NASA Office of Logic Design

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

Independent NASA Test of
Actel SX-A, SX-S, and SX-SU
Field Programmable Gate Arrays (FPGAs)

KU2 Anomalies

1. S/N 54948 Out of family I_CCA

2. S/N 50952: “Set” Pin Stuck High

3. Test Equipment: Burn-In Board (BIB) S/N 335 Incorrect V_CCA Readback

4. Test Equipment: S/N 54906 LED Off

5. S/N 50968: Can not duplicate I/O ring error

6. S/N 54954: Set_n V_IH

7. S/N 50958: Failed V_IH

---

S/N 54948 Out of family I_CCA

Summary:

The initial I_CCA reading for S/N 54948 indicated an in-family, low value.  Subsequent readings, prior to the initiation of any stress tests, showed an increase of the value of  I_CCA with an increase of about 1 to 2 mA.  While this value is within the specification value of 25 mA (I_CCA + I_CCI) the device was removed from the population prior to the beginning of the stress tests.  Preliminary analysis shows no indication of any programmed antifuse damage.  Propagation delays (t_PHL and t_PLH) were within family and stable.

History

• Passed programming
• I_CCA in-family for initial room temperature ATE run
• I_CCA within specification but out-of-family for subsequent ATE runs.
• Will be subject to failure analysis.

S/N 54948 I_CCA History (All values in µA)

Test Sequence	1st	2nd	3rd	4th
Temperature	+25 ºC	-55 ºC	+125 ºC	+25 ºC
S/N 54948	410	1573	2682	1660
KU2 Population Mean	469	415	1233	456
KU2 Population Standard Deviation	127	111	161	120

---

S/N 50952 – “Set” Pin Stuck High

Summary:

S/N 50952 was programmed successfully and passed initial, room temperature ATE testing.  However, S/N 50952 failed subsequent cold temperature (-55 ºC) and room temperature ATE runs.  The device was removed from the population prior to the beginning of stress testing and was submitted for failure analysis.

History

• Passed programming
• Passed initial room temperature ATE run
• Failed subsequent cold (-55 ºC) ATE run
• Failed subsequent room  temperature ATE run
• Silicon Explorer revealed Pin 53 (Set_n) is stuck high
• Failure analysis continuing.

Additional information.

---

Test Equipment: Burn-In Board (BIB) S/N 335 Incorrect V_CCA Readback

Summary:

During KU2-1 stress testing (+125 ºC) Burn-In Board (BIB) S/N 335, located in Slot # 34 of Chamber #1, reported a voltage level of 2.5V for V_CCA; expected value was 2.85V.  A coaxial cable had been attached to the V_CCA plane and an independent measurement indicated the expected voltage level.  In preparation for the KU2-2 test, BIB S/N 335 was installed in slot #29 of Chamber #1 with no anomalies detected.

---

Test Equipment: S/N 54906 LED Off

Summary:

After KU2-1 stress testing (+125 ºC), S/N 54906 was not driving the "IO_mon" LED on the Burn-In Board (BIB).  The expected value of the LED is to be driven.  It was determined that S/N 54906's socket was damaged, indicating a test equipment failure.  The BIB was replaced for KU2-2 testing.

History

• S/N 54906 passed programming
• S/N 54906 passed initial tri-temperature ATE runs
• S/N 54906 subjected to stress testing during KU2-1 (+125 ºC), located on Board S/N 328, socket #5.
• IO_Mon LED observed to be off
• S/N 54906 passed ATE testing and all values and deltas were nominal
• Reference unit installed in BIB S/N 328, socket #5.  I_CCI was found to be anomalously high, 240 mA.  I_CCA was nominal.
• Reference unit installed in a different socket on BIB S/N 328.  I_CCI was found to be nominal (85 mA).  I_CCA was nominal, unchanged from socket #5.
• BIB S/N 328 was replaced with S/N 331. Bench testing indicated nominal performance.
• For KU2-2 testing, BIB S/N 331 will be used in place of BIB S/N 328.
• BIB S/N 328 is in the queue for troubleshooting and failure analysis.

---

S/N 50968 - Can not duplicate I/O ring error

Summary

S/N 50968 initially failed ATE functional testing in the I/O ring and showed small but noticeable changes in leakage current.  It passed later testing and a first round of extended functional testing.  No excessive delays were detected on any net segment.  S/N 50968 is continuing test with the KU2 population into KU2-3 and will be closely monitored.

---

S/N 54954 - Set_n V_IH

Summary

S/N 54954 had an off-nominal reading for the Set_n V_IH level following KU2-4 (completion of 1,000 hours of test at 125 °C).  The data is being reviewed and the part will be subject to analysis, as appropriate.

History after Programming

• 1000 hour HTOL
• V_IH level (3.65 V) shifted by more than 300 mV from time 0 measurement
• Single input pin pin 53 (“Set_n”) exceed CMOS I/O standard specification limit of 3.5 V
• Bench level testing using the IEEE 1149.1 Boundary Scan Register method correlated to ATE measured V_IH
• Curve trace of pin 53 – I/V curve nominal.
• Suspect – Partial ESD damage to transistor N3b in I/O buffer

Current and Near-Term Plans

• I/V Curve Trace of V_CCI with respect to GNDQ of S/N 54954 and reference units
• FIB pad addition underway for signal INX micro probing
• Micro probe FIB Pad

---

S/N 50958: Failed V_IH

Summary

After the completion of 1,250 hours of operation, S/N 50958 failed the V_IH test at the end of step KU2-5 (-55 ºC). An I_CCA increase of approximately 1 mA was noted.  ESD is suspected on one of the JTAG pins.

 

Home Page for Independent NASA Test of Actel SX-A, SX-S, and SX-SU FPGAs

Home - NASA Office of Logic Design
Last Revised: May 09, 2005
Digital Engineering Institute
Web Grunt: Richard Katz