• State
  • Locked Locked
  • Replies 2 replies
  • Subscribers 22 subscribers
  • Views 586 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.

LMX2572: Manual ramping clarification

Mike Neuman
Prodigy 160 points

Part Number: LMX2572

Two questions on manual ramping:

1) Is there a minimum pulse width required for a RampClk edge to be triggered? I know there's a maximum frequency of RampClk of 250KHz (8.1.7.1), and I'm assuming a non-50% duty cycle is okay. So, if I generate a 10ns pulse on RampClk, will it be detected?

2) When manually ramping, does PLL_N automatically roll to the next value when PLL_NUM exceeds PLL_DEN? So, if I start the PLL_N at 50, and PLL_NUM to 0. Can I ramp only until I've maxed out PLL_NUM? Or will it automatically roll PLL_N to 51 and continue adding to PLL_NUM?

Thanks!

  • 0
    TI__Guru** 120130 points

    Hi Mike,

    1. I think that will work. 

    2. No, PLL_N will not be incremented.  In fact, PLL_N is 32-bit but ramp uses 24-bit only.

  • 0 in reply to Noel Fung
    Prodigy 160 points

    The documentation is fairly silent and confusing on some critical issues here. As an aside, PLL_N is only 19 bits, not 32.

    According to the docs for R98, "RAMP0_INC sets the 2's compliment of the number added to the fractional numerator on every ramp cycle".

    7.3.13 says, "PLL_DEN = 2^32 – 1. The actual denominator value being used in ramping mode is 2 24 " which is pretty confusing.

    The formula for RAMP0_INC is (delta fVCO / fPD) * 2^24, but RAMP0_INC is a 30-bit number.  Since RAMP0_INC is 30 bits long, I can actually ramp by 6 bits more than a full fPD. Do these bits not change PLL_N? Are they just thrown out?

    This confusion also exists for RAMP_LIMIT_HIGH (which is a 32-bit number, but whose formula is the same as RAMP0_INC). Why allow you to specify an extra 8 bits of numerator when it's limited to 2^24 unless there's some sort of rolling?

    Also, this rolling behavior is implied by TICS Pro. With a fPD of 80MHz, I can happily enter a manual ramp of 100MHz with no problem, and it computes the RAMP_INC and RAMP_LIMITs as you'd expect for a rolled value. I assume this actually works--I don't have access to the EVM.

    Thanks for your help!