We're using a STM32L4a6 with an ABRACON 32.768KHz crystal (ABS06-127 which is on listed in AN2867 as a compatible crystal) however we seem to be observing error in to order of +200ppm (not the +/-20ppm given in the crystal specs). Using 1p2F caps and drive capability configured to LOW, tested other drive settings but had no effect (as expected). Oscillator starts and stabilizes within expected time, but the resulting clock seems to run significantly fast compared to everyone other clock source we had compared it to.
I know we can calibrate & trim the RTC to compensate for errors in the LSE, however the error does appear to be x10 what is expected, and we're also using the LSE as a source for other peripherals (LCD, LPTIM, MSI auto cal, etc) so a clock more in line with the crystals specs would be preferred and would also make trimming easier, and we haven't even starting on temperature variations yet.
Would love to hear from anyone who has experienced anything like this or has any ideas that might help solve this problem.
