This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LMK04828: LMK04832

Part Number: LMK04828
Other Parts Discussed in Thread: LMK04832

Hello.

In my system, I am driving several LMK04828 Devices using a single LMK04832 in JESD204b configuration.

I supply a 1.2GHz, LVPECL1.6 clock from DCLKOUTx of the LMK04832 to clkin1 of the LMK04828,

and a single sysref pulse from SDCLKOUTx (of the LMK04832 ) to clkin0 of the LMK04828. (also using LVPECL1.6)

The 1.2GHz clock is generated using a 10MHz reference and a 100MHz TCXO, and it's second phase detector is running on 100MHz. 

The second PLL of the LMK04832 generates 100MHz spurious signals around the 1.2GHz carrier that are attenuated  -80dbc.

next, the clkin1 1.2GHz signal in the LMK04828 is directly buffered and sent to an ADC.

when I measure the output of this buffered clock, I see that the 100MHz spurious signals are now -60dbc and not -80dBc from the carrier, and this impacts my ADC performance.

I have tried to use the doubler and send these spurious signals to 200MHz or even 300MHz away, but in either case, this may impact my system's performance.

I forgot to mention that all of the clock signals are routed deferentially, and this situation was repeated using the LMK's EVBs.

Is there a way to cause the LMK04828 to suppress these signals? ( my intention was to use it only for clock division and distribution). if it may help, I can replace it with an LMK04832 as well.

Thank you for helping during the Corona crisis

Itamar Mazaki

  • Hello Itamar,

    I have two thoughts:

    1. The combination of many outputs with 100MHz spurious content is combining to generate a bigger 100MHz spur. This could be tested by disabling all but one outputs using the CLKoutX_Y_PD bits, and seeing if the spurious performance improves at all. If the 100MHz spur is a combination formed by multiple outputs, there isn't much that can be done to reduce these spurs except adding more spacing between the outputs (and possibly leaving some disabled).
    2. There could be a shared power supply coupling spurious energy from the LMK04832 TCXO (sounds like it's a VCTCXO and you're using it as the oscillator for PLL1 of LMK04832, correct?) and adjusting the path of the TCXO or the LMK04832 supply current may help somewhat. However, if the situation is repeatable using the EVBs, it seems odd that they could couple through power supplies (especially if the EVBs run on separate supplies).

    A bandpass filter on the 1.2GHz outputs from the LMK04832 could probably reduce the spurious levels seen by the LMK04828.

    Regards,