This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

SM320F2812-EP: SEE Data for SM320F28335GJZMEP

Part Number: SM320F2812-EP

I need SEE related information for the following part.

SM320F28335GJZMEP

More specifically I need the SEE cross-section (cm2)

A similar Ti part with a published SEE cross-section will work as well.

Was the actual part SEU tested or was a similar part tested?

Was the part tested with protons or neutrons?

  • Casey,
    The Enhanced Product (EP) devices are not intended for use in radiation environments. Thus, TI has not tested them for SEE effects.
    You can see the current -SP space rated product and associated radiation test reports here:
    www.ti.com/.../space.html

    If this answers your question, please click "This Resolved My Question"
    Regards,
    Wade
  • I understand, however I am just looking for the processor cross-section to perform an independent susceptibility analysis per the request of our customer.
  • I understand, however I am just looking for the processor cross-section to perform an independent SEE susceptibility analysis per the request of our customer. However, I need the processor cross-section to perform this calculation. The cross section of a comparable processor could suffice as well.
  • Casey,
    Unfortunately we do not have any SEE data on similar processors.
    There may be some third party data.

    Just to be clear, you are looking for proton or heavy ion cross sections correct? I just want to make sure you are not researching soft error rate (SER) from terrestrial neutrons? We can model SER flip flops and memory.

    Regards,
    Wade
  • Im fairly new to SEE analysis and am trying to meet the needs of my customer.

    My customers requirement reads-

    Use the following equation to Calculate simplified SEE rate:

    SEE rate per device-hour = integrated neutron flux(n/cm2 per hour) x SEE cross-section (cm2 per device)

    My understanding is that there is cross section measurement (in cm2) that I should be able to obtain for my device or a similar device.
  • Perhaps if you could tell me the environment that device will be used in.
    Space, high altitude avionics, or possibly nuclear reactor?
    I believe the only one that I could help with for standard EP product would be the high altitude SER calculation.
    Coming up with error rate would require altitude, voltage, and latitude.
    Our radiation expert is out of the office until after the new year.
    Regards,
    Wade
  • I appreciate your patients in trying to figure this out.

    This is what was provided to me:

    Energy>10MeV:
    the flux for 8202 feet altitude and 60° latitude is 200 n/cm2/h.

    1MeV<Energy<10MeV
    Due to lower SEE thresholds of smaller geometry devices (generally below 100 nm) for these devices when
    calculating SEE rates the flux of neutrons between 1 and 10 MeV should be added to the 6000 n/cm2 per hour

    The flux for 8202 feet altitude and 60° latitude is 348 n/cm2/h.
  • Casey, your description seems to align with neutron SER calculations.
    I will have our radiation expert take a look at this and provide a response.

    Regards,
    Wade
  • "Energy>10MeV:
    the flux for 8202 feet altitude and 60° latitude is 200 n/cm2/h.

    1MeV<Energy<10MeV
    Due to lower SEE thresholds of smaller geometry devices (generally below 100 nm) for these devices when
    calculating SEE rates the flux of neutrons between 1 and 10 MeV should be added to the 6000 n/cm2 per hour

    The flux for 8202 feet altitude and 60° latitude is 348 n/cm2/h."

    Hi Casey,

    The 10MeV neutron cut-off came from the original JEDEC JESD89 and 89A. At the time of the A revision we discussed extending neutron flux down to 1MeV but there was not enough data to justify doing that.  Fully agree that as the critical charge has dropped it makes sense to drop the min neutron energy to 1 or 2 MeV. There is a JESD89B revision coming soon and they may drop it in there but until then Emin is 10MeV. 

    During the discussions we did a rough calculation and found that assuming Emin ~ 10MeV then NYC sea-level is ~ 13 n/cm2-hr. Dropping the Emin to 1MeV gave a flux of about 20 n/cm2-hr.  Not a huge change.

    Using the JEDEC specified tool that we set up for calculating neutron fluxes (www.seutest.com/.../FluxCalculator.cgi) with the 10MeV cutoff, for your conditions alt:8202 feet and 60 N latitude I get:  7.83 x 13 n/cm2-hr ~ 101 n/cm2-hr if you assume conservatively that you double the flux by dropping Emin to 1MeV you get 202 n/cm2-hr.  Hope this helps. Rob

  • Robert, thank you for your answers to this, I am working with Casey on this project.

    Can you state what the cross sectional area of the component is and with the comparison to the customer calculated flux value of 348n/cm2/h, is a lower calculated flux number for the component better or worse? Does a flux of 348n/cm2/h mean that a part of 202n/cm2/hr will not survive?

    Thank you!

    David

  • Hi David,

    I do not have access to the die size.  I would guess that it is much less than 1 cm2. 

    The neutron flux is what creates single event upsets in processors - so more neutrons means a higher failure rate. Please note that you cannot simply multiply the neutron flux by the die area to calculate the number of errors. The neutrons can create errors in semiconductor devices only when a neutron has a nuclear reaction with a silicon or oxygen or other nuclei in the device. This nuclear reaction cross-section is often a  very small fraction.

    Since the part in question is an -EP and in my BU I can support an SEU estimate of the device if this will help.

    Rob

  • Rob, an SEU estimate would be very helpful to have. How long does that take?

    David
  • David

    Please contact directly at rbaum@ti.com. I can support an SEU estimate this week (next week is bad) if I get:

    Minimum Vdd

    Maximum Temp

    Maximum Altitude & Latitude

    Optional = amount of memory used for application

    max flight time.

    Thanks, Rob

  • Hi Casey/David,

    I a[pologize that I did not see the original thread when replying to your e2e.

    "I need SEE related information for the following part. SM320F28335GJZMEP"
    "More specifically I need the SEE cross-section (cm2)"

    Wade is correct that we do not provide space radiation data for -EP parts, however from your follow on message it is clear you are talking about a terrestrial environment (neutrons) and for this we do support SEE modeling. Our models defined and are compliant with JEDEC JESD89A so we do not provide SEE cross-section for parts - we do provide SEE failure rate estimations in FIT if this is helpful.

    "A similar Ti part with a published SEE cross-section will work as well."
    The SEE estimations are based on testing of SRAM and logic test structures in each technology node with each SRAM bit cell thus they are representative fo actual product failure rates (at least for an upper bound).

    "Was the actual part SEU tested or was a similar part tested?"
    No we use SRAM and Logic test chips specifically design to capture SEU, MBU, SEL, etc. these are then adjusted to the amount of SRAM/Logic in the device to come up with an upper bound for the failure rate.

    "Was the part tested with protons or neutrons?"
    The test chips were tested at Los Alamos Neutron Labs which provides one of the better fits to the actual terrestrial spectrum. We also test with alpha particles to account for this failure mechanism as well.