AM3354: Operation at high altitude

Hi Hiroshi-

• Can TI devices like AM335x be used in Highlands – Yes

>>I think that this problem requires not only the device but also board level verification, but please tell me whether there is any problem with these devices.

It depends and the area to focus is soft error rates (SER). Soft errors are only really a concern for storage elements on a device - specifically SRAM’s and registers.  Obviously Soft errors are endemic to all semiconductors, not just TI, as it’s a problem of charge collection and device physics.  So if the board has other devices with SRAM’s and registers then yes there will be a similar impact.

If these other components are from TI then we may be able to provide an analysis report in the same manner as we have for AM335x , but if they are from another supplier then your customer would need to work with those vendors and TI will not be able to provide any guidance on those.

High Level Summary:

• The neutron particle flux density increases with altitude and relative proximity to the poles (north or south)

• By default all base line calculations for soft errors are calculated assuming a geographical location of New York City @ Sea Level to determine the neutron particle flux density.

• At New York City @ Sea Level the geographical scaling factor is 1

• The scaling factor for any location on the planet within terrestrial altitudes (not outside the atmosphere i.e. in space)

can be calculated using the link  

• Any SER analysis report we supply (these are only provided under proper NDA) provides customer control of the geographical scaling factor to account for this.

• Just as a reference for 3000m altitude above New York City the scaling factor comes out to be ~10 

Hope this helps.

regards

Mukul