Hello,
Our product's production line is on hold due to an anomaly discovered with the external 32kHz oscillator startup on the CC2541. We don't understand the reason for this and are looking for some insight whether this could be trouble for CC2541 startup in mass production, and hopefully understand the root cause.
The oscillator starts at approximately 5 KHz, then it slowly transistions to 32.768 KHz after 200 to 300ms. During the transistion period the duty cycle is chaotic before oscillation finally stabilizes. We have not seen this behavior before on MSP430 nor on other processors we've used.
A scope image is attached showing the startup period on the upper window. The lower window shows a zoomed area near the right side of the waveform were the oscillator transistions from 5 KHz to 32KHz. This capture was made with a scope probe on the oscillator drive pin. We verified the same behavior occurs when probing only the digitally buffered signal.
We've tried 3 quality crystals, each from a different manufacturer and compliant with data sheet for CC2541. They all exhibit similar behavior.