PROGRAMMABLE TECHNOLOGIES WEB SITE

A scientific study of the problems of digital engineering for space flight systems,
with a view to their practical solution.

SX-AS SEE Test

MEC Die, 0.25 µm

BNL, October, 2000

 

Summary

This test of the SX-AS, 0.25 µm, MEC die.  This test was run with minimal supply voltages, 4.5/2.25/2.25 for most runs.  Higher voltages were used for SEL and antifuse rupture tests. It used the the TMRSX32_Pattern logical pattern.  For both of these K-Latch test parts, there were very few upsets in any of the shift register strings.

Package: CQFP256

Serial Nos. LAN3421, LAN3422.

Test frequency: 1 MHz.

DUT schematics:  TMRSX32_Logic.ppt.

Total Number of Flip-flops: 800 (user flip-flops).

Clock Configuration: HCLK

Ions used:

210 MeV Cl-35

284 MeV Br-81

345 MeV I-127

Antifuse rupture: not detected.

SEL: not detected.

Loss of control, i.e., JTAG upset: not detected.

 

Since the K-Latch uses redundant structures, the angle of rotation can be critical to making an accurate measurement. This includes not only the magnitude of the angle and the range of the ion, but also the direction of the rotation. The redundant K-Latch circuits are implemented with the redundancy in the "long" direction of the die, as showed in the figure below.  We tested rotating in two directions.  The main angle, which is normally used at BNL, is rotation, and is around a vertical axis as shown below.  The second axis used was roll.  Roll is about an axis that is perpendicular to the part - that is, in the direction of the beam.

 

   ++++++++++++++++++++++++++++++++++++++
   +                                    +
   +               DIE                  +
   +                                    +
   +                                    +
   ++++++++++++++++++++++++++++++++++++++
   ^                 ^
   |                 |
   PIN 1             |
                     |
                     |
                     |
                     |

               axis of rotation

The beam is going directly into the screen.

SEEDutCard.gif (1614112 bytes)

Test Data

BNL
Run #

 S/N

Ion

LET
MeV-cm2/mg

Tilt, Roll
(degrees)

 Supply Voltages

Time
(sec)

Flux

Fluence

 1_Err DOC DOVH DOS DOH 0_ERR Strip Charts
6 3421 Cl 22.9   60, 0 4.5,2.25 218 4.6x104

107

 0 0 0 0 0 0 LAN3421C1.pdf
7 3421 Cl 22.9 60, -90 4.5,2.25 217 4.6x104

107

 0 0 0 0 0 0 LAN3421C2.pdf
28 3421 Br 37.4 0,0 4.5,2.25 202 2.0x104 4.1x106 0 0 0 0 0 0 LAN3421B1.pdf
29 3421 Br 37.4 0,0 4.5,2.25 1.0x107 0 0 0 0 0 0 LAN3421B2.pdf
30 3421 Br 65.1 55,0 4.5,2.25 233 2.4x104 5.6x106 0 0 0 0 0 0 LAN3421B3.pdf
31 3421 Br 65.1 55,0 4.5,2.25 116 1.9x104 2.2x106 0 0 0 0 0 0 LAN3421B4.pdf
32 3421 I 59.9 0,0 4.5,2.25 123 9.4x104 1.2x107 0 1 0 0 0 0 LAN3421I1.pdf
33 3421 I 104.4 55,0 4.5,2.25 235 4.3x104 107 0 2 0 0 0 0 LAN3421I2.pdf
34 3421 I 104.4 55, -90 4.5,2.25 282 3.5x104 107 0 0 0 0 0 0 LAN3421I3.pdf
35 3421 I 104.4 55, -45 4.5,2.25 381 2.6x104 107 0 0 0 0 0 0 LAN3421I4.pdf
36 3421 I 69.1 30, -45 4.5,2.25 304 3.3x104 107 0 1 0 0 0 0 LAN3421I5.pdf
37 3421 I 59.9 0, 0 5.5, 2.75 309 3.2x104 9.9x106 0 1 0 0 0 0 LAN3421I6.pdf
38 3421 I 59.9 0, 0 5.5, 3.0 392 2.6x104 107 0 0 0 0 0 0 LAN3421I7.pdf
39 3421 I 104.4 55, 0 5.5, 3.0 775 1.3x104 107 0 0 0 0 0 0 LAN3421I8.pdf
40 3422 I 59.9 0,0 4.5, 2.25 ~0.75x107 0 0 0 0 0 0 LAN3422I1.pdf
41 3422 I 59.9 0,0 4.5,2.25 137 8.4x104 1.1x107 0 2 1 2 0 0 LAN3422I2.pdf
42 3422 I 104.4 55,0 4.5,2.25 247 4.1x104 107 0 0 0 0 0 0 LAN3422I3.pdf
43 3422 I 104.4 55, -89 4.5,2.25 289 3.5x104 107 0 0 0 1 0 0 LAN3422I4.pdf
44 3422 I 104.4 55, -45 4.5,2.25 339 3.0x104 107 0 0 0 0 0 0 LAN3422I5.pdf
45 3422 I 69.1 30, -45 4.5,2.25 314 3.2x104 107 0 0 1 0 0 0 LAN3422I6.pdf
46 3422 I 59.9 0, -60 5.5, 3.0 302 3.3x104 9.9x106 0 0 0 0 0 0 LAN3422I7.pdf
47 3422 I 104.4

55,-89

 5.5, 3.0 608 1.6x104 107 0 0 0 0 0 0 LAN3422I8.pdf

 

Analysis

Preliminary findings:

Most of the upsets detected in this test are in the "DOC" channel.  This one has a buffer between flip-flops in the shift register, most or if not all of them using direct connect.  It was hypothesized that the direct connect, which is the fastest connection in this architecture (3 different types, direct connect, fast connect, and routed) would be the most susceptible.  This preliminary finding, subject to more analysis and data reduction, suggests that additional tests should be run with the buffer placed and routed such that no direct connects are used but fast connects and routed signals are used for two of the channels.

Comparing DOH to DOVH, we see no errors on DOH but some on DOVH.  DOVH has buffers hanging on the preset and clears.  Note that the error monitor for DOVH [these two strings are TMR-hardened at the user level].  This strongly suggests that the buffers had a small cross section for SETs.

Lastly, S/N LAN3422 had a few errors on DOS, a strong of 200 flip-flops.  No errors for this string were detected for S/N LAN3421.

Home
Last Revised: January 09, 2002
Digital Engineering Institute
Web Grunt: Richard Katz