TRF3722: Output carrier phase jumps and spurious

It looks like the device is may not be locked properly. The desired Fout is 400 MHz. The marker is at about 401 MHz. I suppose that could be reference error or spectrum analyzer resolution, but it may be the device trying to push the VCO to the desired frequency and not being able to reach it. The Vtune voltage should be a constant around 1.1V. If it is below 0.7V or above 2.2V then the device is not locked, regardless of LD voltage.

Based on your set-up parameters I assume that you want to operate in integer mode. Verify the device is selected to operate in integer mode. Check to see if other frequencies have the same problem by programming to, say, 500 MHz and 350 MHz. Verify the reference is truly 20 MHz without any of the spurious. Any spurious on the reference will transfer to the output.

--RJH

Hello,
Thanks for reply RJH

the output frequency is 401 MHz I have typed wrong. I have seen large phase jumps in output. Control voltage is a constant between 1.1 to 1.65 V , changes after each turn on.

phase jump increase with increasing output frequency.

Refernce has no spurs I have checked.

When you initiate a auto-calibration the device searches for the proper cap array setting to achieve lock near the VCO Cal Ref target.  It is very conceivable that each power cycle or even each auto-cal that the device will find a slightly different cap array setting.  This will result in a slightly different Vtune voltage and corresponding phase delta compared to previous setting.  You can read back the VCO selection and cap array setting to verify/view the cap array settings are shifting.  Also note, that it is possible the desired fundamental frequency is located near a VCO transition point and each auto-cal could put you on a different VCO selection with significantly different cap array settings.  At 400 MHz, the fundamental will be 3200 MHz. I cannot recall off hand whether that is a transition point but it is easy to tell if you can readback the VCO and cap array settings.

Once you know the proper VCO selection and range of the cap array setting you can opt to program those values directly and not go through the auto-cal.  This will give you confidence that the settings are not changing on every power up.  If you are using the TI EVM this will be easy to do with the GUI.  On your board it may be more difficult if the readback functions are not implemented.

You are reporting that the phase jump increases with output frequency.  I do not completely follow this issue as programming different frequencies will result in a phase jump.  Once settled it will locked to the reference.

I presume that the 2.9 kHz spurs are still present.  If you set up the same way do you see any issue on the TI EVM?  I am curious if there is an inadvertent programming register that is causing some problem.  Your report of stable Vtune, LD high, and no spurious on the Reference seems to indicate that the device is locked.  I cannot think of a mechanism that results in 2.9 KHz spurs.  That is why I am curious if different frequencies exhibit this issue and if the TI EVM has this issue.

--RJH