Questions about Radio Systems

[hotmailcompany52]

Not quite. A RasPi isn't radiation-hardened - over time, cosmic rays striking the silicon chips will cause errors and damage, culminating in component failure.

But couldn't you shield it from radiation tho, but anyway if he isn't gonna use a Raspberry Pi he would still need to protect it from radiation because to my knowledge not a lot of off the shelf components are radiation-hardened.

[NGTOne]

I don't think of that as a problem. Once you're in LEO, you have all the time in the world to get to the orbit you need. What you really care about is ISP. So if there is a good commercial option for a high ISP ion drive, then it's doable.

But if it has to be a 3U, that's going to roughly tripple the launch cost. Then again, since 1U would probably be too small for this mission anyhow...

The problem is, the thrust might be too low to resist atmospheric drag, even in LEO. The electric thrusters they sell develop tiny fractions of a newton of thrust.

But couldn't you shield it from radiation tho, but anyway if he isn't gonna use a Raspberry Pi he would still need to protect it from radiation because to my knowledge not a lot of off the shelf components are radiation-hardened.

A lot of off-the-shelf components AREN'T rad-hardened, but I'd be very surprised if you couldn't find some. Besides, effective rad-shielding would weigh too much (a 3U CubeSat has a 4kg weight limit).

[Nibb31]

LEO To GEO requires 4000m/s. That's a lot of dV to pack into a 3U cubesat. I don't think it's even possible with an ion thruster.

You also have to include provisions for disposing of a GEO sat. They have to move to a higher orbit at end-of-life, which means even more dV.

[Bunsen]

But couldn't you shield it from radiation tho, but anyway if he isn't gonna use a Raspberry Pi he would still need to protect it from radiation because to my knowledge not a lot of off the shelf components are radiation-hardened.

Radiation shielding is way, way too heavy to be practical. Keep in mind that the comsats the size of small school buses still use rad-hardened electronics, because significant amounts of radiation make it clear through something that size.

For a short mission, it's possible to get away with non-rad-hard components through careful design. Radiation problems fall into two general categories: total ionizing dose effects (things that slowly and predictably get worse with more radiation exposure), and single event effects (things caused by individual particle impacts, you can only predict how often they're likely to happen). TID effects take a while to show up (months or years, usually) if you don't pick uncommonly vulnerable parts, and are a big part of what rad-hardening focuses on. SEEs are always a problem, and cause things like memory corruption, incorrect instructions, random resets, and occasional internal short-circuits that let the magic smoke out of the chips if you aren't careful. But SEEs can be mitigated with defensive circuit design and paranoid programming. This is the approach cubesats usually take, because they're built for short-term operation and don't have the budget for rad-hardened electronics (which are obscenely expensive).

Raspberry Pis don't have those circuit-level protective measures, and are kinda power hungry for what they do, but embedded computers in that ballpark are a decent choice for some small satellites. Sufficiently simple missions can run on something closer to Arduino-scale hardware.

[NGTOne]

LEO To GEO requires 4000m/s. That's a lot of dV to pack into a 3U cubesat. I don't think it's even possible with an ion thruster.

You also have to include provisions for disposing of a GEO sat. They have to move to a higher orbit at end-of-life, which means even more dV.

I actually did the math - with one of the commercially available electric thrusters I looked at, and about 1.5kg of fuel, you could get ~5km/s of dV. And disposal from GEO is actually pretty easy - boosting it to a higher "graveyard" orbit takes less than 30 m/s.

[Crush]

Could someone explain the process to me technically, from start to finish (I mean like from eletrical input to antenna efficiency)?

A good website for questions like that is http://electronics.stackexchange.com/ but you better get more specific, because when you ask it like that it will likely be closed as too broad.

(Which I will not say to protect my uncopyrighted intellectual property)

Ideas are worthless. Everyone has great ideas, but most people lack the resources, skills and/or determination to execute them. But it's the execution which has value.

A visionary is not someone who has a great idea, a visionary is someone who had the balls to execute it and the ability to execute it well.

Edited November 9, 2013 by Crush

[Kryten]

I actually did the math - with one of the commercially available electric thrusters I looked at, and about 1.5kg of fuel, you could get ~5km/s of dV. And disposal from GEO is actually pretty easy - boosting it to a higher "graveyard" orbit takes less than 30 m/s.

But with what kind of power requirements? The power available to cubesats tends to be very low.

[NGTOne]

But with what kind of power requirements? The power available to cubesats tends to be very low.

The power requirement was fairly modest - about 20W, I think.

[Kryten]

Cubesat power budgets currently top out at about 8 watts.

[Bunsen]

Pulsing an electric thruster isn't hard. The size is a bigger problem -- the efficient electric thrusters I've seen aren't so much designed to be a supporting component of a cubesat with a payload as they are technology demonstrations just barely squeezed into a 3U with no room for much of anything else. The ones I've seen that are small enough to coexist with a respectable payload have terrible propellant fractions and correspondingly terrible system Isp.

[hotmailcompany52]

What about some highly pressurised gas and few nozzles?

Edit: You would need about 1 KG tank with a psi high enough for about 107.4m/s escape velocity. This will provide about 30.9 delta-v.

Edited November 9, 2013 by hotmailcompany52

[Idobox]

Building radios for space is difficult.

As other people said, you will need to work at least in UHF, and probably will require K band or higher. At these frequencies, everything works differently: you can't use wires, the thickness and width of strips is important, etc...

Also, with the large temperature differences, you will get some expansion and contraction, and that will detune everything. You will need to use specific substrate like duroid or alumina, rad-hardened transistors or chips, etc...

I've designed antennas professionally for 4 years, and just building one for this project would have taken me a few weeks, and I used to have access to professional tools (an network analyzer that works in K band costs tens of thousands of dollars)

[K^2]

What about some highly pressurised gas and few nozzles?

Edit: You would need about 1 KG tank with a psi high enough for about 107.4m/s escape velocity. This will provide about 30.9 delta-v.

This is fine for maneuvering, but you aren't going to do any significant orbit changes with that. You need about 500m/s just to transfer to Molniya assuming you had correct inclination to begin with. If you are shooting for GEO, the budget is closer to 2km/s from LEO.

Keeping in mind that 1U is capped at 1.33kg, and the maximum you can have is a 3U at 4kg, your options are very limited. The only viable way to do significant orbit changes with a cube sat is by using ion drives. These are energy-limited, and with a solar cell you can get as much as you need over time.

So yes, if you need to collect data near GEO and transmit it to a base on Earth, by far your best bet is a 3U with an ion drive in Molniya orbit.

If I was contracted to build this, I'd be quoting something in the $300k - $400k range, depending on some specifics. This is $100k for launch, $100k-$200k parts and anything I have to special-order, and $100k commission for the work. Later includes nav and comm software, testing, etc.

If anyone's interested, send me a PM with details and I'll get you a more precise quote. There is no way OP is building it for less.

[Kryten]

Is he likely to actually find a launch opportunity to molniya orbit, though? I'm pretty sure most or all satellites put into those orbits were ultimately replaced by GEO sats, and after a quick search I can't find any recent launches to it.

[K^2]

Is he likely to actually find a launch opportunity to molniya orbit, though? I'm pretty sure most or all satellites put into those orbits were ultimately replaced by GEO sats, and after a quick search I can't find any recent launches to it.

That's kind of my point. You won't. What you can find is a high inclination launch to LEO. Say, something en-route to ISS. The ion drive will have to push the rest of the way. You'll have to do 12° of inclination change, requiring about 500m/s of dV, followed by a 2.3km/s apogee push. Because you are already in stable LEO, you can do it over many revolutions, so low thrust isn't a problem. And 3km/s of dV on a small ion drive is entirely doable. You can get an electrospray thruster with I_SP of up to 1,300s. So you'd only need 26% of your sat's mass, or just over a kg in propellant for this. Another kg of mass is taken up by the engine, leaving you with 2g of payload. Entirely reasonable. Thrust is a bit more problematic. At 9W, you'll get only 1mN of thrust. Hauling the 4kg, it will take over four months of thrusting to reach the target orbit. Realistically, way longer because you have to time the puffs.

But who cares, right? This isn't enough for orbit to decay significantly, and there is enough overhead in above to make corrections. And it will be the computer doing the pushing, so it's just launch and forget until it's time to collect the data.

There are certainly faster ways to get this done, but not cheaper ones.

[NGTOne]

This is fine for maneuvering, but you aren't going to do any significant orbit changes with that. You need about 500m/s just to transfer to Molniya assuming you had correct inclination to begin with. If you are shooting for GEO, the budget is closer to 2km/s from LEO.

Keeping in mind that 1U is capped at 1.33kg, and the maximum you can have is a 3U at 4kg, your options are very limited. The only viable way to do significant orbit changes with a cube sat is by using ion drives. These are energy-limited, and with a solar cell you can get as much as you need over time.

So yes, if you need to collect data near GEO and transmit it to a base on Earth, by far your best bet is a 3U with an ion drive in Molniya orbit.

If I was contracted to build this, I'd be quoting something in the $300k - $400k range, depending on some specifics. This is $100k for launch, $100k-$200k parts and anything I have to special-order, and $100k commission for the work. Later includes nav and comm software, testing, etc.

If anyone's interested, send me a PM with details and I'll get you a more precise quote. There is no way OP is building it for less.

Your values are off, rather significantly. Most of the parts, even including the thruster and fuel, can be obtained for under $50K (in total). Launch costs for CubeSats, depending on exactly what else is going up and where it's going, range between $20-30K. I'm assuming that whoever is attempting to launch the thing has all the skills required to put the pieces together, and is willing to invest the time to do so. But I've seen (admittedly simplistic) CubeSat missions for as low as $50K. There was a successful Kickstarter a while back - dude launched a CubeSat that did nothing but tweet canned messages from contributors.

The specifics may vary, but it's reasonably simple to design a 3U CubeSat mission for under $100K. My theoretical mission was for a lunar flyby from LEO (entirely possible, though the payload capacity shrinks rather significantly), and my estimated budget (given the research I did) came to about $130K, and that was with SIGNIFICANT padding.

What I'm interested to see is the 6U (2x3) CubeSat standard I've heard about. That has some REAL potential (CubeSat Mars flyby, anyone?)

Edited November 12, 2013 by NGTOne

[Bunsen]

Your values are off, rather significantly. Most of the parts, even including the thruster and fuel, can be obtained for under $50K (in total). Launch costs for CubeSats, depending on exactly what else is going up and where it's going, range between $20-30K. I'm assuming that whoever is attempting to launch the thing has all the skills required to put the pieces together, and is willing to invest the time to do so. But I've seen (admittedly simplistic) CubeSat missions for as low as $50K. There was a successful Kickstarter a while back - dude launched a CubeSat that did nothing but tweet canned messages from contributors.

The specifics may vary, but it's reasonably simple to design a 3U CubeSat mission for under $100K. My theoretical mission was for a lunar flyby from LEO (entirely possible, though the payload capacity shrinks rather significantly), and my estimated budget (given the research I did) came to about $130K, and that was with SIGNIFICANT padding.

What I'm interested to see is the 6U (2x3) CubeSat standard I've heard about. That has some REAL potential (CubeSat Mars flyby, anyone?)

Got a source on those launch costs? I've taken a look around, and the only prices I can find explicitly stated are well over $100k for a 3U launch (these guys, for example, charge $325k for a 3U P-POD, integration, and launch, and that seems in line with other vague figures I've seen).

[kiwi1960]

Now I remember why I gave up my ZL2 call sign... ha, those were the days...

[dharak1]

What I'm interested to see is the 6U (2x3) CubeSat standard I've heard about. That has some REAL potential (CubeSat Mars flyby, anyone?)

If you go to the Clyde space website they have a 12U structure category but barely anything for it, not even structures. I was looking at Structures on another webstite and they say that a few 6U cubesats are prepping for launch.

Here is the 6U Website: http://www.cubesatshop.com/index.php?page=shop.product_details&flypage=flypage.tpl&product_id=45&category_id=1&option=com_virtuemart&Itemid=66&vmcchk=1&Itemid=66

And the clyde Space product list: http://www.clyde-space.com/cubesat_shop

[K^2]

Your values are off, rather significantly. Most of the parts, even including the thruster and fuel, can be obtained for under $50K (in total).

Transmitters, receivers, and testing equipment will bring it up to $100k easily. The rest depends on exactly what it is meant to do up there.

Launch costs for CubeSats, depending on exactly what else is going up and where it's going, range between $20-30K.

That's per unit. 3U is 3 units, hence the name. It will cost around $100k for launch, perhaps rounded up a bit. Keep in mind that suitable orbit is also important, so it's going to narrow down the market.

The specifics may vary, but it's reasonably simple to design a 3U CubeSat mission for under $100K. My theoretical mission was for a lunar flyby from LEO (entirely possible, though the payload capacity shrinks rather significantly), and my estimated budget (given the research I did) came to about $130K, and that was with SIGNIFICANT padding.

And you were planning to do telemetry how? You need to hit an object over 300,000km away with a satellite that needs more than a year to reach the necessary velocity, during which its orbit is going to become more and more elongated. There is simply no way in hell you are placing the apogee where you need it without correction in transit, and you aren't going to make corrections in transit without telemetry. You could, normally, make final adjustments using a camera, recognizing stars, and watching when they become blocked by the Moon, assuming you can actually write the software that does all that, but unfortunately, with thrusters you can actually get for this sort of dV, you won't have enough thrust to make a correction that late in flight.

I suggest you take a look at history of unmanned lunar fly-by missions and see how complex they were, and how much development they took. Your estimates are naive. And that's not even mentioning the fact that you took a launch estimate for the wrong size sat.