Hello once again!  I hope that you have been enjoying my recent series on single event effects with high speed ADCs.  So far we have looked at TID (Total Ionizing Dose), SEL (Single Event Latchup), and SET (Single Event Transient).  This month the focus is on SEU (Single Event Upset).  This is quite closely related to the SETs we looked at over the last two months.  It is possible that a SEU could result in an SET depending on the ion strike and the resultant behavior of the circuit.  For the purposes here though we will look at them separately to help provide a bit of clarity into the underlying cause.  For a high speed ADC we will consider an SEU as a configuration register upset.  The AD9246S is the device under test.  As it turned out there were no configuration register upsets observed for the AD9246S but in my Planet Analog blog I discuss what an SEU might look like.  A brief idea is that an ion strike results in a configuration bit(s) upsetting.  It is expected that the bit(s) would not permanently flip but would revert back to the expected value.

Recall that it is important to perform this testing so that the device behavior in space can be predicted.  The Weibull fit curve is generated from the data and then input into a model factoring in the expected orbit to generate a probability of device upset.

Example Weibull Fit Curve - AD9246S


It is imperative to perform the testing to generate this data and predict the device performance in space since, as we all are well aware, it is not very easy, dare I say impossible in many cases, to repair/replace devices once they are put into an application in space.  Obviously, there are things like the ISS (International Space Station) that are routinely visited and repaired, but something like Juno which is orbiting Jupiter that would not be able to be repaired nor worth the cost required to attempt to do so. If you'd like to learn more I encourage you to check out my blog over on Planet Analog.  I hope you have learned a lot so far and found SEEs interesting.