Link to NASA Agency Website
NASA-NEPP Logo - Click to return to the NEPP Home Page    + View the NASA Portal
NEPP Website Search  
Powered By Google
Link to the NEPAG Home Page Link to the Photonics Page Link to the NEPP Parts Page Link to the NEPP Packaging Page Link to the NEPP Radiation Page Link to the DC/DC Converters Page Link to the NASA Parts Selection List
Link to the NASA Technical Standards Server Link to the NASA Workmanship Website Link to the Wirebond Website Link to the Tin Whiskers Home Page Link to the Plastic Encapsulated Microcircuits Page Link to the NEPP Publications Search Link to the NEPP Experts Page Link to the NEPP Calendar
Your Path: Home » Parts » Parts Links » One-NASA Position on EEE Parts Issues
  Market Survey: SSR's
  CALCE
  Reliability Information Analysis Center (RIAC)
  NASA GSFC/Xilinx Meeting Presentations
  Search
 
 
 
 
Positions Shown Herein:

One-NASA Position on the Use of Actel RTSX-S, RTSX-SU, and RTAX-S FPGAs in Space Mission Hardware, Ken LaBel, August 2005. 


######

Actel RTSX-S, RTSX-SU, and RTAX-S FPGAs
Author: Ken LaBel, NASA GSFC

 

The question of Actel device usage for space missions is a complex question with many variables. What follows below is a brief NEPP recommendation for project usage, for the latest generations of devices.


RTSX-S devices (0.25 um MEC foundry)
These devices have been undergoing investigation for reliability-related issues (NASA/NESC/NEPP/OLD, Aerospace, others). At this time, the critical failure mode is such that device speed is reduced on a single net, from a damaged programmed antifuse. Thus, if you plan on using these devices (as in, already have bought them) , there are several recommendations:

- Utilize most recent programming algorithms for this device, and
- Ensure designs satisfy all specifications (many designs do not).
- Recognize that these devices are extremely ESD sensitive.
- Lot specific DPA for wire bond integrity is recommended.

There are three programming algorithms available: the original programming algorithm, the "new programming algorithm" which addresses one of two classes of antifuse problems found, and the "modified new programming algorithm" which addresses both classes of antifuse problems found.  These improvements appear to mitigate and not eliminate the issue.

=====================================

RTSX-SU devices (0.25 um UMC foundry)
These are the "new" RTSX-SU devices for space and have also been undergoing reliability tests (NASA/NESC/NEPP/OLD, Aerospace, others). They, at this point, seem to have improved reliability versus the MEC foundry parts. They are the current "recommended" device for projects that have not yet procured devices. The recommendations are:

- Utilize the most recent programming algorithm available, and
- Ensure designs satisfy all specifications (many designs do not).
- Recognize that these devices are extremely ESD sensitive.
- Lot specific DPA for wire bond integrity is recommended.

There are two programming algorithms.  The first is the original programming algorithm.  The improved programming algorithm is called UMA for "UMC Modified Algorithm."  Also, the routing software is being modified to minimize peak currents through a net using only a "fast connect" called SAL for "S Antifuse Loading."

=============================================

RTAX-S devices  (0.15 um UMC foundry)
These are the latest generation space devices from Actel. Reliability testing by Actel has shown promise. However, until independent reliability and radiation tests are performed, we recommend that near-term projects not baseline these devices. This opinion will likely change within 6 months once independent data exists such as that generated for the RTSX family of devices.

All devices:
- More complete single event data is desired on all device types above (high-temp latchup and high-speed single event upset/transient).
- In addition, we recommend lot-specific TID tests be performed. This is standard practice for Actel.
It should also be noted that traditional "good engineering practices" such as clock derating should also be applied.

There is currently debate among the NASA centers on the need for post programming electrical test and post programming burn in.  Proper risk management practices would dictate the need for an independent verification of the programming procedure.  This can only be done via an electrical test on a sophisticated tester with through test vectors.  This needs to be built into planning for cost and schedule well in advance of any testing.  In all cases, particularly if post programming testing and burn in is not conducted, significant operation at use conditions and temperatures for a minimum of 500 hours is recommended.


 

More information about Programmable Technologies


 

Privacy Policy and Important Notices NEPP Help for Site Navigation NEPP Feedback Form Link to the NEPP Site Map
USA.gov  
NASA Logo - Click to visit the NASA Agency Website
NEPP Program Manager: Michael J. Sampson
Responsible NASA Official: Michael J. Sampson
Webmaster: Carl M. Szabo, Jr.
      Last Updated: May 12, 2008