View Full Version : [queued] Teensy has joined the space program

10-09-2017, 01:23 AM
Teensy 3.2 is replacing a grossly under-powered PIC processor in the camera simulator used to validate the commands we plan on sending to the cameras on the OSIRIS-REx spacecraft.
Pictured is a solid model of the simulator board (PCBs are in fab), and the current breadboard I'm using for testing the ported code.


10-09-2017, 05:10 PM
Sounds out of this world... (ok it sounded better in my head :P)

Do you have to worry about any kind of radiation? Whats the expected mission lifetime?

10-09-2017, 05:13 PM
This is running a simulator that will stay planted firmly on the ground. I think the odds of this processor surviving more than a week in space are vanishingly small :)

10-09-2017, 05:47 PM
That's awesome!
Ishkabbible; do you mind me asking who you work for and how you got into working with OSIRIS-REx? Dream job stuff right there!

10-09-2017, 06:13 PM
I am an EE at the University of Arizona Lunar and Planetary Lab - We have a long history of providing instruments for NASA missions, going back to Apollo. We have more recently moved into doing entire missions - from concept through space operations and science data analysis. The Mars Phoenix lander was entirely run by the UofA, as is OSIRIS-REx.
How I got into it is a long story going back to the 70s - in a nutshell, I got very lucky and more or less stumbled into it.
Not sure about the "dream job" aspect. Mostly LONG periods of boredom, writing proposals, and hoping there is enough money to keep me employed for another year, punctuated by moments of panic when we actually get a contract and have 3 years to turn our brilliant ideas into hardware on a completely unrealistic budget and timeline. But we somehow manage :D

10-09-2017, 06:31 PM
That's really interesting! Thanks for sharing!

10-10-2017, 12:53 AM
Just check the website for OSIRIS since I never heard of that one before. In case anyone is wondering:

OSIRIS-REx’s key science objectives include:

Return and analyze a sample of Bennu’s surface
Map*the asteroid
Document the sample site
Measure the orbit deviation caused by non-gravitational forces (the Yarkovsky effect)
Compare observations at the asteroid to ground-based observations

And since I was interested in the camera instrumentation:

The OSIRIS-REx Camera Suite (OCAMS) consists of three cameras: PolyCam, MapCam, and SamCam. These cameras will “see” asteroid Bennu as the spacecraft first approaches it. OCAMS will then provide global image mapping and sample site imaging and characterization. Finally, OCAMS will record the entire sampling event during the touch-and-go (TAG) maneuver.

PolyCam, an 8-inch telescope, is the first to acquire*the asteroid from 2 million km away. Once the spacecraft is closer, it will image Bennu at high resolution.
MapCam searches the area around Bennu for satellites and outgassing plumes. It maps the asteroid in 4 different colors, informs our model of asteroid shape, and provides high resolution imaging of the sample-site.
SamCam will continuously document the sample acquisition event and TAG maneuver.

OCAMS was developed by the University of Arizona. The OCAMS team is led by Bashar Rizk (Instrument Scientist, UA), Christian d’Aubigny (Deputy Instrument Scientist, UA), and Chuck Fellows (Instrument Manager, UA).


10-10-2017, 03:46 AM
The Teensy is simulating the mechanical portion of the cameras. Poly has a much more complicated mechanical system than map and sam. THe board is designed to simulate both map and sam, or just poly, depending on which code is loaded.

And to add a little cachet to the OREX story above, the asteroid we are going to visit has a greater-than-zero probability of hitting earth in either 2198 or 2199 - earth is within the error bars on its orbital position in those 2 years. We don't know its orbit very well, because we only get to observe it for a small fraction of its orbit as it whizzes by earth each year. So OREX will spend 2 full orbits hanging out with Bennu and being a radio beacon that we can use to track its exact orbit, so we can shrink those error bars and determine for sure whether earth is a target.

10-10-2017, 12:43 PM
I hope you'll post some photos when it's built? :)

10-10-2017, 02:11 PM
Absolutely. Ordered the quick-turn PCBs yesterday, Teensys and other parts should arrive tomorrow. Software is already ported from the PIC environment and tested on my simple breadboard. I hope to have the first board up and running by early next week.

10-19-2017, 05:57 AM
Breadboarded and software verified.
Had some debugging fun - the optoisolators have a slight assymetry between the turn on and turn off time, and the Teensy is actually fast enough to capture it (unlike the old PIC) - had to add another state to the state machine to handle the case where the coil polarity is both positive and negative at the same time.....


10-19-2017, 06:01 AM
PCBs from Advanced Circuits - tried out their $33 for any size 2-layer board offer, very happy with the quality.

10-19-2017, 06:04 AM
Fit check - fits.

Tomorrow, the smoke test.....

11-09-2017, 05:36 AM
Heading up to Denver next week to install it in the Software Test Lab at Lockheed, where it will be connected to an exact copy of the spacecraft computer.