liked: 317
  • Likes Given: 251
Re: standard electronics in space
« Reply #8 on: 01/14/2015 07:52 am »
http://forum.nasaspaceflight.com/index.php?topic=36552.msg1315454#msg1315454

in this thread, I asked about using standard electronics with a radiation shielding on satellites, and mentioned regular notebooks on the ISS. the discussion quickly started revolving around the notebooks, that they don't last long...

so, I'm starting this thread here, because that discussion clearly doesn't belong into the other thread (spaceX entering the commercial satellite market) with a few questions:

1. how long does consumer-electronic last? so far I saw some notebooks on ISS-videos and Nikon full frame cameras. or are those devices specially designed to work (or last longer) in space?

2. what causes electronics to fail in space, whether directly exposed to space, like on satellites, on mars-rovers or perhaps on philae, or in a shielded environment like on the ISS?

3. how about repairing those devices? especially on missions to other planets, a failing laptop can't be easily replaced, but they might have a few spare parts for those devices. is that more complicated in zero-g? (I'm thinking about small screws floating around).

4. which parameters (radiation-shielding, temperature) are required to allow standard electronics to work for several years?


I’ll try to add some useful information here. Keep in mind that there are big differences in the “Space” environment. LEO is much less toxic than the Van Allen belts just outside of LEO. Likewise the temperature differences at Mars vs. Mercury are different. There are also special cases with LEO like The South Atlantic Anomaly  http://en.wikipedia.org/wiki/South_Atlantic_Anomaly

1.   You’re asking about longevity but you only note devices you’ve seen in LEO. “Space” is much bigger and usually more lethal to man and machine than LEO.

2.   Gamma rays, high velocity z-particles, and stronger generic radiation can all cause malfunction/damage to conventional electronics. Some are benign while some may be potentially crippling depending on the circuit design and to some extent the software.

3.   A re-boot may resolve a bit flip, but depending on the operational environment the overall design (hardware and software) may require either rad-hardened or a complex multi-redundant voting scheme.

4.   The design itself (and the operational environment) will dictate general parameters on the needed shielding and temperature requirements. “Standard” is almost too ambiguous.

For further reading ex-NASA engineer extraordinaire John Muratore outlines the SpaceX approach to a non-radhardened solution. 
http://aviationweek.com/blog/dragons-radiation-tolerant-design
http://en.wikipedia.org/wiki/Byzantine_fault_tolerance

Offline Proponent

  • Senior Member
  • *****
  • Posts: 7277
  • Liked: 2782
  • Likes Given: 1462
Re: Standard electronics in space
« Reply #9 on: 01/14/2015 08:56 am »
At what voltages and frequencies (or DC) is power supplied on ISS and other spacecraft these days?  I seem to recall reading about some quite high frequencies (400 Hz?) being used and have wondered why (maybe less potential for arcing?).
« Last Edit: 01/14/2015 11:51 am by Chris Bergin »

Offline robertross

  • Canadian Member
  • Senior Member
  • *****
  • Posts: 17939
  • Westphal, Nova Scotia
  • Liked: 659
  • Likes Given: 7692
Re: Standard electronics in space
« Reply #10 on: 01/14/2015 03:36 pm »
At what voltages and frequencies (or DC) is power supplied on ISS and other spacecraft these days?  I seem to recall reading about some quite high frequencies (400 Hz?) being used and have wondered why (maybe less potential for arcing?).
http://www.boeing.com/boeing/defense-space/space/spacestation/systems/eps.page

160V DC

400Hz is a standard for many aircraft. As I found out online, it allows many of the components to be lighter & smaller (IE: transformers) as they don't need to store as more power in the magnetic core like 60 Hz does.

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 37441
  • Cape Canaveral Spaceport
  • Liked: 21451
  • Likes Given: 428
Re: Standard electronics in space
« Reply #11 on: 01/14/2015 04:24 pm »
At what voltages and frequencies (or DC) is power supplied on ISS and other spacecraft these days?  I seem to recall reading about some quite high frequencies (400 Hz?) being used and have wondered why (maybe less potential for arcing?).

ISS is different than other spacecraft.  Most spacecraft operate around 28V DC

Offline wolfpack

  • Full Member
  • ****
  • Posts: 743
  • Wake Forest, NC
  • Liked: 159
  • Likes Given: 4
Re: Standard electronics in space
« Reply #12 on: 01/14/2015 07:15 pm »
At what voltages and frequencies (or DC) is power supplied on ISS and other spacecraft these days?  I seem to recall reading about some quite high frequencies (400 Hz?) being used and have wondered why (maybe less potential for arcing?).

The iron core in a transformer will saturate at a certain magnetic flux density. For 60 Hz (16.7 ms period) you need more iron than 400 Hz (2.5 ms period) to avoid that. When the core saturates it stops coupling flux from the primary winding to the secondary winding and then the primary winding just becomes a short. Think fire.

So aircraft like 400 Hz for weight savings. It is annoying though because it is quite audible.

Future technology should be all DC with DC/DC converters everywhere. It would also eliminate the constant speed drives (CSDs), which are basically giant automatic transmissions bolted to aircraft engines to spin at 400 Hz regardless of turbine speed.

Offline ThereIWas3

  • Full Member
  • ****
  • Posts: 948
  • Liked: 500
  • Likes Given: 338
Re: Standard electronics in space
« Reply #13 on: 01/15/2015 03:44 am »
Higher voltage means less current for the same amount of power transfer.  Less current needs thinner wires, which can save a lot of weight.  With the distances inside ISS being an order of magnitude greater than in most smaller spacecraft, it makes sense to use higher voltages.

If the wire you use is too small, it heats up.  Way too much, and it melts and starts fires.  But even if it just gets a little warm, that heat is lost energy that does not make it to the far end.

Offline AJA

  • Full Member
  • ****
  • Posts: 889
  • Per Aspera Ad Ares, Per Aspera Ad Astra
  • India
  • Liked: 146
  • Likes Given: 212
Re: Standard electronics in space
« Reply #14 on: 01/15/2015 09:53 am »
Let me also add that radiation doesn't only cause bit-flips. Ions from this radiation can get deposited inside the semi-conductor (depending on your radiation flux of course) and alter the semi-conductor physics in a manner where the range of voltages produced by an element can shift and expand/contract. This can manifest itself as sensor drift.. and depending on the software control/sensing logic, can lead to a spurious readings and instrumentation anomalies after a period of extended operation. Some sensors are re-calibrated from time to time, to establish a new baseline... but again, you have to have a greater degree of confidence in your calibration and test equipment, if you want to do this. So you use well defined, constant environmental parameters - e.g. if you wanted to re-calibrate an image sensor; you image the calibration target on your rover, compensate for photochemical fading (if any) of the various colours on it, and image the target at a known time of day, under similar illumination conditions as a previous calibration run. Then you tweak your image rendering algorithm to give you "accurate" hues etc. Alternatively, you compare sensor readings of the same target (at the same/similar times) with a newer sensor (of course, this requires access to it, and frequent increments/supplies).


Chang'E's (Yutu's) adventures also remind me of how cold-fractures in electronic circuit relays are a matter of concern. Oh..that reminds me of another challenge. The different ambient electrical environment of non-atmosphere-shielded planetary surfaces. The moon, for example sustains a large potential difference between the tenuous exosphere (ionised by solar radiation, so much so that microscopic dust particles are charged and repel each other, and end up levitating), and the surface. That's a spark hazard, and can screw around with your sensors, your communications, and probably even electronics - depending on how you've shielded various elements in the design.

At what voltages and frequencies (or DC) is power supplied on ISS and other spacecraft these days?  I seem to recall reading about some quite high frequencies (400 Hz?) being used and have wondered why (maybe less potential for arcing?).
The iron core in a transformer will saturate at a certain magnetic flux density. For 60 Hz (16.7 ms period) you need more iron than 400 Hz (2.5 ms period) to avoid that. When the core saturates it stops coupling flux from the primary winding to the secondary winding and then the primary winding just becomes a short. Think fire.


Calculations here


For further reading ex-NASA engineer extraordinaire John Muratore outlines the SpaceX approach to a non-radhardened solution. 
http://aviationweek.com/blog/dragons-radiation-tolerant-design
http://en.wikipedia.org/wiki/Byzantine_fault_tolerance



here's a quick source document I found on Texas Instruments Space Qualified components. Has a fairly good write-up. Check the beginning & end sections

http://www.ti.com/lit/sg/sgzt006d/sgzt006d.pdf


Thanks!
« Last Edit: 01/15/2015 09:56 am by AJA »

Offline MattMason

  • Full Member
  • ****
  • Posts: 1062
  • Space Enthusiast
  • Indiana
  • Liked: 772
  • Likes Given: 2016
Re: Standard electronics in space
« Reply #15 on: 01/16/2015 02:05 pm »
If someone hasn't already noted this, for an interesting contrast, the Dragon cargo spacecraft uses a "radiation-tolerant" practice, rather than rad-hardening. In short: Add lots of voting computers.

http://aviationweek.com/blog/dragons-radiation-tolerant-design
"Why is the logo on the side of a rocket so important?"
"So you can find the pieces." -Jim, the Steely Eyed

Offline AS-503

  • Full Member
  • ****
  • Posts: 494
  • Orion Fab Team
  • Colorado USA
  • Liked: 317
  • Likes Given: 251
Re: Standard electronics in space
« Reply #16 on: 01/16/2015 05:13 pm »
If someone hasn't already noted this, for an interesting contrast, the Dragon cargo spacecraft uses a "radiation-tolerant" practice, rather than rad-hardening. In short: Add lots of voting computers.

http://aviationweek.com/blog/dragons-radiation-tolerant-design

Yup, mentioned twice already on a one page (thus far) thread.  ;)

Offline Llian Rhydderch

  • Full Member
  • ****
  • Posts: 1237
  • Terran Anglosphere
  • Liked: 1299
  • Likes Given: 9683
Re: Standard electronics in space
« Reply #17 on: 01/17/2015 03:36 pm »
If someone hasn't already noted this, for an interesting contrast, the Dragon cargo spacecraft uses a "radiation-tolerant" practice, rather than rad-hardening. In short: Add lots of voting computers.

http://aviationweek.com/blog/dragons-radiation-tolerant-design

Yup, mentioned twice already on a one page (thus far) thread.  ;)

It may have been mentioned a couple of times, but I'll add here that the "fault-tolerant" design, at both the hardware-design and software-design level, will be the key critical piece of technology that will make the use of lower-cost (non-rad-hardened) electronics in space possible. 

So it is an important part of making standard-electronics work in space, and thus very relevant to this thread.  And when solved, it also gets us flying electronics in space that are only a year or two-old, and thus flying electronic technology that is much closer to the state-of-the-art.
Re arguments from authority on NSF:  "no one is exempt from error, and errors of authority are usually the worst kind.  Taking your word for things without question is no different than a bracket design not being tested because the designer was an old hand."
"You would actually save yourself time and effort if you were to use evidence and logic to make your points instead of wrapping yourself in the royal mantle of authority.  The approach only works on sheep, not inquisitive, intelligent people."

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 37441
  • Cape Canaveral Spaceport
  • Liked: 21451
  • Likes Given: 428
Re: Standard electronics in space
« Reply #18 on: 01/17/2015 03:46 pm »

It may have been mentioned a couple of times, but I'll add here that the "fault-tolerant" design, at both the hardware-design and software-design level, will be the key critical piece of technology that will make the use of lower-cost (non-rad-hardened) electronics in space possible. 

 And when solved, it also gets us flying electronics in space that are only a year or two-old, and thus flying electronic technology that is much closer to the state-of-the-art.

Nonsense.  It may never happen.   There has yet to be a long duration test to prove out the practice yet.   The Dragon is quiescent for most of its ISS mission.  It is only really active for less than 3 days. 

That doesn't even happen in the aircraft business.  And if reusability happens, it will be no different than now due to reuse of the avionics.
« Last Edit: 01/17/2015 03:50 pm by Jim »

Offline baldusi

  • Senior Member
  • *****
  • Posts: 8356
  • Buenos Aires, Argentina
  • Liked: 2539
  • Likes Given: 8273
Re: Standard electronics in space
« Reply #19 on: 01/17/2015 04:39 pm »
If someone hasn't already noted this, for an interesting contrast, the Dragon cargo spacecraft uses a "radiation-tolerant" practice, rather than rad-hardening. In short: Add lots of voting computers.

http://aviationweek.com/blog/dragons-radiation-tolerant-design

Yup, mentioned twice already on a one page (thus far) thread.  ;)

It may have been mentioned a couple of times, but I'll add here that the "fault-tolerant" design, at both the hardware-design and software-design level, will be the key critical piece of technology that will make the use of lower-cost (non-rad-hardened) electronics in space possible. 

So it is an important part of making standard-electronics work in space, and thus very relevant to this thread.  And when solved, it also gets us flying electronics in space that are only a year or two-old, and thus flying electronic technology that is much closer to the state-of-the-art.
Have you stopped a bit to think about it? Who does the voting? How do they communicate their votes? How does the bonding, the layers and the doping react to radiation? How does it handle parasitic charges? They start from COTS, but at least industrial grade COTS parts. And that only works for LEO short duration.
Are you aware that gamma rays and high energy protons can actually physically damage parts? How do you make sure you get the correct message from each IC? You need ECC in every single step, even in the cache lines. But that's not the worse part. What happens if you get an error in your voting logic? What happens if your voting IC gets damages because of a relativistic nucleus?

Offline Llian Rhydderch

  • Full Member
  • ****
  • Posts: 1237
  • Terran Anglosphere
  • Liked: 1299
  • Likes Given: 9683
Re: Standard electronics in space
« Reply #20 on: 01/17/2015 05:21 pm »
If someone hasn't already noted this, for an interesting contrast, the Dragon cargo spacecraft uses a "radiation-tolerant" practice, rather than rad-hardening. In short: Add lots of voting computers.

http://aviationweek.com/blog/dragons-radiation-tolerant-design

Yup, mentioned twice already on a one page (thus far) thread.  ;)

It may have been mentioned a couple of times, but I'll add here that the "fault-tolerant" design, at both the hardware-design and software-design level, will be the key critical piece of technology that will make the use of lower-cost (non-rad-hardened) electronics in space possible. 

So it is an important part of making standard-electronics work in space, and thus very relevant to this thread.  And when solved, it also gets us flying electronics in space that are only a year or two-old, and thus flying electronic technology that is much closer to the state-of-the-art.
Have you stopped a bit to think about it? Who does the voting? How do they communicate their votes? How does the bonding, the layers and the doping react to radiation? How does it handle parasitic charges? They start from COTS, but at least industrial grade COTS parts. And that only works for LEO short duration.
Are you aware that gamma rays and high energy protons can actually physically damage parts? How do you make sure you get the correct message from each IC? You need ECC in every single step, even in the cache lines. But that's not the worse part. What happens if you get an error in your voting logic? What happens if your voting IC gets damages because of a relativistic nucleus?

Absolutely have thought about it.  I worked in an industry building high-availability, fault-tolerant electronics and software for commercial markets, and did both engineering in, and management of, the embedded software group.  Prior to that, I designed dense integrated circuits for a living, ICs that required extra silicon real-estate to support thorough post-fab line 100% circuit testing (of, then, 100s of millions of transistors), and could do the same as a part of an extended power-on-self-test if it were necessary to do so. 

Our chips were not extra-special, and certainly not anything like the whole rad-hardened game in the space industrial complex.  We made it work to our high-availability specs by using parallel capabilities, or as Elon and SpaceX are doing, triple-redundant capabilities.  The trade is more chips, circuit-board real-estate, and power--plus some software complexity--for higher-reliability.  When moving this to the space environment, don't forget how much capability is now possible in smaller and lower-power electronic packages; eventually, the cost to a design of extra redundant componentry, even when connected by sophisticated embedded software and very high-reliability testing requirements, is much less than the alternative "big iron" approach, the approach Musk calls "Battlestar Galactica".  I know, because our company regularly won big contracts competing against "big iron" solutions from the larger established players.

I am aware of both the temporary rad events as well as potential permanent rad-induced events in the space environment.  There are solutions, even using the relatively newer industrial grade electronics you mentioned, to both.

In my preliminary thinking, Musk will likely do two things to achieve his system reliability objectives here: 1) increase (but not super gold plate) his design with standard (or novel) fault-tolerant techniques to provide high-availability of any single satellite; as well as 2) design in system-level redundancy that can handle single-sat faults for days at a time (and possibly even double-consecutive-in-orbit sats for periods of time), where the system keeps running, but possibly at temporary reductions in communication throughput.

So, ya, I have thought about it, quite a bit.

Re arguments from authority on NSF:  "no one is exempt from error, and errors of authority are usually the worst kind.  Taking your word for things without question is no different than a bracket design not being tested because the designer was an old hand."
"You would actually save yourself time and effort if you were to use evidence and logic to make your points instead of wrapping yourself in the royal mantle of authority.  The approach only works on sheep, not inquisitive, intelligent people."

Offline Remes

  • Full Member
  • ****
  • Posts: 434
  • Germany
  • Liked: 246
  • Likes Given: 142
Re: Standard electronics in space
« Reply #21 on: 01/17/2015 11:04 pm »
But that's not the worse part. What happens if you get an error in your voting logic? What happens if your voting IC gets damages because of a relativistic nucleus?
3 Computer, 3 networks, every computer have each sensor, every computer does the voting on sensors, makes its control loops, error detection, diagnostics, ... , sends it result to the other 2. All Computer have all values, they compare, and if one is rogue, he gets deactivated.

Actuators: A space shuttle TVC had 4 servovalves per actuator, every valve was controlled by one computer. Many valves can have 3 coils on one core. Most servovalves have 2 windings. There shouldn't be a single point of failure. I even saw a mission critical actuator with a electromechanical relais in it. It could switch between two control authorities.

I have seen 4 valves arrangements in satellites, 2 in series parallel to 2 other in series.

You can also build mixing circuits which are inherently safe agains radiation.

But I agree, this still might not work outside LEO without rad hardening.

My guess for the future: radiation hardened FPGA. Price should go down, if most companies agree on a few devices  (2 supplier/4 sizes = 8 devices covering everything).
« Last Edit: 01/17/2015 11:15 pm by Remes »

Offline A_M_Swallow

  • Elite Veteran
  • Senior Member
  • *****
  • Posts: 8906
  • South coast of England
  • Liked: 500
  • Likes Given: 223
Re: Standard electronics in space
« Reply #22 on: 01/18/2015 04:25 am »
By using 3U and 6U cubesats with on board ion thrusters NASA should be able to get 1kg masses to high radiation orbits like GEO for about $1 million, starting in 4-5 years time.  At that sort of price the agency can afford to perform long term radiation testing on components and circuit boards.

Offline Hotblack Desiato

  • Full Member
  • ***
  • Posts: 369
  • Austria
  • Liked: 74
  • Likes Given: 52
Re: Standard electronics in space
« Reply #23 on: 01/18/2015 09:36 pm »
http://www.heise.de/ct/artikel/Alexander-Gerst-Da-mussten-wir-in-MacGyver-Manier-reparieren-2514587.html

I found this article on heise.de (I'm reading there regularly). an interview with alexander gerst regarding the repair-tasks at notebooks, and the cameras on the ISS. sadly, it's just in german, therefore use a translationservice to read it.

for me, the most interesting parts of the interview were:

Quote
Den mussten wir dann in MacGyver-Manier mit LeatherMan-Tool, Sägeblatt und Rasierschaum reparieren. Der Rasierschaum hat verhindert, dass die Späne unkontrolliert wegflogen.

they had to repair something MacGyver-style with LeatherMan, sawblade and shaving cream. the shaving cream was necessary to avoid metal particles to float away (and cause shortages).

Quote
...dass nichtkritische Systeme der Raumstation am besten auf einem standardisierten Notebook laufen, weil man dann Ersatzteile hat. Das sind diese Lenovo Thinkpads, früher waren es die A31p, jetzt A61p.

non-critical parts of the station use lenovo thinkpads, A61p and formerly A31p.

Quote
Von denen haben wir sehr viele an Bord. Wenn die Weltraumstrahlung ein System zerschießt – und das passiert ab und zu mal, besonders an Tagen mit starker Aurora – nimmt man sich einfach ein neues, baut die Festplatte ein und es läuft innerhalb von 10 Minuten. Wenn die Festplatte zerschossen ist, nimmt man die heraus, die neue wird vom Boden aus mit den aktuellen Daten konfiguriert und das Ding funktioniert wieder.

they have a lot of them on board. and if the radiation kills one, and that happens occasionally, especially on days with strong auroras - they just grab a new one, replace the hard drive and continue, within 10 minutes. if the harddrive died, ground control sends a new image to a fresh harddrive and then they configure it properly, and it works again. (rough translation, maybe someone is better at it).

he then talks about taking pictures in space, that the Nikon D2XS has vacuum-certification and therefore can be used during EVAs without external casing. and that they have problems with hot pixels, and so on (he even took dark images... only the hot pixels glowed, and his colleagues down on earth subtracted those images to clean them).



so at least that answers (toghether with your postings)  quite a bit regarding radiation hardening, and that the ISS-shielding is not sufficient to operate regular hardware in space without problems, however it's possible to circumvent this situation with clever hardware-design. I gues, for satellites which should operate at least 5 years in space (more is better), they need at least some rad-hardening. there is a fair chance that quite a part of the systems fail within that time.

and without protection of earths magnetic field, the problems will increase.

Offline mlindner

  • Software Engineer
  • Senior Member
  • *****
  • Posts: 2908
  • Space Capitalist
  • Silicon Valley, CA
  • Liked: 2204
  • Likes Given: 818
Re: Standard electronics in space
« Reply #24 on: 01/19/2015 10:39 pm »
Just a note: Cubesats have been doing standard phone technology electronics in space for years.

Personal experience here with this too. They're not hard. Only thing I'll say is don't use SD cards, those things were by far the fastest element to fail in our experience. Something in them dies really quickly in radiation environments.

Other than that just put watchdogs everywhere (and watchdogs watching watchdogs). If anything fails, the whole satellite or just the component gets a power cycle. Ignore and continue.
« Last Edit: 01/19/2015 10:41 pm by mlindner »
LEO is the ocean, not an island (let alone a continent). We create cruise liners to ride the oceans, not artificial islands in the middle of them. We need a physical place, which has physical resources, to make our future out there.

Offline wolfpack

  • Full Member
  • ****
  • Posts: 743
  • Wake Forest, NC
  • Liked: 159
  • Likes Given: 4
Re: Standard electronics in space
« Reply #25 on: 01/20/2015 03:30 pm »
Personal experience here with this too. They're not hard. Only thing I'll say is don't use SD cards, those things were by far the fastest element to fail in our experience. Something in them dies really quickly in radiation environments.

SD cards need +12V to program/erase the "bits" inside them. Since most portable electronics don't have this voltage, there's usually a charge pump integrated onto the NAND flash chip that generates it. If that charge pump uses P/N junction diodes then I'd wager they're the failure point. Slam enough charged particles into the depletion region of a P/N junction and you will render the diode permanently open ("off") or permanently shorted ("on").

Offline AnalogMan

  • Member
  • Senior Member
  • *****
  • Posts: 3431
  • Cambridge, UK
  • Liked: 1602
  • Likes Given: 50
Re: Standard electronics in space
« Reply #26 on: 01/20/2015 06:30 pm »
Personal experience here with this too. They're not hard. Only thing I'll say is don't use SD cards, those things were by far the fastest element to fail in our experience. Something in them dies really quickly in radiation environments.

SD cards need +12V to program/erase the "bits" inside them. Since most portable electronics don't have this voltage, there's usually a charge pump integrated onto the NAND flash chip that generates it. If that charge pump uses P/N junction diodes then I'd wager they're the failure point. Slam enough charged particles into the depletion region of a P/N junction and you will render the diode permanently open ("off") or permanently shorted ("on").

You are correct that the charge pump circuit is usually the first part of flash memories to fail under increasing irradiation, rather than the memory cells.

My understanding is that the thicker oxide gates in the high voltage MOSFETs of the pumps are more susceptible to charge trapping within the gates, leading to a change in voltage switching thresholds.  The charge pumps are quite sensitive to shifts in the gate threshold voltages and stop working properly.  This stops the memories being erased or written, but they can often still be read correctly since this relies only on the low voltage circuitry functioning correctly.

Offline watermod

  • Full Member
  • ****
  • Posts: 519
  • Liked: 177
  • Likes Given: 153
Re: Standard electronics in space
« Reply #27 on: 01/23/2015 12:53 am »
Curious if optical processors have been tried in space.
If so, I wonder if they are more immune to radiation problems.

Offline Jarnis

  • Full Member
  • ****
  • Posts: 1313
  • Liked: 830
  • Likes Given: 201
Re: Standard electronics in space
« Reply #28 on: 01/23/2015 03:56 pm »
Curious if optical processors have been tried in space.
If so, I wonder if they are more immune to radiation problems.

Has anyone tried an optical processor yet? Anywhere?

I thought they were still firmly vaporware. Interesting research, sure, but still quite a long way from practical applications.

Offline watermod

  • Full Member
  • ****
  • Posts: 519
  • Liked: 177
  • Likes Given: 153
Re: Standard electronics in space
« Reply #29 on: 01/23/2015 04:09 pm »
Curious if optical processors have been tried in space.
If so, I wonder if they are more immune to radiation problems.

Has anyone tried an optical processor yet? Anywhere?

I thought they were still firmly vaporware. Interesting research, sure, but still quite a long way from practical applications.

here is a hybrid partial optical DSP from 2003 http://www.theregister.co.uk/2003/10/31/israelis_ship_eight_teraops_optical/

Tags:
 

Advertisement NovaTech
Advertisement Northrop Grumman
Advertisement
Advertisement Margaritaville Beach Resort South Padre Island
Advertisement Brady Kenniston
Advertisement NextSpaceflight
Advertisement Nathan Barker Photography
0