LMX2572LP: Phase Modulation

Gordon McKenzie

Part Number: LMX2572LP

Hello,

I'm starting on a design that needs a PLL and some form of phase modulation. On the initial scan of datasheets, the LMX2572LP looks like it might be the best option if it can do both of these together. I realise the device is targetted more at FSK than PSK, but I'm hoping that if it can do one it can can also do the other.

Section 7.3.12 of the datasheet seems to be the relevant section, but from this I'm not quite sure if it will function as I require. I want to do phase shifts in both directions, but it looks like it might only be possible to addvance the phase whereas I need both advance and retard. If I told it to advance 350 degrees in one go, would it try to add 350 degrees (almost a full cycle) or would it go -10 degrees? I also need to limit how quickly the phase changes - if I set the loop filter bandwidth sufficiently low, would this control the phase rate of change and prevent any almost-immediate step change?

If this isn't workable, I guess the alternative is to ramp the frequency one way then the other and get the phase shift as the integral under the ramp. This seems prone to errors though, since I wouldn't be controlling phase directly.

Any suggestions as to whether or not this is workable and, if so, the best way to do it?

Thanks,

Gordon.

over 5 years ago

0 Shawn Han over 5 years ago

TI__Genius 9920 points

Had forwarded it to the team in US. Maybe reply your later because of holidays.

0 Noel Fung over 5 years ago in reply to Shawn Han

TI__Guru** 107110 points

Hi Gordon,

The phase shift goes in forward direction, you can make it move forward 10deg then move forward 340deg (looks like -10deg).

The phase shift effective time is proportional to the fractional DEN value. If you want the response to be immediately, make DEN smaller.

0 Gordon McKenzie over 5 years ago in reply to Noel Fung

Intellectual 610 points

Hi Noel,

Thanks for that. That's a shame there isn't the same control of negative phase shift that there is for positive shifts. If I did a shift of 350° and had DEN set to a small value, would the output signal shift +350° or -10°, i.e. would it add the extra unwanted cycle or not? It might seem that it makes no difference, but if it quickly adds an extra cycle when I don't want it to this breaches one of my other requirements.

I'll be using an output divider of 8, would that make it more forgiving in terms of hiding the extra cycle than if I was using a smaller divider value?

Thanks,

Gordon.

0 Noel Fung over 5 years ago in reply to Gordon McKenzie

TI__Guru** 107110 points

Hi Gordon,

If you start from 0deg and has it shifted 350deg, it will shift 350deg, but the next rising edge will look like -10deg.

0 Gordon McKenzie over 5 years ago in reply to Noel Fung

Intellectual 610 points

Hi Noel,

from what you say, it looks like it should be possible to get something to work so will try it out on an eval board. Thanks for your help with this.

Gordon.

Clock & timing

Clock & timing forum

LMX2572LP: Phase Modulation