Sending a rocket into space.

[made2hack]

Hi all,

Wanted to share this video from Copenhagen Suborbitals on the design of their onboard computer using an Atmel microcontroller and other design considerations.

[HAL-42b]

Sourcing components that won't outgas in space must be very difficult.

[HAL-42b]

Sourcing components that won't outgas in space must be very difficult.

Why that matters?

It may change the value of the component or break some of them completely.

[calexanian]

One solution is to pre outgas the parts and then encapsulate the circuits.

[mackrc1]

something doesn't feel right about strapping an arduino to a rocket.

[made2hack]

something doesn't feel right about strapping an arduino to a rocket.

well, to be pedantic, they are not actually strapping the whole arduino. Just using the Atmel chip with their own custom Arduino layout / pcb.

[cdev]

Satellite interiors make extensive use of gold foil and I think one of the reasons is for shielding.

[cdev]

Cats in space

[mackrc1]

I feel if you were going out your way to put a rocket into space and design your own flight computer you wouldn't then go and put the arduino boot loader on it.

[HAL-42b]

Definitely not. I doubt the atmel chip is even involved in the realtime loop.

[m98]

Definitely not. I doubt the atmel chip is even involved in the realtime loop.

Why? I don't see any reason to doubt the reliability of any reasonable uC for such a simple application.

[HAL-42b]

Why? I don't see any reason to doubt the reliability of any reasonable uC for such a simple application.

Engine control + attitude control + sensors + telemetry all in realtime doesn't seem all that simple to me.

[cdev]

It would be interesting to compare the Apollo onboard computer to an Arduino.. seriously, how do they compare?

maybe we're being elitist here when the facts - I wouldn't be surprised if the Apollo missions used basically something with the power of an Arduino. I'm pretty sure that was the best they had in 1969 that could fit on a rocket within the weight limitations.

Also, remember HAL, in 1969, when the film 2001 was made, they visualized HAL as basically filling an entire room. (Still, I think they got a lot right in that film, they really did their research as to what the next few decades would bring well)

[HAL-42b]

Gyroscopes as big as basketballs and accelerometers consisting of a hunk of metal on a spring and a linear pot. And yet it was binary digital, it was not analog and it was not even binary coded decimal.

I admit I have a soft spot for fully analog control loops.

[jt]

Damn, a thread finally within my expertise - but I never know how much I can say because of the ITAR BS!   

[Fungus]

Definitely not. I doubt the atmel chip is even involved in the realtime loop.

Why? I don't see any reason to doubt the reliability of any reasonable uC for such a simple application.

Engine control + attitude control + sensors + telemetry all in realtime doesn't seem all that simple to me.

That's a software problem, not a hardware problem. If the chip can run the software then what's the problem if it's the same chip used in Arduinos or not?

Those AVR chips can do a lot if you program them right (some of the Arduino libraries are awfully inefficient...)

[Rerouter]

Engine control is On / off for most systems, variable thrust is rare to see apart from larger systems, equally when in free fall, the choice of propellant also gets interesting, as it may not be at the nozzle but rather a blob in the middle of the tank, possibly using the reaction wheels before any maneuvers to shift it to the nozzle,

Attitude control is Up at an angle, with some fine control to get the inclination and pre-circularisation apoapse in the right place, quad copter controllers deal with harder tasks than that with an ATmega 328, heck at the velocities involved, it wouldnt be that hard to predict the orbit off gps points alone on the initial ascent,

Now telemetry is what you do when all the other bits are idle, even in predicting its final orbit, it will have slabs of time free, as the calculations are well known,

[made2hack]

Damn, a thread finally within my expertise - but I never know how much I can say because of the ITAR BS!   

Please tell us as much as you can. We won't report you!

[calexanian]

I am surprised that more of the components did not unsolder themselves when directly exposed to that much heat and flame.

[jt]

Please tell us as much as you can. We won't report you!

Well links shouldn't hurt.  For further reading, other companies/activities using commercial-of-the-shelf components (including Atmel MCUs) for LEO space operations include:

Spacecraft and spacecraft hardware vendors:
Pumpkin:  www.cubesatkit.com
Tyvak /Terran Orbital: www.terranorbital.com  &  www.tyvak.com
Blue Canyon:  www.bluecanyontech.com/portfolio-posts/cubesatsystems/
Clyde Space:  www.clyde-space.com/
Berlin Space:  www.berlin-space-tech.com
Sinclair: http://www.sinclairinterplanetary.com/

A few companies with business plans based on this class of hardware:
Visible earth observation: www.planet.com/
Weather science:  www.spire.com/
Lightsail propulsion:  http://sail.planetary.org/ (not actually a business plan) 

Also more nano-launch vehicles banking on cheaper commercial electronics based avionics for more economical solutions:
Garvey Spacecraft:  www.garvspace.com/
Generation Orbit:  www.generationorbit.com/

Related University programs:
Uof Michigan: http://umcubed.org/
Cal Poly SLO: http://polysat.calpoly.edu/
Montana State:  https://ssel.montana.edu/

Related Archived papers & presentations: 
CubeSat Workshop:  http://dww.cubesat.org/index.php/workshops/past-workshops
Small Sat Conference: http://digitalcommons.usu.edu/smallsat/

[calexanian]

How about in helium balloon satellites at just the edge of space.  Come up with a gas ballasting system to keep them at a stable height. Just a sleep deprived thought.

[Rerouter]

seems most balloons only get you to an altitude of about 40km max. at which point, the atmosphere is less than 3% of ground level,

Ignoring drag, and assuming you magically have enough thrust to weight to accelerate with minimal gravity losses, your looking at about 7600m/s from a balloon, stealing the planets rotational speed, and a little bump from the wind going in the right direction,

If you can pull that off, while making it light enough to float up that high on a balloon, then you can get to space, barely. compared to the norm of about 10,000m/s for a marginal ground landing (larger rockets have lower drag losses)

[Howardlong]

I am part of a team who've put up several satellites, what do you need to know?

I'll warn you now though that you will find that if you ask three different rocket scientists the same question you'll get three or more different answers though.

In short in LEO if you're prepared to accept the odd glitch using COTS parts can be a reasonable way ahead. Most outgasing will be caught in the thermal vacuum testing anyway, but also it's common to conformally coat boards anyway, which also maitains rigidity for launch (also tested in pre launch shake tests).

"The odd glitch" in LEO is usually caught by watchdog timers and the effects usually fairly benign. We see resets once every few months. Our designs also use dual active devices, an MCU for the more computationally complex day to day stuff and a CPLD for a very basic mode where we can still command the spacecraft even if the MCU goes titsup. Using two completely different designs complementing each other also removes common mode failure.

As well as one off events ("single event effects") over time there are cumulative effects ("toal ionising dose") which slowly degrades the insulating gate area of MOSFETs, and eventually will render a transistor inoperative. There are many classes of one-off SEEs though. The worst is a single event latchup leading to single event burnout which is catastrophic, where the totem pole transistor pair turn on together causing a short across the supply. This is often unrecoverable. You can mitigate it by introducing current limiting power down in you power supplies but this adds to complexity and it may be too late anyway as the amount of charge in you decoupling caps may have already sealed the fate.

One of our team also sent up Strand-1 a few years ago, this was an off the shelf Android phone.

So why use COTS parts rather than so-called "space rated" parts? There are several reasons. Cost is the obvious one. Off the shelf parts tend to be far better characterised than limited volume parts. In a lot of cases, it is a valid business decision to have a 90% success rate for 1/4 the cost of a 99% success rate. After all you could send up four times as many and have a better aggregated success rate that way.

My area of specialisation is in the comms segment including telemetry, control and transponders, and the RF segments in general. Among other things currently I'm working on the embedded hardware and software for a digital transponder for use with a geostationary satellite scheduled for launch late this year.

I'm not in the US so ITAR and EAR have little direct effect on me, but I have to be aware of them. ITAR and EAR keep many non-US citizens in jobs as a result!

[Howardlong]

Q. Why is LEO much more benign than HEO?

A. The Van Allen belts (there are two) protect LEOs from most of the harmful ionising radiation.

There are modelling programs like creme96 that will model radiation on a given orbit.

[made2hack]

I am part of a team who've put up several satellites, what do you need to know?

Well, I do have a question unrelated to sending something into space. I want to, in the future, time permitting, send up a weather baloon or other type of "craft" that can circumnavigate the earth and send out a Hello World message (or something to that effect) say every hour / minute / second or so.

1- So, where do I begin? What I mean is, what would you concieve as a craft? Would a weather baloon work? Will it just simply climb until it pops? Or can I limit it say to 15km or so, just above commercial air ceiling?

2 - How can I communicate with it? Either active or passive? Radio? GPS? Voodoo? Is it possible at that distance to have something affordable?

3 - How do I know in what direction it will take? Does it just follow the wind? Are there air currents that will make sure the baloon goes around the world?

I know my question is vague but perhaps you can point me in the direction of where to start my searches.