• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 37 subscribers
  • Views 1638 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

Operating Point not found: AC Transfer Characteristic

Saqib Yaqub1
Prodigy 60 points
Other Parts Discussed in Thread: TPS5450, TINA-TI, TPS54140

Hi 

I want to estimate the Phase and Gain Margins of the simple application circuit mentioned in TPS5450 datasheet. I use the method described in the below link

www.linear.com/.../4673

However, if I inject the sine wave as mentioned on the website, I cant run the AC analysis. I always get an error "Operating point not found".  I tried different .PRM files from your library, but no success. I also add a shunt resistor in parallel to VG1, but still fails.

Can you please help take a look if any change need to be made and if that change wont change the loop characteristics I am trying to measure.

Attached are the .TSC file and .TSM files (for TPS5450 and BC540)

To_TINA.zip

  • 0
    TI__Mastermind 41310 points

    Saqib,

    There are a lot of things working against you here. First, the recommended method is running transient simulations at different frequencies of injection, not an AC analysis.

    Next, TINA-TI cannot use a frequency parameter, so you would have to run the simulations for each of the frequencies desired individually. You currently have the circuit set up to run 1 frequency (1Khz). Next, you would need to record the data for each frequency point and then re-assemble the data and plot it using some external plotting tool for the values versus frequency.

    We do not recommend this method for TINA-TI.

    It would be much easier to simply design your circuit in WEBENCH® using the TPS5450 from the product folder, provide your simple inputs and required output voltage and current and use the simulation capability there (or just read off the cross over frequency and phase margin from the OpVal table). You can change the design and BOM values and re-run to see the impact of your choices on the Bode Plot or on the values in the OpVal table.

  • 0 in reply to Britt Brooks - Online Design Tools
    Prodigy 60 points

    Hi 

    Thanks for your feedback. I will try WEBENCH, but I was trying understand the concept for measurement of Closed loop response with a controller in the loop. Is it not possible by any way to use TINA-TI to do it? 

    I have made the circuitsame as WEBENCH and I am able to run AC transfer characteristic on it. However, the Gain and Phase plots look totally different (I guess TINA-TI is plotting Vo/VGS and Vi/VGS. I want to plot 20*log*(Vo(s)/Vi(s)) , but the log function doesn't work in Pos Processor.

    Any suggestions?

    b340A.TSM

    INA826_2.TSC

  • 0 in reply to Saqib Yaqub1
    TI__Mastermind 41310 points

    Saqib,

    You are trying to use a transient model to do an AC analysis. If you wanted to look at the frequency response (AC simulation or Bode Plot), you would need an average model for that purpose. This is not a limitation of TINA-TI or any other simulator. It is a limitation of the model. The control loop in the transient model is closed and you cannot break it externally to do your AC injection that you need for the Bode Plot.

    In WEBENCH®, there is a transfer function of the control loop and power stage that can be calculated and evaluated at frequency points so you can get the required information in the OpVal table. You can also simulate the Bode Plot in the simulation page. If there was an average model for the TPS5450 for TINA-TI in the product folder, you could do the same thing in TINA-TI.

    The transient models for the SMPS devices were not made to be used in an AC simulation. A different type of model is required. Please do not use the transient model for an AC simulation.

    The log function will not work because the DC source on the input is 0V (shorted) for an AC simulation. You would basically be trying to divide a very small number by 0. Take a look at the TPS54140 TINA-TI average reference design for a little more information about what an average model looks like. You will see that only pins that effect the loop are used  and the inductance and frequency are passed as parameters.

    You can also look at the work Basso has done for representing the SMPS in an average model for AC analysis.