Closed hartytp closed 5 years ago
@marmeladapk I looked at the output of the HMC830 on my board on a fast oscilloscope. It seems locked, but sometimes it jumps to a different frequency. I'll post a video in a second. But, if that is happening on other boards it could explain some of the issues we've seen!
NB I'm running my HMC830 output at 2.4GHz not 1.2GHz
Probing on the UFL connectors.
It was at 2.4GHz at first, but then it jumped and is now wandering between 2.9GHz and 3GHz.
https://drive.google.com/open?id=1OgUiDcEbpp-GNcApitAWPpE_TnNefFv1
The signal at the reference input looks a bit weaker than I expected (only about 400mVpp)
Okay, another observation: the UART reports that the HMC830 is locked but when I trigger a scope from the HMC830 reference input and look at the RF output on J59 it's pretty clear that it isn't locking even transiently.
I don't have a fast enough probe to look at the HMC830 output directly so it's possible that the 3GHz RF I'm seeing is actually just a clock buffer free-oscillating.
@marmeladapk is there any way you can track down the board I sent you and post it to me? That board didn't have any of these issues, and this is just killing me
@hartytp I observed output with spectrum analyser and to make sure it locks, I changed the input frequency by a few tens of Hz to observe if output follows it
@gkasprow
Cool. Poked around a bit to find what was oscillating and it started working! Whoo!
Anyway, @gkasprow one thing I still find concerning is that it is clearly possible for the HMC830 to indicate lock when it's not in fact locked. That's a pain and important to know