• State
  • Locked Locked
  • Replies 7 replies
  • Subscribers 21 subscribers
  • Views 563 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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: CLK Output Divider

Erik Jarva
Prodigy 80 points
Part Number: LMK04828
Other Parts Discussed in Thread: AFE7950, AFE7950EVM

Tool/software:

I am looking at the AFE7950EVM reference design and there is a SDCLKOUT3 differential clocking signal going from the LMK04828 to the AFE7950. The signals run through a series 100 ohm resistor and then a 100 ohm resistor to ground before reaching the differential 100 ohms on the AFE7950 SYSREF input.

Presumably this termination scheme is meant to keep the output of the LMK04828 within the allowable voltage range of the AFE7950.

Given the format options for the SDCLK output and the termination chosen the output is likely LVPECL.

Is my assumption about the output format correct? If it is, how can one calculate the common mode voltage at the output of the driver? The termination resistors are spread across the clocking and transceiver schematic pages. Where should they be placed in the layout relative to the driver and receiver chips?

  • 0
    TI__Mastermind 29542 points

    Hi Erik, 
    The appropriate terminations are found on page 105 of DS. 

    The common mode voltage for an LVPECL output is 2V. 

    I just quickly glanced the AFE DS for sysref and saw this input. 

    According to below VCM for sysref input should be 0.6V-0.8V with a 1.8V swing maximum. 

    I found an IBIS model for the SYSREF input of the AFE and used the LMK04828 SYSREFOUT IBIS model to drive it in Hyperlynx (the top signal is a 100MHz LVPECL SYSREFOUT from LMK04828 and you can see that VCM is also 2V)

    The schematic you report seems to indicate the Rx will see a 1V VCM and the swing is around 400mV (bottom traces U4)
    This does not seem to meet the SYSREF input spec from the AFE DS but I am unfamiliar with the AFE devices unfortunately thus I cannot say for certain, so I recommend reaching out to the AFE team. 

    Best Regards, 

    Vicente

  • 0
    Prodigy 80 points

    Hello Vicente,

    Thank you for your help! Do you have a good contact for the AFE team? In the meantime I will crosspost this to the other forum.

    Erik

  • 0 in reply to Erik Jarva
    TI__Mastermind 29542 points

    Hi Erik, 
    Unfortunately, I do not but you can simply make a question and an AE from their team will respond to your question on the thread. 
    Can you share a picture of the schematic you're referencing? I would like to see how the LMK04828 and the AFE part are connected. 

    Best regards, 

    Vicente 

  • 0 in reply to vicente flores prado
    Prodigy 80 points

    Hello Vincente,

    Here is the connection between the LMK04828 and the AFE7950.

    Erik

  • 0 in reply to Erik Jarva
    TI__Mastermind 29542 points

    Hi Erik, 
    It is the same circuit I simulated - I just wanted to ensure I set it up correctly. 
    I see they're other resistors that are not DNI 
    If those are populated the VCM becomes 0.9V assuming, we AC couple the LMK04828 beforehand. 

    It's a bit odd because the DS snippet I sent earlier specified a 100Ohm differential termination which I don't see in this schematic - I would also ask if that's internal to the AFE part or needs to be done externally. Once you have the AFE questions/specs sorted out feel free to follow up so I can assist with the clocking. 

    best regards, 

    Vicente 

  • 0 in reply to vicente flores prado
    Prodigy 80 points

    Hello Vicente,

    The internal differential termination on the SYSREF pins is configurable between 100, 150, and 300 ohms according to the datasheet.

    Erik

  • 0 in reply to Erik Jarva
    TI__Mastermind 29542 points

    Hi Erik, 
    Not sure if that's modeled inside the IBIS model.
    Assuming it's not adding another 100Ohm differential resistor near the AFE would further reduce the swing. 

    Best regards, 

    Vicente