LMK05318B: LMK05318B shows spurious after change registers

Hi Steve, 

Registers 0x135 - 0x139 correspond to the DPLL REF FB Divider Numerator, which could cause unpredictable behavior if they're changed on the fly. Does the spurious behavior go away after doing a soft reset? 

Regards, 

Connor 

No, they want to change the output frequency (several Hz offset), so they only change the Divider Numerator without change the input reference. But maybe the changed values are not the most proper values compared to the script calculation, and leading to the spurious.

Steve Sun

Hi Steve, 

If they want to change the output frequency in small step sizes, I would recommend using the DCO feature. We have some details about this in Section 9.4.4 of the datasheet, and registers can be configured on the Advanced -> DCO page of the LMK05318B TICS Pro profile. 

Regards, 

Connor 

Hi Connor,

Why I can not change the value of DPLL LBW target in TICS Pro. And I don't find the registers that change this value.

You should be able to update the DPLL LBW Target frequency, then after you click "Run Script" you should see the actual value update to be close to the target value. There are several registers related to the DPLL filter settings. The calculations for setting these registers aren't straightforward which is why the MATLAB script is needed. Let me know if this helps. 

Regards, 

Connor 

Hi Connor,

Since it's not easy to calculate these values, if it's possible to simulate these parameters by PLLATINUMSIM-SW 应用软件和框架 | 德州仪器 TI.com.cn？

Steve

Hi Steve, 

Unfortunately PLLatinumSim can only calculate analog loop filter parameters, and not digital filter parameters. The Matlab script in TICS Pro is the only way we support programming the DPLL filter settings. 

Can you clarify what the customer is trying to accomplish? 

Regards, 

Connor