10
\$\begingroup\$

I know that many chips, such as the ATMEGA328P-AU list flash storage lifetimes at certain temperatures, but they usually cap out at 100° C.

I know that, ideally, one should include leads on their board to program the chip after soldering, but I just want to know how the flash memory is affected at reflow temperatures of ~230° C.

\$\endgroup\$
  • 1
    \$\begingroup\$ Good question. Ask the vendor. Also, scour the datasheet to see if this information is present in the datasheet. Many flash parts can survive reflow without memory errors provided the recommended reflow profile is followed. But I have heard of memory errors after reflow (maybe this was more of a problem back in the "old days.") \$\endgroup\$ – mkeith Aug 1 '17 at 3:47
  • 1
    \$\begingroup\$ We actually ran PIC18's at an operating temperature of \$150\:^\circ\textrm{C}\$ for an hour or two (not longer) on a rotating platform. They collected data and survived the effort. But they were not spec'd for it. I think we used some rated for \$125\:^\circ\textrm{C}\$. No, this is of no help regarding guarantees on the reflow temperature you mentioned. But it was an interesting durability experimental result. (This would have been more than a decade ago.) I think your question is a good one, too. I'll be interested in a good answer. \$\endgroup\$ – jonk Aug 1 '17 at 3:57
4
\$\begingroup\$

I don't think there's much risk of data corruption at reflow temperatures for a short cycle, although maybe it takes a small fraction off the usable retention lifetime.

Most of the major microcontroller vendors will sell you devices pre-programmed with your firmware - so it stands to reason they expect you to solder them. I'm talking about pre-programmed into flash, not mask ROM, and it can be done in modest quantities.

It's also common reverse-engineering practice to desolder a flash IC from a PCB to read it out.

There are certain more exotic nonvolatile memory types to which this definitely does not apply - data stored in phase-change RAM will not survive reflow, and IIRC older MRAMs were also susceptible though nowadays MRAM manufacturers are saying it's not a problem.

| improve this answer | |
\$\endgroup\$
  • 5
    \$\begingroup\$ It's also common to have chips programmed by a third-party service before assembly, if you don't like the price/lead-time offered by the chip manufacturer. \$\endgroup\$ – The Photon Aug 1 '17 at 4:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.