Post Go back to editing

ADF4159 Sawtooth

Hi team,

I trying to develop a FMCW Radar application using the ADF4159 PLL. Using this PLL, I'm generating a sawtooth waveform which looks like below image(taken from the ADIsimPLL software when I was analyzing whether the ramp generation is meeting my specifications are not)

As you can see, when ramp reached the final step it fell rapidly but the next ramp is starting from somewhere 9.375GHz instead of desired start freq., which is 9.385GHz. Though surprisingly, the ramp is covering a range of 30MHz and ending at 9.415GHz, which is a desired specification.

Using the spectrum analyzer, the sawtooth ramp generation from PLL looked like below image 

which I believe is not suitable in a Linear FMCW Radar application. 

My questions are as follows

  1. Is it possible to eliminate the frequency component (highlighted in the above image)?
  2. If the answer is no for the above question, does that frequency component has any affect on target range detection?
  3. Looking at 2nd image, what are your thoughts on using that signal in FMCW radar application? Is it a good practice? 
  4. If possible, please enlighten on Linear FMCW and how can one use ADF4159 PLL is such Radar application?

Thanks in advance.

*2nd image not suitable in a Linear FMCW radar application.
[edited by: geeth_anurag_uts at 3:23 PM (GMT -4) on 31 Mar 2022]
  • Hi, 

    Some undershoot is generally acceptable, you can adjust the charge pump current, the filter loop bandwidth and the number of steps and the ramp speed to try to minimise it, but note that this energy might be pushed back into the downramp resulting in unlinearity which may not be acceptable depending on the exact application. I have attached a simulation of an example ramp based on the loop filter we use on our TinyRad FMCW radar kit (it uses ADF4159 and ADF5901) - you can see here we also have some undershoot and the system still has good radar performance

    For your measurement the undershoot does seem to be affecting the start of the ramp, this might cause issues with the radar system operation or issues reaching compliance with emission regulations. The first step would be to try and adjust CP current in PLL to increase LBW and see if that helps. You may need to use a slower ramp also. Below I have an example of ramping with the ADF4159 and ADF5901.

    I also notice you are using an active loop filter. One thing to note is that the GBP of the op amp you are using should be at least 10*LBW of the filter. So assuming OP184 is used the max loop filter BW you could use is ~400kHz. The ADA4084 is a pin compatible alternative with more bandwidth.



  • Thank you Mr. Andrews, for the response. I appreciate it.

    Just another question, as MUXOUT signal is representing the ramp complete status, can I feed it to a RF switch(say, ADRF5032) to overcome the undershoots?