:-)) I think you are pulling my chain. I am just an electronics designer. I've designed electronics for several satellites currently in geo orbit, which is one of the more difficult areas to design electronics for. To test my electronics, I often have to write software programs or scripts to automate the testing.
I'm sure you are a much better message board designer and plugin designer then I would be. You do it for a living! and I have never done it. But you know how programming is... you learn one language and you pretty much know them all (or you do with a little online language reference manual).
:-D
Just teasing you
But I agree about the whole "learn one language, know them all". Nothing about school pissed me off more than "learning java". Just teach programming concepts in general. Thx.
The idea of designing hardware for satellites is pretty fokken sweet though. I assume theyre communications satellites?
no, they're the ones armed with nukes they been puttin in orbit
So I've had this Dallas cowboys mini helmet digital clock in my room as long as I can remember and I have NEVER changed the batteries in it and it still works perfect. But here's the kicker this thing was made in 1991 and like I said it still works and I have never changed the batteries.
How the fuck does that happen?
d00d I has a laser pointer from 2007 and the fukkin thing still workz
[-(
blue turbins
From Those Fishes - I Fingered An Old Bitch (i got Aids on my finger)
WakeOfAshesPosts: 21,665destroyer of motherfuckers
The idea of designing hardware for satellites is pretty fokken sweet though. I assume they're communications satellites?
Generally I am just a communications hardware design engineer. I've worked on really high speed terrestrial fiber applications (SONET/SDH), microwave comms, satellite comms, and many different radio comms. Currently just working for a radio comm place.
The last satellite I designed was called 3GIRS. It is actually a big 4 megapixel 70 fps inferred video camera in a geo orbit. I designed about a third of the camera electronics, but was overall the main system integrator of all the camera electronics. The basic idea was to track ICBM's that happened to be coming our way. It is really just an engineering test at this point in time, and not a full scale development but it will be in orbit soon.
But yeah, in generally designing spaced based electronics offers a lot of interesting challenges. Like the electronics will go from behind the earth at -55 degrees C, to in direct view of the sun and 125 degrees C, almost instantly. And so the electronics not only need to survive that instant temp change, but also not hiccup. Not only that they are always being irradiated by massive amounts of sun radiation which in time will degrade the electronics and cause the thing to just stop working. No ways around that. There are interesting challenges, like you have to be careful not to use ROHS parts because they are mostly tin and tin will grow "whiskers" in space and has been know to brick your craft. Also the radiation will cause errors in your data, and so you need to do everything three times and then take the best 2 of 3 answer. Also... the parts you use are mega expensive. as an example... when I make earth based electronics, the fully populate board might cost 50 dollars to maybe 5 thousand dollars. A spaced based boards cost anywhere from 250k-1 million. It's pretty nerve racking programming a one time programming only computer chip that costs $100,000+. Try telling your boss you fucked that one up.
Wow... thats intense. I had no idea about the corruption from radiation, though it makes a lot of sense. How do you determine the "best" 2? Some kind of similarity algorithm?
WakeOfAshesPosts: 21,665destroyer of motherfuckers
Yeah they call the corruption from radiation "Single Event Upsets" and they can happen on earth because we receive small amounts of radiation here, but not enough for it to happen frequently. Sometimes radiation can actually cause very serious hard errors. There are like 3 types of those (Single Event Gate Rupture is one example) but they basically destroy your computer chips if they happen.
To prevent hard errors, they are very very picky about what silicon they use to make the parts. They shield the fuck out of the electronics. and they run the parts through a lot of radiation to make sure they wont just lock up and die. All of this is why space based electronics costs so much.
Most of the time, real processing work is being done by FPGA's and not processors. These are basically computer chips that can be programed to behave a specific way (I use the language VHDL or Verilog to define the logic functionality). Like you could take a Intel processor and program it to these FPGA's and then the FPGA would act like an Intel processor. The space based FPGA's are one time programmable devices and they actually natively perform the "best of 2". I only need to think in terms of there being 1 instance, and the tools automagically make 3 and automagically take the best of 3. I do understand it more then what I am saying here, but for the sake of my work, I dont really need to know it because it is done for me by the tools. This is why one of these FPGA's cost 100+ thousand dollars.
I think you are thinking in terms of software... and in general processors and software is not often used on orbit. It can be used, but you are limited to what you can do because so many errors can be injected into the system. I think when people use processors, all data packets use some sort of error correction. like convolution, or Turbo Product Codes.... Plus you shield the fuck out of it, and make sure it isnt doing anything that can't recover from radiation errors.
The science behind this stuff is pretty interesting... When you get into this stuff, you realize how funny it is that we take the earth ozone for granted.
WakeOfAshesPosts: 21,665destroyer of motherfuckers
Oh and you can think of radiation as kinda like a sun burn.... over time the burn gets much worst. At some point in time your electronics will get cancer and die and there isnt anything you can do to stop that, however if you glop on enough bullfrog, you can prevent it for a while :-D
I thought earth's radiation shield was primarily based on our magnetic field warping the particles around the planet instead of through it. So couldn't something like that be built, on a smaller scale? Though powering it would probably be a nightmare...
WakeOfAshesPosts: 21,665destroyer of motherfuckers
edited April 2011
I thought earth's radiation shield was primarily based on our magnetic field warping the particles around the planet instead of through it. So couldn't something like that be built, on a smaller scale? Though powering it would probably be a nightmare...
I'm clearly no expert on the happenings of Earth. I do seem to recall watching some show that talked about the magnetic fields deflecting radiation... I was under the assumption that the atmosphere was also a big part of the radiation not hitting the Earth. I could be wrong.
They actually have software programs that you run to figure out your radiation dosage per year. Like on this last program, in a geo orbit, with the amount of shielding we had (this thing was the size of a minivan), the electronics should expect to see roughly 30 kRADs of radiation per year. I know the amount of radiation we see on the ground is less then 1 kRAD per year. I actually have no idea off the top of my head how much radiation you would see in a geo orbit with no shielding. Much higher then 30 kRAD I am sure.
I remember reading once that the space shuttle and ISS only receive like 2 KRad of radiation inside. That is pretty crazy they could shield it that much. it surprised me when I heard that.
WakeOfAshesPosts: 21,665destroyer of motherfuckers
I thought earth's radiation shield was primarily based on our magnetic field warping the particles around the planet instead of through it. So couldn't something like that be built, on a smaller scale? Though powering it would probably be a nightmare...
oh and I havent heard of anything doing shit like that on a small scale. Typically they just throw a shit load of lead around the electronics and that takes care of a lot of the radiation.
Something else you have to think about.... Can't really use fans in space to cool down your electronics. I mean you could run them, but they wouldnt create a breeze. you pretty much gotta use heat sinks and push the heat to the outside of the craft.
That makes sense. I wonder if heat sinks work better in a vacuum than in a fluid. This shows my lack of physics knowledge, but I'm still curious.
Do you use like, copper heat pipes? Or else how do you "push" heat to the outside of the craft? How do you deal with preventing reverse conduction when you're in direct sun? I'm guessing cooling is easy on the dark side.
WakeOfAshesPosts: 21,665destroyer of motherfuckers
yeah all that stuff is outside of my area of expertise. All of those issues are why we hire Mechanical engineers. The ME's usually do the thermal modeling. Most of the time I tell them how many watts my parts will burn, and then they do modeling on that and devise some sort of heat sinks that will work. Or perhaps they might come back and tell me to cut my wattage down.... Which ties back to the original point of this thread of designing stuff that is low power
and yeah I am sure they deal with reverse conduction in direct sun, but I dont know how they do it. I'm too busy designing the engine to worry about the frame
Comments
[-(
blue turbins
From Those Fishes - I Fingered An Old Bitch (i got Aids on my finger)
The last satellite I designed was called 3GIRS. It is actually a big 4 megapixel 70 fps inferred video camera in a geo orbit. I designed about a third of the camera electronics, but was overall the main system integrator of all the camera electronics. The basic idea was to track ICBM's that happened to be coming our way. It is really just an engineering test at this point in time, and not a full scale development but it will be in orbit soon.
But yeah, in generally designing spaced based electronics offers a lot of interesting challenges. Like the electronics will go from behind the earth at -55 degrees C, to in direct view of the sun and 125 degrees C, almost instantly. And so the electronics not only need to survive that instant temp change, but also not hiccup. Not only that they are always being irradiated by massive amounts of sun radiation which in time will degrade the electronics and cause the thing to just stop working. No ways around that. There are interesting challenges, like you have to be careful not to use ROHS parts because they are mostly tin and tin will grow "whiskers" in space and has been know to brick your craft. Also the radiation will cause errors in your data, and so you need to do everything three times and then take the best 2 of 3 answer. Also... the parts you use are mega expensive. as an example... when I make earth based electronics, the fully populate board might cost 50 dollars to maybe 5 thousand dollars. A spaced based boards cost anywhere from 250k-1 million. It's pretty nerve racking programming a one time programming only computer chip that costs $100,000+. Try telling your boss you fucked that one up.
To prevent hard errors, they are very very picky about what silicon they use to make the parts. They shield the fuck out of the electronics. and they run the parts through a lot of radiation to make sure they wont just lock up and die. All of this is why space based electronics costs so much.
Most of the time, real processing work is being done by FPGA's and not processors. These are basically computer chips that can be programed to behave a specific way (I use the language VHDL or Verilog to define the logic functionality). Like you could take a Intel processor and program it to these FPGA's and then the FPGA would act like an Intel processor. The space based FPGA's are one time programmable devices and they actually natively perform the "best of 2". I only need to think in terms of there being 1 instance, and the tools automagically make 3 and automagically take the best of 3. I do understand it more then what I am saying here, but for the sake of my work, I dont really need to know it because it is done for me by the tools. This is why one of these FPGA's cost 100+ thousand dollars.
I think you are thinking in terms of software... and in general processors and software is not often used on orbit. It can be used, but you are limited to what you can do because so many errors can be injected into the system. I think when people use processors, all data packets use some sort of error correction. like convolution, or Turbo Product Codes.... Plus you shield the fuck out of it, and make sure it isnt doing anything that can't recover from radiation errors.
The science behind this stuff is pretty interesting... When you get into this stuff, you realize how funny it is that we take the earth ozone for granted.
[-(
blue turbins
From Those Fishes - I Fingered An Old Bitch (i got Aids on my finger)
They actually have software programs that you run to figure out your radiation dosage per year. Like on this last program, in a geo orbit, with the amount of shielding we had (this thing was the size of a minivan), the electronics should expect to see roughly 30 kRADs of radiation per year. I know the amount of radiation we see on the ground is less then 1 kRAD per year. I actually have no idea off the top of my head how much radiation you would see in a geo orbit with no shielding. Much higher then 30 kRAD I am sure.
I remember reading once that the space shuttle and ISS only receive like 2 KRad of radiation inside. That is pretty crazy they could shield it that much. it surprised me when I heard that.
Something else you have to think about.... Can't really use fans in space to cool down your electronics. I mean you could run them, but they wouldnt create a breeze. you pretty much gotta use heat sinks and push the heat to the outside of the craft.
Do you use like, copper heat pipes? Or else how do you "push" heat to the outside of the craft? How do you deal with preventing reverse conduction when you're in direct sun? I'm guessing cooling is easy on the dark side.
and yeah I am sure they deal with reverse conduction in direct sun, but I dont know how they do it. I'm too busy designing the engine to worry about the frame