Failures in Aerospace Applications, Part 1

Integrated analog ICs are often utilized in aerospace applications to reduce the dimension of power management modules that are inserted, for example, in space stations or satellites.

Complex integrated circuits increase the density of devices that can be inserted in the aerospace modules, and they enhance the scenario of possible functionalities in these very efficient aerospace electric modules. The main goal of the aerospace application engineering team is to ensure a zero failure condition.

In several of my blogs, I've described the damages that radiation sources — electrons, protons, heavy ions, gamma rays — may cause. Radiation is present in the open space environment (see Figure 1 below). It can create some failures in the elementary components of a power management IC (i.e., MOSFET, BJT, IGBT) and, consequently, in the whole IC.

Figure 1

The radiation sources causing the single event error: cosmic rays, solar particles,and protons trapped in magnetic fields.(Source: ESA)

The radiation sources causing the single event error: cosmic rays, solar particles,
and protons trapped in magnetic fields.
(Source: ESA)

The sources of radiation are the causes of failures in the integrated circuits for aerospace applications. Now let's consider the effects of these failures of a part of the circuitry on the overall functionality of a complex integrated system on a chip. The main relevant failures that may occur in an aerospace application integrated board are known as single event errors (SEEs) and are caused by two of the four radiation sources of the aerospace environment: the protons and the heavy ions, as shown in Figure 2 below.

Figure 2

The mechanism of the single effect event caused by protons and heavy ions.(Source: ESA Microelectronics)

The mechanism of the single effect event caused by protons and heavy ions.
(Source: ESA Microelectronics)

The mass of these radiation particles is not negligible compared to the atom's dimension; hence, this radiation can cause displacement damage in the silicon structures.

These particles create electrical charges all along their path, so the radiation sources generate parasitic currents inside the material they cross. These currents generate disturbances that can lead to functionality losses. These losses may be temporary or permanent, but either way, the whole integrated system will be corrupted because of the interconnections among the different blocks.

There are several types of single events:

  • Single event latchup
  • Single event upset
  • Single event transient
  • Single event burnout
  • Single event functional interrupt
  • Single event gate rupture

In addition, post-irradiation gate stress may occur during the electrical stress of the gate oxide of, for example, a power MOSFET by means of a reverse bias testing procedure and the gate oxide breakdowns.

Beginning in part 2 of this series, I will describe in deeper detail the single events listed above. Have you experienced SEEs? If so, what kind did you experience? What do you think about the importance of providing a SEE screening for the IC to be utilized in radiation hard applications?

34 comments on “Failures in Aerospace Applications, Part 1

  1. etnapowers
    February 26, 2014

    The most common Single Event Error, is the SEL, that is described in details in the part 2 of this Blog series. However all of these errors (SeL,SEU,SEGR…) may happen.

  2. etnapowers
    February 26, 2014

    SEGR failure and PIGS rupture are two ruptures that can destroy the device during and after the irradiation, these failures may compromise partially or even totally the functionality  of the whole aerospace model, so it's important to avoid the both of these ruptures.

  3. etnapowers
    February 26, 2014

    A PIGS failure implies a possible failure during the lifetime of the integrated component , so it's dangerous as well as SEGR.

  4. etnapowers
    February 26, 2014

    “The mass of these radiation particles is not negligible compared to the atom's dimension; hence, this radiation can cause displacement damage in the silicon structures.”

    The DD (Displacement Damage) effect depends on the device that is exposed to the radiation, in the part  6 we will see that an N-channel power MOSFET for example is insensitive to Displacement Damage DD, whilst a power BJT is sensitive to the radiation's effect.

  5. chirshadblog
    February 27, 2014

    @etnapowers: Yes the risk is really high and in most scenarios this do backfire very badly.         

  6. chirshadblog
    February 27, 2014

    @etnapowers: You cannot prevent errors happening. Electronics are built by humans and these humans are bound to make errors. Also with errors only you figure out the other possibilities and alternatives.          

  7. chirshadblog
    February 27, 2014

    @etnapowers: Does it have a very strong effect and also how does it impact the user ?           

  8. etnapowers
    February 27, 2014

    @chirshadblog: the SEGR  PIGS is described in details in the part 5 of this blog series. I can anticipate that the silicon oxide SiO2 is impacted by both of these ruptures , hence all the technologies presenting this material may be affected

  9. etnapowers
    February 27, 2014

    @chirshadblog: agreed, this BLOG series describe all the possible failures, because all of these errors may happen, my consideration was only a statistical evaluation.

  10. Netcrawl
    February 27, 2014

    Its a tough place to work because we're dealing with a different environment, the design and development of space components should follow rigorous process to ensure their suitability for operations and survivability in safety-critical environment.

  11. Netcrawl
    February 27, 2014

    thanks for that @etnapowers, the reality is displacement damage could introduce different failure modes, such failures could occur in both hardened and unhardened circuits. Failure in voltage regulators is especially important because their failure can impact the operations of key circuits elements and space applications.

    Protons still the dominant source of ionization damage and proton damage will remain an important and complex issue in every space applications.   

  12. etnapowers
    February 28, 2014

    Yes Netcrawl, DD is really dangerous because it's very difficult to estimate with a good precision what will be the functionalities that are going to be affected by the sources of radiation. Moreover there are some ICs , like VR as you correctly said, that play a key role for the application. A good protective screen of these ICs is a must in aerospace application where the control is many times by remote and repairing a failed part is difficult.

  13. eafpres
    February 28, 2014

    @Paolo–I have watched with some skepticism the emergence of “phonesats”.  I think it is sad that NASA is reduced to such PR stunts to maintain enough popular appeal to fund real science.  (See this marketing piece.)

    Clearly smart phone electronics are not as reliable as many general purpose electronics.  The density is very high in the processors, and the target lifetime is only a few years in normal use.  Also, they already suffer failures–who has not seen glitches on their phone?  Is it a harware single event, or bad software?  Who knows.  But now you put that phone in space where you have all the radiation and particles, and it won't last too long.

    So, just like almost everything nowadays, we have entered the era of truly disposable satellites.

  14. Davidled
    February 28, 2014

    Space industry might review component material that are flying on space, not only temperature requirement, but also insensitive to radiation. Material engineer continues to develop the robust particles for tough environment.  I remembered that semiconductor class that presented all electronic components.

  15. amrutah
    March 1, 2014

    @etnapowers: Is it that, since the BJT is a junction device the we see a damage due to displacement of charge carriers?

  16. amrutah
    March 1, 2014

    @Paolo: Though I my daily bread winner is IC design and semicon industry, I have never worked on IC's for aerospace applications.

       The single event damages that you mentioned are something which we never consider while design, but are concerned with hot and cold process, technology yield etc.  Are aerospace application IC built using 200mm-300mm wafers are the wafers are high yield wafers.  Are these IC's reliabity target of 9σ?

    Thanks for sharing and eager to learn more here.

  17. amrutah
    March 1, 2014

    @Paolo:  In the figure above, You have shown only n-channel devices getting affected due to radiations or sub-atomic particles (I read one of previous comment where DD affects BJTmore than a MOSFET).  Is this true for all devices?

  18. Netcrawl
    March 2, 2014

    @Daej successful space operations and missions requires an understanding of the mechanisms that cause degradation and radiation testing of the space components in order to ensure that they will withstand the harsh environment. The reality is protons in space can cause permanent damage in some type of electronics which could lead to total operational failure. Proper shielding is  “a must things” in every space operations, solar flares are most important here, an intense solar flares could be devastating capable of disrupting anything.   


  19. etnapowers
    March 3, 2014

    @Blaine: I think the “phonesats” project might be successful only if the overall reliability will be guaranteed, the basic idea is interesting, but it has to work well to be successful.

    Anyway we will see very soon the test results:

    “During the PhoneSat Project's time in orbit, a global amateur radio community will be engaged, able to download and upload packets of data”





  20. etnapowers
    March 3, 2014

    Yes amrutah, you are right, the DD impact the doped areas as I expose in the part 6 of this series of blogs.

  21. etnapowers
    March 3, 2014

    @amrutah: you're welcome, the aerospace requirements are 3σ at wafer sort step , some market require als a geografical test, to put on evidence the  parameter drift in a particular area of the wafer. For these kind of qualification, the requirement is 3 to 4 σ drift.

  22. etnapowers
    March 3, 2014

    @amrutah: MOSFETS, whatever having  channel n or p, are not affected by DD because the electric current flows in a low doped area, and the protons or heavy ions cannot damage too much the devices. This is true for all the devices in which the electrical current flows in parts having a low doping level.

  23. Davidled
    March 3, 2014

    Well, I wonder whether or not MEMS device would be impacted by DD. In generally, MEMS is more sensitive component than any other high voltage MOSFET or transistor. I understand that operating voltage range is much lower than any component.

  24. eafpres
    March 4, 2014

    @DaeJ–when you talk about MEMS being more sensitive, what kind of sensivity are you describing?  I'm guessing more sensitive to ESD?

  25. etnapowers
    March 4, 2014

    @DaeJ: MEMS are sensors whose basic principle of functioning is based on the piezoeletric effect. This effect does not involve high current values and high doped areas into the device, so I guess that MEMS won't be affected by DD. Anyway we will know very soon if this is right, as soon as the “PhoneSat” project (cited by eafpres) will start, utilizing a smartphone , MEMS powered, in the space environment.

  26. amrutah
    March 4, 2014

    @Netcrawl: “Proper shielding is  “a must things” in every space operations,…”

       Inside the IC we shield lines (signal routings) either to a ground or supply plane.  What does a shield mean here? Does the chip package comes with a proton or heavy ion absorbing material or does the entire system sit inside a box of vaccumm which can reflect EM radiations and repel protons and heavy ions?

  27. eafpres
    March 5, 2014

    @etnapowers–there are Piezo MEMS devices like microphones etc., but there are other devices that have cantilevers at nano scale, and tunable capacitors where the MEMS distorts a capacitor plate to tune it.  I think these devices might be more susceptible to proton damage or induced currents.

  28. samicksha
    March 10, 2014

     The drain and source may be doped of opposite type to the channel, in the case of depletion mode FETs, or doped of similar type to the channel as in enhancement mod.

  29. chirshadblog
    March 10, 2014

    @eafpres1: Yes there is a risk as you correctly pointed out but what I feel is that if you are to climb to the next level there should be some risks taken. Anyway assessing the damage is very vital before taking the risk. 

  30. Sachin
    March 31, 2014

    I might be getting ahead of the teacher here but I couldn't help responding to this topic, particularly when it comes to the threats posed by Single Event Errors. In particular, the Single event latchup errors, which by the way are the most common of the SEEs, are easy to overlook and ignore but can lead to very serious repercussions in the long run especially when they occur in IC chips used in Aerospace applications where precision is ultimate.

  31. SunitaT
    March 31, 2014

    What I know and believe is that, reliability must be maintained through testing and regular checking. But we must also consider risk taking when creating a product and this is true for somebody to move to the next level. @chrihadblog, I agree with you, assessing damage is more very important before taking the risk.

  32. SunitaT
    March 31, 2014

    Exposure to radiation is one of the main reasons behind the SEEs that can occur in aerospace applications. The effects of this radiation could take different forms. For instance, it could induce or create a new current within an IC circuit making the entire IC unstable; as is the case that happens when you encounter Single Event Latchup errors.

  33. yalanand
    April 30, 2014

    I agree that exposure to radiation is the main source of single event errors that are usually common in aerospace environments. These single event errors if overlooked may cause grave damages on the aerospace environment. The errors may reduce the density of devices that are to be installed in the aerospace modules thereby reducing the functionality and efficiency of the aerospace electric modules. The aim of aerospace application engineering which is to achieve a zero failure rate is therefore compromised.

  34. yalanand
    April 30, 2014

    @chirshadblog, what you have just brought to sight is more than true. Errors are not things that you can fully avert in any type of activity or with any device. That`s why they always say, nothing is perfect. Only God is perfect. When we sit down and focus on errors that have already caused effect, we get to prevent more errors and to come up with much better electronic devices. In other words, try to look at errors in a positive way. This is the only solution to betterment.

Leave a Reply

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