0
\$\begingroup\$

I have devices that use Xilinx XC9572XL CPLDs. They haven't been used (as in: not powered at all) for almost seven years, and symptoms would point to the CPLDs being at fault.

I couldn't find much on the durability of non-volatile storage in CPLDs. Is it possible that storage has deteriorated enough over 7 years to cause problems?

More generally, how long is a CPLD expected to last once programmed?

\$\endgroup\$
  • \$\begingroup\$ What are the symptoms that make you think the NVM is degraded? \$\endgroup\$ – Funkyguy Sep 1 '15 at 16:15
  • \$\begingroup\$ @Funkyguy: I'm just guessing at this point. But I have two failed devices, each failed the same way. There are two CPLDs in each device, one responsible for half the functionality. And I already had a similar failure years ago, after a static discharge, which wiped the memory of one CPLD. Symptoms were similar then. There are no electrolytic caps in the device, only ceramics, CPLDs, RAM, flash, a DSP and some custom chips. \$\endgroup\$ – Jan Rychter Sep 1 '15 at 16:40
  • \$\begingroup\$ Large flux of ionising radiation during storage could reduce the retention time. I would look elsewhere, if you have a working unit you can compare with then power them up and compare all marked and likely test points for supplies and external input conditions if possible. Good luck, more disclosure gets more interest. \$\endgroup\$ – KalleMP Sep 1 '15 at 19:01
3
\$\begingroup\$

The Xilinx data sheet XC9572XL datasheet states that the programming should be good for 20 years. I would expect even more if the temperature is moderate.

From XC9572XL datasheet

I would look somewhere else for problems. Electrolytic caps are the usual suspects for deteriorating with age. Make sure the power supplies for the CPLD are good - measure with a scope as well as a meter as there may be a large ripple if the caps in the PSU have degraded.

\$\endgroup\$
  • \$\begingroup\$ I was wondering if the data retention numbers also apply if the device is never powered on during all this time. \$\endgroup\$ – Jan Rychter Sep 1 '15 at 16:36
  • 1
    \$\begingroup\$ I would think that there would be less stress and thus it would improve retention when not powered. The act of reading in Flash memories does tend to discharge the floating gate (much less stress than writing however). If unpowered the temperature would probably be lower as well which would help. \$\endgroup\$ – Kevin White Sep 1 '15 at 16:56

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.