0
\$\begingroup\$

Has anyone run this part to its rated temperature of 150 Celsius? We are able to run up to and beyond this temperature, but if power is cycled the part will not start reliably. We have tried running with and without the PLL, with the internal oscillator, a crystal, a high temperature oscillator, and an oscillator outside of the oven, with brownout on and off, and with various circuits on the reset pin and at various frequencies. The part is only using two output compare modules, one UART, one ICD interface, and powering a couple of LED's. All of the other pins are floating. We have monitored the voltage on VCAP and it stays around 2.5 volts. The part recovers after cool down. We have also tried letting the part soak unpowered for long periods to make sure it is not self heating.

If there are any tricks to make it start or if you have had a similar experience, I would be grateful to hear.

Thanks.

\$\endgroup\$
  • 1
    \$\begingroup\$ Any way to determine thermal rise inside the chip due to operating conditions (clock speed, supply voltage...)? The datasheet specifies +155ºC max junction temp. Perhaps this is being exceeded. \$\endgroup\$ – AlmostDone Jul 26 '18 at 20:51
  • \$\begingroup\$ Interesting stuff, John. I have no suggestions for you. However, I am curious about what solder you're using at these temps. We've tested various alloys under temperature and stress, and noticed trouble for 63/37 and SN60/PB29/CU1 at 165C, SAC at 195C, and a high-temp variety at 220+C. Joints would fail if stressed at/above these temps. \$\endgroup\$ – Chris Knudsen Jul 26 '18 at 20:52
  • \$\begingroup\$ @ChrisKnudsen we are using Sn96 \$\endgroup\$ – John Birckhead Jul 26 '18 at 21:06
  • \$\begingroup\$ @AlmostDone we have tried powering down and soaking at 150C to make sure there is no self-heat issue. \$\endgroup\$ – John Birckhead Jul 26 '18 at 21:08
  • \$\begingroup\$ Monitor the VDD, to ensure the bypass capacitors achieve low ripple at those high temperatures. \$\endgroup\$ – analogsystemsrf Jul 27 '18 at 3:55
5
\$\begingroup\$

Thanks for everyone's good suggestions. It turns out that there are two versions of silicon for this chip, and the earlier version is only rated to 140 Celsius. Sheesh. This is noted in the "errata sheet." This is the second time I have been bitten by not reading the errata sheet. For reasons that are not clear to me, Microchip publishes errata sheets for their parts rather than revising the data sheets. Anyway the moral is: if you use Microchip, read the errata sheet!

Thanks again.

\$\endgroup\$
  • \$\begingroup\$ I remember when we first used the pic32mz, the errata was nearly as long as the datasheet... \$\endgroup\$ – Colin Jul 27 '18 at 14:59
  • \$\begingroup\$ Usually you want both information, the old and the new one and a way to determine which applies to the thing you have in your hands \$\endgroup\$ – PlasmaHH Jul 30 '18 at 12:44

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.