SBAK021 December   2024 AFE7950-SP

 

  1.   1
  2.   2
  3.   Trademarks
  4. 1Introduction
  5. 2Single-Event Effects
  6. 3Device and Test Board Information
  7. 4Irradiation Facility and Setup
  8. 5Test Setup and Procedures
  9. 6Destructive Single-Event Effects (DSEE)
  10. 7Single-Event Effects (SEE)
  11. 8Event Rate Calculations
  12. 9References

5 Test Setup and Procedures

SEE testing was performed on AFE7950-SP devices soldered down to AFE7950EVMs. The units were de-lidded and were then thinned from the backside down to 50um to enable heavy ion penetration through the Si .The test setup also includes a TSW14J56EVM to aid in FFT captures of all 6 RX Channels and a spectrum analyzer to capture the 4 TX Channel DAC outputs. All RX inputs are connected to a R&S SGS100A signal generator using a splitter and the TX outputs are connected to the Aligent E4445A Spectrum Analyzer through a RF switch.

The device configuration used for testing is as follows:

  1. Device/EVM was powered up to Vnom using an external 5.5V, 5A max power supply.
    1. The 5.5V is broken out to 0.925V, 1.2V, and 1.8V by on-board LDOs
  2. Internal PLL is used. Reference clock is 491.52MHz, sent from the on-board LMK04828.
  3. Fin on the ADCs was set to 9520MHz, Fout on the DACs was 9520MHz. The NCO was set to 9500MHz for both the RX and TX Channels.
  4. RX and TX channels were tested independently. TX channel was not looped back to RX.
  5. LMFS Config was set to the following:
    1. ADC LMFS: 6-12-4-1-0
    2. DAC LMFS: 8-8-2-1-0

Before the start of every run, a software reset and reprogramming of the device was done and then frequency spectrum plots were recorded for all the RX and TX channels. RX channels were captured in the HSDC Pro software using the TSW14J56EVM. TX channels were captured directly from the spectrum analyzer. The beam was then started and subsequently stopped when we observed SNR degradation. However, while we observed and captured all of the RX and TX channels individually, we did not monitor all the channels at the same time. This means that for our testing, only RX channel FB1 was taken into account when determining when to stop the run. More details on how this affects our calculations can be found in section Single Event Upsets. After the beam was stopped, frequency spectrum captures were taken of all the RX and TX channels. Software resets were then performed to make sure failing channels were able to be recovered.

All equipment was controlled and monitored using a custom-developed LabVIEW™ program (PXI-RadTest) running on a HP-Z4™ desktop computer. The computer communicates with the PXI chassis through an MXIExpress cable and a NI PXIe-8381 remote control module. During SEL and SET testing the device junction temperature was maintained at 125°C by using a Closed-Loop PID controlled heat gun (MISTRAL 6 System (120-V, 2400-W)) that was set to only cool air. Die temperature was verified built-in die temp sensor.

AFE7950-SP AFE7950 Block Diagram Figure 5-1 AFE7950 Block Diagram