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.

LMK04803:

Part Number: LMK04803
Other Parts Discussed in Thread: PCM1792

Hello,

First of  all, the LMK04803 is a fantastic chip! Lock and holdover work very well to get a consistent signal with very little jitter at the CLKout ports. Even if the reference signal at CLKin has terrible jitter and frequently initiates holdover.

The jitter of the output signal is very low, but unfortunately there are no clean square wave signals. At 2.8224 MHz the signal at CLKOUT_2 is somewhat ok, but with the rising edge at the top and the falling edge at the bottom, it has an obliquely flattened area (please note blue circles). The oblique flattened area extends with increasing frequency, 12.288 MHz. At 24,576MHz, it is an ugly sawtooth signal.

Due to the subsequent circuits, the CLKout ports are operated in LVCMOS mode. The trace of the measured signals CLKOUT_2 lies on the top of the board, the trace lengths are very short and there are no Vias. Therefore, I suspect a faulty termination of the CLKout ports. But you may also know other reasons.

Unfortunately, the LMK04803 datasheet gives no recommendation regarding the termination of the CLKout port in LVCMOS mode and my schematic looks as follows:

Best regards

Andreas Nicklas

  • Hello Andreas,

    For CMOS clocks. I recommend a simple direct connect as the LMK0480x outputs are ~50 ohms for LVCMOS.

    -

    At LMK04803 LVCMOS output:
    The LVCMOS outputs have 50 ohm outputs so no extra Rs is required. Set R6/R7 to 0 ohms.

    R11/R12 1 kohm pulldowns are not necessary and can be removed.

    The series 50 ohms to input is not required. The input for LVCMOS is high impedance, so the 50 ohm series resistance has little effect. This results in full source termination.

    Are you trying to do a passive split of the CMOS output to many inputs?

    73,
    Timothy
  • Hello Timothy,

    many thanks for your feedback.

    Yes I am trying to do a passive split of the LMK04803 LVCMOS output to serve eight PCM1792 in order to provide the Bitclock.

    Best regards
    Andreas
  • Ok Andreas,

    Just to confirm...  The schematic update did give you the improved edge you were looking for?

    73,
    Timothy

  • Hi Timothy,

    I will run the next measurement Thursday this week and will let you know the outcome.

    However your advise was very helpful and a big help so far.

    Best regards
    Andreas