Advertisement

Blog

Single Event Effects (SEEs) with High Speed ADCs: Single Event Latch-up (SEL)

Back in January blog here on Planet Analog, A Quick Overview of Radiation Effects – Single Event Effects, we took a look at single event effects in general. Now we will take what we learned there and take a look specifically at high speed ADCs and what single event effects can be observed. I plan to dive into these single event effects in the next few installments of my blog. In this installment we will look at single event latch-up (SEL) in particular for high speed ADCs. SEL is similar to the case of traditional latch-up where the device exhibits an abnormally high supply current caused by an induced path from supply to ground. The device will not return to normal operation after an SEL without a power cycle. Recall that single event upsets (SEUs) and single event transients (SETs) are soft errors that occur during radiation exposure where the device typically recovers quickly. Unlike SEL, the device does not require a power cycle or device reset to restore normal operation. This is one of the reasons SEL testing is generally performed prior to testing for SEUs and SETs.

During SEL testing, the device case temperature is set to 125o C and the supply currents are monitored to observe for any latch-up. The test is run up to an energy level such as 80 MeV-cm2 /mg and out to a fluence of 107 ions/cm2 . The goal is to determine the threshold for latch-up for the ADC with the worst case conditions for temperature and supply voltage. The setup diagram for SEE testing of the AD9246S 14-bit 125 MSPS ADC is shown below. This test setup was used for collecting results for SEL, SEU, and SET performance of the AD9246S .

Click here for larger image 

AD9246S Single Event Effects (SEE) Test Setup

AD9246S Single Event Effects (SEE) Test Setup

A general procedure for SEL testing would be similar to the one used for the AD9246S. For this device the following procedure was used:

  1. Power up the AD9246S and wait for it to attain desired test temperature (125o C in this case).
  2. Select the desired ion and incidence angle for an effect LET of 80 MeV-cm2 /mg.
  3. Turn on the ion beam and observe/monitor/log AD9246S supply currents.
  4. If no latch-up is observed and the fluence reaches 107 ions/cm2 , run is considered passing.
  5. If the test passes, run the remaining AD9246S units at the same effective LET of 80 MeV-cm2 /mg beginning at step 1. At least one run should be performed with no current limit on the power supplies.
  6. If no latch-up or destructive events occur on any runs, the SEL test is done.
  7. If the AD9246S latches up, it is considered a failing run.
  8. Shut off the ion beam and attempt to recover to initial current levels by first reprogramming the AD9246S registers. If that fails, power down the AD9246S.
  9. Reapply power to the AD9246S and check the current levels and check for a destructive latch.
  10. Proceed with SEL characterization.

The results from the SEL testing of the AD9246S are tabulated below. The four different AD9246S devices tested exhibited no signs of latch-up out to an effective fluence of 107 ions/cm2 up to an LET of 80 MeV-cm2 /mg.

Table 1

AD9246S Radiation SEL Test Results

AD9246S Radiation SEL Test Results

As mentioned in the test procedure steps above, the power supply currents of the AD9246S are monitored while the device is irradiated by the ion beam. For the case of the AD9246S on the test board used, the power supplies are AVDD, DRVDD, and DRVDD_HK (DRVDD supply used for housekeeping purposes during the test). While performing the SEL test, the power supplies are set to the maximum value for the AD9246S which is 1.9V for AVDD and 3.6V for DRVDD and DRVDD_HK. The supply currents are monitored for latch-up while the device is exposed to the ion beam. Power supply currents for AVDD and DRVDD for a typical SEL test run on the AD9246S device with serial number 39 are shown in the plots below.

AD9246S AVDD Power Supply Current During SEL Test Run

AD9246S AVDD Power Supply Current During SEL Test Run

AD9246S DRVDD Power Supply Current During SEL Test Run

AD9246S DRVDD Power Supply Current During SEL Test Run

Notice that the AD9246S power supply currents exhibit very little variation from their nominal values during the SEL test run. As reported in the Table 1 above there was no latch-up observed during the test. The AD9246S performed very well in this test exhibiting no latch-up at an LET of 80 MeV-cm2 /mg out to a total fluence of 107 ions/cm2 . More details on the SEE test results can be found in the SEE report on the Analog Devices website here: AD9246S Single Event Effects Test Report.

Performing the SEL test is a good first test when looking at SEEs for a particular device. In some cases the latch-up event may not be catastrophic, but in many cases an SEL event can be destructive so it makes sense to begin radiation testing for single event effects with this test. If a device has destructive latch-up at a low LET value then it is most likely not suitable for many space applications. Recovery can generally be made from other SEE events such as SEUs and SETs but a destructive latch-up is typically non-recoverable since many times the device is rendered unusable. If a device exhibits early onset of latch-up at lower LET values it is an indication that the device may not be suitable for space applications. However, thus far we see that the AD9246S exhibits no latch-up at an LET of 80 MeV-cm2 /mg which puts us off to a great start in our look at single event effects testing with the AD9246S. In my upcoming blogs we will continue looking at SEEs for high speed ADCs as we move the discussion on to SEUs and SETs. As we have discussed these events are not typically destructive like an SEL event could be. I hope you’ll join me in the next installments as we continue looking at SEE testing for high speed ADCs.

0 comments on “Single Event Effects (SEEs) with High Speed ADCs: Single Event Latch-up (SEL)

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.