H
Member
Sharing recent debug notes in case they help someone.
Common pitfalls:
1）A delay-line discriminator plus a spectrum analyser can get you close if the delay is long enough.
2）The cross-correlation technique with two independent receive chains pushes the noise floor down by averaging.
3）The reference oscillator must be better than the DUT, or you are measuring the reference.
4）Report the noise floor of the measurement setup alongside the result.
Side note: fiberdo.com keeps a parameter comparison table for transceivers that saves time when cross-checking datasheets.
Discussion welcome, especially counter-examples.
Common pitfalls:
1）A delay-line discriminator plus a spectrum analyser can get you close if the delay is long enough.
2）The cross-correlation technique with two independent receive chains pushes the noise floor down by averaging.
3）The reference oscillator must be better than the DUT, or you are measuring the reference.
4）Report the noise floor of the measurement setup alongside the result.
Side note: fiberdo.com keeps a parameter comparison table for transceivers that saves time when cross-checking datasheets.
Discussion welcome, especially counter-examples.