• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Subscribers 49 subscribers
  • Views 1047 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

THS3120 and THS3120EVM

francesco mazzilli
Prodigy 70 points
Other Parts Discussed in Thread: THS3120

Hi,

I am planning to use THS3120 to measure PSRR of my low-drop-out regulator. Moreover, I bought THS3120 EVM and I think they will deliver it to me tomorrow.

However, what I do not understand from http://focus.tij.co.jp/jp/general/docs/lit/getliterature.tsp?literatureNumber=slaa414&fileType=pdf if I need to calibrate

my network analyzer 4395A in order to get correct results. Do you have any hint for this? In case I need a calibration, what type of calibration

I need to perform (e.g. thru, short, open,..) ? 

Concerning calibration, I know that I need to connect both probes to the output of your EVM but what is next? I have only used VNA for S-parameter measurement.

 

 

  • 0
    TI__Expert 5640 points

    We also use the 4395A in our lab. I calibrate using the "thru" setting and this provides good results.  If you are in doubt about the calibration of the 4395A then you can use an oscilloscope to measure VIN and VOUT while sweeping the frequency.  You can then calculate the measured PSRR at different frequencies using the scope and see how it compares to what the 4395A is reporting. 

  • 0 in reply to Scot Lester
    Prodigy 70 points

    Thank you for the suggestion.

    Just one more thing, how do you perform the calibration? You connect the two active probe via a thru? Or you connect the two active probe at the output of your EVM and select thru calibration?

  • 0 in reply to francesco mazzilli
    TI__Expert 5640 points

    First, let me clarify my definition.  The app note you referenced shows an amplifier EVM driving the load (a LDO EVM). I generally connect the two together with the amplifier driving the LDO. The oscillator output of the network analyser is connected to the input of the amplifier EVM.  Then both active probes are connected together at the output of the amplifier, then run the 'thru' calibration.  Having the LDO board connected to the amplifier allows the network analyser calibration to calibrate out any errors due to the impedance of the load versus the drive capability of the oscillator or amplifier. This isn't the text book way to do it but it provides good results and doesn't require any additional equipment.

     

  • 0 in reply to Scot Lester
    Prodigy 70 points

    We got the THS3120 EVM and I guess in order to perform the measurement as stated in the application note http://focus.tij.co.jp/jp/general/docs/lit/getliterature.tsp?literatureNumber=slaa414&fileType=pdf

    we should modify the EVM.

    Is it true? Or I am missing something.

     

  • 0 in reply to francesco mazzilli
    TI__Genius 17905 points

    Yes, you will have to modify the EVM to replicate Scot's setup. The EVM by default is populated as described in the EVM section of the THS3120 datasheet (http://focus.ti.com/lit/ds/slos420e/slos420e.pdf) starting on p.22.

  • 0 in reply to Kris F
    Prodigy 70 points

    OK, we did it and measurements are fine.

    However, I still have one question. In the application note, you mention that s possible to measure PSRR via EVM or oscilloscope. If you do a comparison between results EVM and oscilloscope, then results are close.

    But, let's say we set the output of VNA to 0 dBm, the ac signal to Vdd should be 795 mVpp because not matched to 50 Ohm.

    So, in order to take into account this voltage change, it is better to set the output of VNA to -5 dBm or even lower value depending of the LDO.

    I would like to know if what I say make sense to you or not?

    In the application note, while measuring the PSRR via an oscilloscope, you used a signal generator. How did you handle the problem of the impedance?

  • 0 in reply to Scot Lester
    Prodigy 70 points

    But when you use the oscilloscope to check the measurement of PSRR, and you use an signal generator to sweep frequency. How do you deal with the impedance mismatch?

    Signal generator is also matched to 50 Ohm as VNA, checking with the oscilloscope you read each time twice the voltage you inject due to reflection problem.

  • 0 in reply to francesco mazzilli
    TI__Genius 17905 points

    Francesco,

    I would not worry about impedance mismatch in this measurement, especially at these fairly low frequencies. The main concern here is measuring the peak-to-peak AC signal at the input of the LDO and comparing it to the measured peak-to-peak signal at the LDO output. Yes, depending on the LDO being used, you may want to reduce the applied AC signal so that you don't push the LDO into drop out.