Clock stability above 60°C

Discussion created by lpcware Employee on Jun 15, 2016
Latest reply on Jun 18, 2016 by lpcware
Content originally posted in LPCWare by tbeauchant on Wed May 04 11:27:51 MST 2016
Hi Folks,

I'm running into an odd issue on a LPC1752 (verified also on a 1751). When the IC package temperature rises above 60°C the clock starts shifting in frequency. I can see this on all my outgoing signals (UART, PWMs, ...) whose timebase gets longer (the UART goes from a steady 115Kbps to about 106Kbps, PWMs are longer by the same amount, ...). This causes mayhem on the board.

I've reproduced the problem on 3 different boards, and also tried replacing the LPC1752 with a 1751 I had in spare and ran into the exact same problem.
I've checked the crystal clock and I am positive the quartz is not causing this.

The datasheet seems to say the component is rated +85°C so I'm very surprised by this. If I was to guess, I'd say the PLL in the IC is the root cause.

Is this a known problem? Has anyone experienced this? Does anyone have an idea for a workaround?

Thanks a lot, any help would be appreciated!