RTEMS on orbit

Gerald Needell needell at espmail.sr.unh.edu
Fri Aug 31 10:58:30 UTC 2012


Just a quick update.
The EMFISIS instruments on both of the RBSP Spacecraft have been turned on and are running the RTEMS application. They are both working normally and already producing interesting data! We'll be continuing to commission the instrument suites over the next few months but are off to an excellent start.

Another pair of notches in the RTEMS belt!

- Jerry

Begin forwarded message:

> From: Needell Gerald <jerry.needell at unh.edu>
> Subject: Re: RBSP Launch Delayed
> Date: August 26, 2012 12:02:39 PM EDT
> To: Joel Sherrill <joel.sherrill at oarcorp.com>
> Cc: Needell Gerald <jerry.needell at unh.edu>, "rtems-users at rtems.org" <rtems-users at rtems.org>
> 
> Joel, thanks for the "shout-out" !
> 
> For more detail on EMFISIS see http://emfisis.physics.uiowa.edu  (it even has a link back to the RTEMS EMFISIS page !)
> 
> A few more details relating to role of RTEMS :
> 
> The EMFISIS instrument is controlled by a CDPU board (see photo in http://www.rtems.org/wiki/index.php/EMFISIS_on_RBSP ) designed and built at the University of New Hampshire.
> 
> The CDPU uses the Radiation Tolerant version of the LEON 3 FPGA from Gaiser Research (now AeroFlex http://www.gaisler.com) programmed into an ACTEL RTAX 2000.
> The LEON3 FPGA is run at 21 MHz.
> The resources available to RTEMS are:
> 2 Mbytes of BAE CRAM ( we believe this is the first launch of these space qualified CRAMS )
> 2 Mybtes of BAE SRAM 
> 512 Mbytes of 3DPlus SDRAM.
> 
> The RTEMS development platform is CENTOS Linux ( the flight versions are built on a CENTOS 5.6 system, but we also use CENTOS 6.2 - the generated images are identical :-) )
> 
> The flight software application is built on a customized version of the RTEMS 4.9.2 SPARC/leon3 BSP.
> The main customization was in the start.S initialization code to copy the Trap table to SRAM and allow the application to execute from the CRAM. Since RTEMS self-modifies the Trap table when when installing interrupt and trap handlers, the Trap Table must reside in RAM.
> 
> The executable image is stored in and executed from the CRAM using the SRAM for the Trap tables, stack and general RAM.
> The SDRAM is used as a data buffer to store science data before it is transferred to the spacecraft computer for transmission to the ground.
> 
> Should it become necessary, the RTEMS application can be loaded to and executed from the RAM  or SDRAM as well.
> 
> Notably absent from the design are PROMs. The BOOT code is built into the Advanced High Speed Bus Read Only Memory (AHBROM) of the LEON3 FPGA. From this boot code, we can reload flight software into any of the memory devices if necessary. Under normal operation, the boot code verifies the functionality of the CRAM containing the RTEMS application and jumps to it for execution.
> 
> RTEMS is used for command and control of the EMFISIS suite and for formatting all telemetry for transmission to the spacecraft computer for eventual transmission to  the ground.
> 
> The RBSP mission will be flying trough the Van Allen Radiation Belts ( that is the purpose of the mission ;) and will be exposed to high levels of radiation. All of the memory devices employ EDAC to correct single bit errors and detect Multibit errors. RTEMS Trap handlers are invoked when such events occur.
> 
> The main science application is driven by a 64 Hz interrupt cycle that is used to schedule all of the sensor operations. The RTEMS application acquires data directly from the Fluxgate (DC )  magnetometer at 64 samples/second. The magnetometer was developed by the Goddard Space Flight Center. The electric field and search-coil (AC) magnetometer sensors, developed at the University of Iowa, are operated by custom DSP co-processors also developed at the University of Iowa. THe RTEMS application schedules the coprocessor operations and then downloads sampled data from the coprocessors. 
> 
> 
> This is my first time using RTEMS on a flight project. I have been working with RTEMS for about 10 years with a lot of help from this community.  I want to express my thanks to the entire RTEMS community and to Joel in particular for continued support throughout this project.
> 
> The 2 scrubbed launch attempts were frustrating, but we are looking forward to Thursdays attempt at 0805 UTC.
> 
> Spaceflightnow.com has been providing excellent coverage of the countdowns and NASA-TV has been providing launch video starting few few hours before launch time.
> 
> - Jerry
> 
> On Aug 25, 2012, at 5:51 PM, Joel Sherrill <joel.sherrill at oarcorp.com> wrote:
> 
>> Hi
>> 
>> For those who noticed the news, the Radiation Belt Storm
>> Probe(RBSP) satellites were not launched today due to
>> Hurricane Isaac.
>> 
>> The EMFISIS instrument on the RBSP is controlled by RTEMS.
>> 
>> http://rbsp.jhuapl.edu/spacecraft/instruments/instruments_emfisis.php
>> 
>> http://www.rtems.org/wiki/index.php/EMFISIS_on_RBSP
>> 
>> I won't mention those who developed it by name just in case I shouldn't.
>> If they want to pipe up and share some details, I for one would be happy
>> to hear more.
>> 
>> I am sure I speak for the entire RTEMS Community when I say that
>> I hope the mission is a success.
>> 
>> --joel
>> _______________________________________________
>> rtems-users mailing list
>> rtems-users at rtems.org
>> http://www.rtems.org/mailman/listinfo/rtems-users
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.rtems.org/pipermail/users/attachments/20120831/d8da5073/attachment.html>


More information about the users mailing list