2 questions about XTAL trimming:
1. In the FAQ, it is mentioned that the BTLE spec assumes a +/- 50ppm deviation in which it should still operate. In the same chapter, you advise to reach an accuracy of +/- 15ppm to compensate for future aging, is this including influences by temperature?
2. In the 'AN-B-020_v1.3_End product testing and programming guidelines.pdf’ it is advised to reach an accuracy of +/- 5ppm, this contradicts to the value mentioned in the FAQ. Which one is the correct one, 5 or 15ppm, and does that include influences by temperature?
Thanks,
Jeroen
Hi Jeroen,
it's indeed confusing. Let me explain. See also the latest datasheet v3.0: 16 MHz Crystal Oscillator: the crystal accuracy specification has changed.
When doing 16MHz Xtal oscillator trimming, we now aim for an accuracy (at 25C) of +/- 10 ppm. Better is not needed, and would take more time in production, but is allowed to do of course.
Temperature variation would result in an additional +/- 10 ppm or +/-20, and assuming +/- 5 ppm for aging, we end up at +/- 25 ppm or +/- 35 ppm. This is well in the BLE limits of +/- 50ppm.
When executing the 16MHz crystal oscillator trimming, a 16MHz crystal with an initial accuracy of +/- 40 ppm can be selected. This is inclding aging and temperature drift.
We will have a look at our documents, and adapt the figures where necessary.
Best regards, BB_Dialog.
Thanks for the clarification,
Jeroen