KSP Community CubeSat

[Mazon Del]

It's Mazon Del who has access to the moss samples. So I forward the question to him. Can you do it? Like an IR camera filming a small sample of the moss (doesn't need to be fancy at this point) for a few days with the most FPS (one frame picture per hour?) So we can then see if that's overkill and, if it is, speed it up through post-editing to see what's the sweetspot of least FPS with good video for analysis.

Technically I could. I don't have an IR camera (though really you'd just need any camera because all we are doing is looking at growth, not any changes in wavelength). Mostly though I'm sure I can just ask Luis what is a good rate.

[MBobrik]

That's unorthodox indeed. What does [orthogonal circular coils] even mean? I'm a visual kind of guy, so the lack of 3D models is making this thinking quite a challenge of imagination

At the end of my post from the previous page is a picture of two possible shapes

[stopping the spin, transmitting, and spinning back up again] sounds awfully complicated to me, and is it really necessary?

We have first to aim the panels to the sun to get maximum power to charge the batteries. Then,, we have to aim the antenna towards receiver on the ground to get reasonable bps. And then back to the sun.

[50 RPM/min] Really? I think that's really overkill.

It's now just 44 RPM/min. And that's only when earth's magnetic field direction and strength is optimal. In reality, we will get maybe 1/3 of it most of the time.

[Magnetotorquers outside of the cube]

no longer necessary. the 44 RPM/min figure is with torquer coils directly around the craft.

[LordQ]

A lot of discussion since I went to bed last night, great to see!

@K^2: Yep, understanding precession isn't the problem, the problem is that the coning manoeuvre analysis that I attempted last night just isn't suitable for analysing how fast we can precess our cubesat. This is because the coning manoeuvre analysis assumes instantaneous impulses, like you might see with monopropellant thrusters, and a magnetorquer, at least the one I analysed, is not powerful enough to be able to provide "instantaneous" impulses. Similar to how it's a terrible idea to analyse an ion probe using standard Hohmann transfer analysis.

@MBobrik: Nice work again! I don't have much time right now, but your numbers certainly look better now. When I get home, I'll be sure to go through it. It might also be worth investigating whether the magnetic fields induced by our magnetorquers get close to or exceed magnetic field limits for cubesat hardware. Probably a bad example, but you wouldn't hold a magnet up to your phone.

Edit: So on the train now, so I've had some time to look through your calcs at least a little MBobrik. I noticed that you're putting in an input power of 10 watts, which is a lot of power. In comparison, the coil I found onlineonly uses 200mW, 50 times less power. With an input power of 200mW, your coil would produce less than 1 RPM/min of acceleration. I'm not 100% sure if just changing the power like that us valid however, and I will have to check properly once I get home.

Edited December 4, 2014 by LordQ

[henryrasia]

Endersmen, I forgot to reply to your post! I personally think this thread is getting too small to house our multiple discussions, especially with new people joining in without context. That website you were building would fit us like a glove, especially that mini-forum section! Are you still working on that?

And MBobrik, ok fair enough. I still think stopping transmitting and spinning back up is exceedingly wasteful of time, electricity, and maneuvering. We should really prioritize having all systems passive as much as possible. That's why I suggested having all nonspinning components along the rotation axis in a structure that remains stationary in relation to the sun (solar panels) and a point on the ground (ground station). It sounds complicated, but I think solving that problem is MUCH better than solving the stop-restart one. If anything else the latter would mess with the gravity experienced by the moss, potentially ruining the usefulness of the whole experiment. Mazon Del could correct me on this, of course.

[MBobrik]

So on the train now, so I've had some time to look through your calcs at least a little MBobrik. I noticed that you're putting in an input power of 10 watts, which is a lot of power. In comparison, the coil I found onlineonly uses 200mW, 50 times less power. With an input power of 200mW, your coil would produce less than 1 RPM/min of acceleration. I'm not 100% sure if just changing the power like that us valid however, and I will have to check properly once I get home.

well, it's √(P). so reducing power 50 times will reduce acceleration to 6 RPM/min.

I still think stopping transmitting and spinning back up is exceedingly wasteful of time, electricity, and maneuvering. We should really prioritize having all systems passive as much as possible. That's why I suggested having all nonspinning components along the rotation axis in a structure that remains stationary in relation to the sun (solar panels) and a point on the ground (ground station).

Sure, reorienting the whole craft costs power. But having antenna that can turn and track the ground as the satellite is spinning, could be mechanically very complex.

[LvL]

(Just jumping in here) could we do some experiments with sound in a later cubesat? Sound travels at a different speed in vacuum, right? maybe we could study that... no idea how to though. would it be worth it? is it even possible to do it in a cubesat? (feel free to prove that it can't be done )

[K^2]

Sound travels at a different speed in vacuum, right?

Sound doesn't travel in vacuum. LEO isn't perfect vacuum, but satellite orbits are far above the altitude at which any discussion of sound is sensible on relevant scales. For all intents and purposes, molecules/atoms up there are ballistic projectiles.

[LvL]

Sooooooo, I then guessing that we can't measure the speed of light? (half joke, I mostly have no idea what I'm talking about)

[K^2]

One of the things I'd like to have on board of the sat is at least one optical gyro for spin rate/orientation control. So it will be measuring speed of light about a hundred times a second. Or rather, it will measure a quantity that depends on spin rate and the speed of light. But if we grab accelerometer data to establish spin rate independently, it does give us a speed of light measurement. Problem is, it'd be imprecise measurement due to accelerometer measurements being of limited quality. Which is the reason for having optical gyros in the first place. By assuming the known value for speed of light, we can get very precise rate of spin measurement instead.

Just thought you might find it interesting.

[henryrasia]

K^2 and Mazon Del, what do you guys think about the discussion on stopping-restarting vs independent axis? MBobrik thinks the latter would be too mechanically complicated, but I think the former would be too electricity-expensive and risky (if that fails the whole mission is a goner). But more than that I would think the spin rate should be constant for the experiment, as decreasing and increasing would interfere severely with the results. The independent axis could be maintained by two small and weak magnet holder contraptions* that could serve as their own mini pseudo magnetotorquers (for the rod, not the whole thing, see the asterisk). The only problem with this is that the Sat should be spinning very stable at all times, which is a non-problem since it should be regardless (to keep the experiment ok).

*The sat itself does not need to be pressurized, only the moss samples. If there's two holes along the spin axis through which passes a rod (to which solar panels, transmitters and what have you are attached ) the connecttion would be like this: two disks on the rod, one in each side of the sat wall. Magnets in these disks and in the sat wall would keep them in place. This construction would be mirrored on the other hole for balance. Then needed, extra electromagnets would be activated to decrease/ increase rotation OF THE ROD. This of course would affect the sat, but since the sat is more massive its own magnetotorquers would easily make up the difference. And since this would be a passive system (only being used to correct small imperfections), both the power usage and risk of failure would be low. Also, the contraptions can be very small and light as we aren't expecting big forces (remember the stability requirement?). Is this description clear or do we really need 3D animation? Because I'd really love if we did, text is hard...

[MBobrik]

K^2 and Mazon Del, what do you guys think about the discussion on stopping-restarting vs independent axis? MBobrik thinks the latter would be too mechanically complicated, but I think the former would be too electricity-expensive and risky (if that fails the whole mission is a goner). But more than that I would think the spin rate should be constant for the experiment, as decreasing and increasing would interfere severely with the results. The independent axis could be maintained by two small and weak magnet holder contraptions* that could serve as their own mini pseudo magnetotorquers (for the rod, not the whole thing, see the asterisk). The only problem with this is that the Sat should be spinning very stable at all times, which is a non-problem since it should be regardless (to keep the experiment ok).

There won't be any spinning up and down. the torquer would continuously change sat's direction of rotation without changing RPM. And there is no need to actually run it at 10 Watt. We can say 22 RPM/min and 2.5 W is fine, or 10 RPM/min and 0.6 W. At 0.2 W it still got 6 RPM/min, which means it is about 3x as efficient and weights 2.3 times less than the one LordQ found on the internet. It is just, that we can make it change direction that fast if we need to.

If we had the antenna stick out on a rod, and aim for the ground while the satellite is turning, it would involve one (small) motor that runs nonstop to compensate for friction, and another servo to change antenna's pitch. And the whole contraption would have to be inside of the satellite, or tightly folded, during the orbit trip, and then unfurl after release.

There is of course a third option. We won't aim the satellite's antenna at all, and use omnidirectional ( or go with the 4x switched dipole with measly 1.5 dBi gain ), and make it up by using a good tracking phased array antenna, and a high quality receiver on the ground. Might be cheaper if we can weld the antenna by ourselves ( over at cubesat shop costs a ready-made tracking station $118 000 ).

Edited December 5, 2014 by MBobrik

[Mazon Del]

Personally my recommendation is to go with an Omni directional system on the cubesat. It involves the fewest operations for the sat to undergo (just needing to make sure the antenna is within visible range of Earth).

If we REALLY wanted to get fancy and had the budget for it, it might be possible to come up with a phased array using Omni-directional antennas that we electronically steer towards the Earth. This requires a nice clock to do and will increase costs, but if we have to care about directionality this is MUCH more preferable than anything that requires physical steering. Electronic steering doesn't change any physical parts.

[K^2]

Both are bad. Spin rate must stay constant for experiment. Mechanical rotation is problematic in space. Consider, e.g., problem of bearing lubrication kn vacuum.

Any ways, the only reasonable way to reorient the axis is precession.

Phase array is only viable in microwave, and even then, very limited with 10cm. UHF array is practically useless at this scale.

I would also lean towards omni, and consider precession as plan b.

Edited December 5, 2014 by K^2

[Nicholander]

Is this description clear or do we really need 3D animation? Because I'd really love if we did, text is hard...

I guess we kinda do... I can't really understand that.

[MBobrik]

Phase array is only viable in microwave, and even then, very limited with 10cm. UHF array is practically useless at this scale.

What I meant was phased array on the ground to make up for the low gain of an omnidirectional antenna on the sat.

[Newt]

Also, for the antenna, we have to find a frequency that works with all the countries that we will be receiving in, and the licenses of the station operators therefore. You cannot plan to transmit commands in a band that is illegal to use (well, maybe), no matter how efficient it is, and doing this is several different countries complicates matters.

Additionally, I agree with concern about a directional antenna, already needed are rotation for the experiment, it sounds like, and proper orientation to receive sunlight. Trying to track both the sun and a ground station, while spinning no less, is going to be a challenge. Easier just to align the sun to the axis of rotation,and have the solar panels there, with the antenna sticking out somewhere (or several antennae?). Of course, that takes a lot more power, on both ends.

Oh, and, hi! I would be interested in contributing.

Edited December 5, 2014 by Newt

[Mazon Del]

K^2, as far as inside the experiment is concerned unless we load up on cameras or minimize the moss in question, we are going to have to have some physical movement of parts in there. I have to say that I like the idea of the ring of moss capsules that rotates about the center axis of the sat with fixed cameras (one shining straight down to capture the growth of horizontal-growth bred moss, and one from a perpendicular angle for up/down-growth bred moss).

Do you think this is an acceptable amount of mechanical movement for the sat?

Incidentally I was thinking that what we might want to do is provide each capsule with a small bit of text or even a QR code that would be in the frame of the shot. This would allow for identification of a given sample with more certainty than the internal position sensing. Imagine for whatever reason there was drift. Worst case we wasted some ink and an infantismally small amount of our mass. Best case, our images require less effort to sync up with each other (I'm sure you'd be able to look closely and figure out which was which by hand if you had to).

Additionally, a thought that occurred to me. While we cannot replicate it on Earth, at least not easily, would we want to have a Geiger counter or other radiation sensor on board? I am uncertain about sizes, masses, or power requirements for this.

Welcome Newt!

Edited December 5, 2014 by Mazon Del

[LordQ]

My investigation of your results is ongoing MBobrik, but I've found something odd. I'm not sure how I missed this before, but I think you may have converted from rad/s^2 to RPM/min incorrectly. Your a = 0.012rad/s^2 converts to 6.88RPM/min using the conversion 3600*("input")/(2*pi), instead of the much higher value you calculated. I also found this website, which correctly verifies the conversion method I used.

[artwhaley]

Crazy comm idea... not worth doing... but amusing enough I thought I'd share it.

If you could be precise enough... you could split the sat in two, very gently, once you've deployed but before you begin spinning. The main body could spin and take measurements... while relaying data via low power omni radio to the other half that includes the high power directional radio.

Hoping that the spin inducing thrusters don't slam the two parts together... or spiral them out of short range comm range would be... iffy. Not to mention that you now need two solar arrays and two batteries.... though the overall power budget is almost the same (other than the short range radios)... so you don't really need MORE of either... just split up, and twice as many potential points of failure.

Definitely not a sane and practical solution. But maybe it will spark something that IS practical with somebody.

[K^2]

What I meant was phased array on the ground to make up for the low gain of an omnidirectional antenna on the sat.

Ah. Yeah, that might actually be cheaper and easier to make than a mechanical tracking station. Definitely worth considering.

Do you think this is an acceptable amount of mechanical movement for the sat?

Very high chances of it seizing up. I'd try to come up with something that's either more reliable or has redundancy.

[LordQ]

Alright, so my investigations of MBobrik's calcs have come to a sort of end. Unfortunately, I wasn't able to get the same numbers as you did MBobrik, but that's probably because I was putting in wrong numbers. Could I ask you to put all your calculations in a google spreadsheet?

In the end, what I did was calculate the magnetic moment of MBobrik's system for an input power of 0.2W (200mW, the same input power as the magnetorquer coil I found earlier). Interestingly, I found that MBobrik's system only has a magnetic moment of 0.09Nm/T at this input power. In comparison, the magnetorquer coil I found gave a magnetic moment of 0.2Nm/T, and thus has a much higher level of performance for the given level of input power. Of course, that magnetorquer would probably catch fire or melt if you tried giving it 10W of power. Which highlights the fact that the theoretical coil that MBobrik has described doesn't have any actual physical parameters attached to it, particular wire cross sectional area or number of turns, so we don't actually know if it would be able to withstand 10W of power.

So I took another approach, and am currently experimenting with the basic t = B * I * N * Sc formula with the R = ÃÆ’ * N * Cc / A and P = I^2 * R formulas, with inputs including the wire cross sectional area and number of turns that weren't present in MBobrik's analysis. I'll see what I can get.

[Endersmens]

Guys? Is anyone gonna tel me what I've missed?

[MBobrik]

My investigation of your results is ongoing MBobrik, but I've found something odd. I'm not sure how I missed this before, but I think you may have converted from rad/s^2 to RPM/min incorrectly. Your a = 0.012rad/s^2 converts to 6.88RPM/min using the conversion 3600*("input")/(2*pi), instead of the much higher value you calculated. I also found this website, which correctly verifies the conversion method I used.

Well, I suspected initially that there is a mistake there ... lost a 1/2pi factor at the very end . True, the real value is 6.88 RPM/min.

In the end, what I did was calculate the magnetic moment of MBobrik's system for an input power of 0.2W (200mW, the same input power as the magnetorquer coil I found earlier). Interestingly, I found that MBobrik's system only has a magnetic moment of 0.09Nm/T at this input power. In comparison, the magnetorquer coil I found gave a magnetic moment of 0.2Nm/T, and thus has a much higher level of performance for the given level of input power. Of course, that magnetorquer would probably catch fire or melt if you tried giving it 10W of power. Which highlights the fact that the theoretical coil that MBobrik has described doesn't have any actual physical parameters attached to it, particular wire cross sectional area or number of turns, so we don't actually know if it would be able to withstand 10W of power.

Well, the difference is, that the calculated one is already all 3 axes and weights 13 g and the one on the internet weights 50 g and it is one axis. If I scaled up to 3 x 50 g it would end up with 0.32 Nm/T

... getting the sheet ready

Edited December 6, 2014 by MBobrik

[MBobrik]

Torquer calculation sheet ... enjoy.

.

Sheet is set to the exact input values as the commercial torquer of LordQ just for comparison, and it still gets better numbers by cca 50 %. Feel free to change them to see the effects.

[Nicholander]

Guys? Is anyone gonna tel me what I've missed?

I'm not good at explaining things, but I'll try:

We're currently trying to figure out how fast the magnetotorquer we will be using to spin up our CubeSat can spin us up, and we think it can spin us up to 44 RPM in 1 minute.

Sorry if you already know this, but we have a Google Docs thing where we put all our ideas.

I think that's all I can tell you, like I said, I'm not good at explaining things.