Using PLL’s Phase Detector as a Linear Phase Meter

Hi,

I’m exploring circuits that could make a Linear Phase Meter. While I’d like to operate from 10Hz to 10MHz with 0.1 degree resolution - I’m willing to see what I can get – with standard parts. I’m expecting lower resolution and accuracy at the frequency extremes though.

 

I’ve prototyped a number of solutions including Balanced Modulators and Analogue Multipliers. The AD8302 is very non-linear around 0 degrees.

I’ve also been looking at Phase Frequency Detectors and Charge Pumps – predominantly what’s inside NXP’s 74HCT9046 PLL. I am getting very strange results that don’t make any sense. While I may have got something wrong, I’m also wondering whether parts like this are just not designed for such an application.  I have never seen circuits showing these parts used in such a way. Mainly I see theoretical documents.

I’ve discovered the AD9901 Phase Frequency Discriminator (still need the Charge Balance circuit) which looks promising and realize that Analog Devices may have something more suitable that might work. Might there be a part I should consider?

Is using the Phase Detector and Charge Balance sections from a PLL going to work as a Linear Phase Meter – or are these sections designed to keep the PLL functioning? i.e. don’t use them open loop?

Thanks

Peter

Parents
  • Hi,

    A few people have suggested that PLL ICs have "tricks" inside them - to ensure that they do perform as good PLLs - rather than anything else.

    The ADF4002 is an interesting part regardless.

    I'm not sure whether the data sheet suggests that the ADF4002 won't perform well below 5MHz - unless the slew rate is high. I'd be driving it from a fast comparator (AD8611) and want to get down to 20Hz. It is not clear where the lower limit is though - all the graphs stop about 100kHz.

    ---------------------- 

    You said: "The fact that there is an in-built delay to ensure there is no dead zone may actually not be good for this particular application".

    I've noticed many "papers" (theoretical documents) on the web, that don't mention an "anti-backlash delay" in the Phase Frequency Detectors. Wouldn't the majority of "real" PFD parts have such a feature - or at least an equivalent solution?

    Thanks

    Peter

Reply
  • Hi,

    A few people have suggested that PLL ICs have "tricks" inside them - to ensure that they do perform as good PLLs - rather than anything else.

    The ADF4002 is an interesting part regardless.

    I'm not sure whether the data sheet suggests that the ADF4002 won't perform well below 5MHz - unless the slew rate is high. I'd be driving it from a fast comparator (AD8611) and want to get down to 20Hz. It is not clear where the lower limit is though - all the graphs stop about 100kHz.

    ---------------------- 

    You said: "The fact that there is an in-built delay to ensure there is no dead zone may actually not be good for this particular application".

    I've noticed many "papers" (theoretical documents) on the web, that don't mention an "anti-backlash delay" in the Phase Frequency Detectors. Wouldn't the majority of "real" PFD parts have such a feature - or at least an equivalent solution?

    Thanks

    Peter

Children
No Data