AnsweredAssumed Answered

PNA-X Tunable Converter Characterization

Question asked by alanpna on Jul 13, 2015
Latest reply on Jul 14, 2015 by daras

Like • Show 0 Likes0
Comment • 3

I am using the N5242A for characterization of a frequency converter. My UUT is wide-band tunable, so I have been creating multiple measurement channels for different tuning configurations (channel 1: RF: 3-4 GHz, LO = 5 GHz, IF = 2-1 GHz; channel 2: RF:4-5 GHz, LO = 6 GHz, IF = 2-1 GHz; etc.). Ideally, I would be able to setup 20-30 frequency bands. I am measuring gain, compression, imd, and nf, some of which will not allow segmented sweeps, or I would do that. I then calibrate all my measurement channels and take data. Calibration takes multiple hours. If I want to go back and add another tune frequency, I have to redo my setup and recalibrate.

This is incredibly inconvenient. I wish there was a way to calibrate "generally" so if necessary, I could retune my setup without having to re-calibrate. Is there any possibility of this feature being added?

PLEASE!
Alan

No one else has this question

Outcomes

3 Replies

daras Jul 13, 2015 5:41 PM
Mark Correct
Correct Answer
because your measurements include Compression, IMD and Noise Figure in addition to the conversion parameters (S11, SC21, SC12, S22), there is no single calset that you can use for all these measurements. However, if you can setup all the different measurement channels that you ultimately require, then there is one calibration process that you can go through to calibrate everything. this is called "Cal All" (you can find more about it _*Here*_). Cal All supports all the different measurement classes that are available (with the exception of VMC) and it works by consolidating all the common steps (and frequencies) of the different measurement-specific calibrations into a single optimized calibration routine.

So with Cal All you will be able to calibrate every measurement that you have already setup and it should take a lot less time since all the common steps for each channel is consolidated into a single calibration routine.

you can also use Cal All to calibrate a setup where you have 1 broadband channel of each type of measurement you want to eventually make and then you can create additional channels of each type after performing the Cal All and use the same calsets to correct the new channels you have added. so doing what you asked for is possible, but there are things that you need to consider (like the right frequency density to minimize interpolation errors and making sure that the attenuator settings for each of the broadband channels can be used for all the specific measurements of the same type that you will need in the future). I can provide more detail of how to do that, but if you can setup all the measurement channels that you need up front, using Cal All to correct everything is very easy and there will be no interpolation errors.
Like • Show 0 Likes0
Actions
- alanpna @ daras on Jul 14, 2015 7:02 AM
  Mark Correct
  Correct Answer
  Yes, Cal All is what I have been using. I setup channels for all measurements and tune frequencies I need, and then perform Cal All. The Cal All can take over an hour to complete, and often, with enough measurement channels (> 10 NFx channels, plus other measurements) Cal All crashes after I have spent all the time going through the process!
  
  I would love to hear about setting up a broadband channel that could be used for multiple setups, but in what sense is it broadband? I basically need to sweep RF AND LO. The way I am currently doing it is setting up discrete LO frequencies and sweeping RF in each measurement channel. I don't see how I could sweep BOTH RF and LO at once.
  
  Thanks!
  Alan
  Like • Show 0 Likes0
  Actions
  - daras @ alanpna on Jul 14, 2015 9:23 AM
    Mark Correct
    Correct Answer
    Alan,
    
    NF calibrations, especially if you are using the power sensor receiver characterization method, take a long time. this is because we measure the Gain-Banwidth product of our noise receiver at every RF frequency in the channel(s). We are working on optimizing that process, however, the fact that the application crashes on you is troublesome and we would like to investigate that further. to do that we would need:
    
    1- the information about your PNAX - model number, options, and FW revision
    2- a PNA state file of your setup (.csa) prior to the start of Cal All
    3- a description of all the settings that you adjust in the Cal All wizard. this includes any changes to the default meas class specific settings, changes to calibration channel settings, power sensor and cal kit selections, whether or not you include LO ports in the calibration, and the steps you take before the crash occurs.
    
    If you are not comfortable putting this information on the public forum, please PM me with your email and I will send you my contact information.
    
    as for the broadband approach, you basically need to setup one channel of each type of measurement that you eventually need with a mixer setup that looks something like this:
    
    !https://dl.dropboxusercontent.com/u/87949221/Forum/ThreadID_39926/BraodbandMixer.png!
    
    for example, if you create an SMC, GCX, IMDX, and NFX channel with the above mixer setup and calibrate that setup with Cal All and then save the resulting calsets as "User Calsets", then you can create any number of SMC, GCX, IMDX, and NFX channel where there RF and IF frequencies fall anywhere in the 500 MHz to 30 GHz range shown in the mixer setup above. the one main drawback to this approach is that you will not be able to calibrate the LO power with this setup and you will have to perform a separate source power cal on the LO port for each channel that requires an LO cal. The other caveat is that using this broadband approach means the correction applied to each individual channel will be interpolated. you will have to put enough points in the master broadband channels to create a dense enough frequency spacing to minimize interpolation errors (especially if you are operating at higher frequencies - typically anything above 20 GHz).
    Like • Show 0 Likes0
    Actions

PNA-X Tunable Converter Characterization

Outcomes

Related Content

Recommended Content