• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 2 answers
  • Subscribers 42 subscribers
  • Views 3419 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.

How to make faster simulation using Pspice for TI.

Tevhid ATALIK
Prodigy 50 points
Other Parts Discussed in Thread: LMR50410, LM393LV, OPA2376

Dear Expert

I would like to make a simulation of LMR50410Y3F-Q1 step-down converter. There is no simulation model in Tina or Webench Power Designer so I have to use Pspice for TI. I download simulation model from https://www.ti.com/lit/zip/slum770 and I opened the simulation file.

I change some parameters and I saw the note that " The simulation runs for 2.4ms and takes approximately 60 minutes to run on a 4 core 2.8GHz machine."  I didn't want to believe and I checked simulation parameters. Max step size was set 8ns and I made it 400ns (I removed also and left blank) and I set 2.6ms simulation and it took about 40 minutes. I realized and I made a mistake and I changed one parameter and I started the simulation again and it took another 40 minutes. It is unbelievable. I cannot make the simulation duration short because the IC cannot start to produce output before 1.8ms due to soft-start up feature.

How can I make the simulation faster? I use LTspice and I made similar simulation using LT3502 (it works at 2.2MHz as like LMR50410) and the simulation takes 30 seconds (2.6ms simulation duration).

There is a speed option in the simulation setting window and the default value is 3 and I left it 3. I think the software is so slow or so complex. It may give very very very accurate result but it is so slow and become useless. Who can wait 60 minutes for 2.4ms simulation? I think that if there is a way to make it faster TI didn't write a note that 2.4ms simulation takes 60 minutes.

 

Shall I change Gmin value or what I can do to make it faster as like Ltspice?

Best Regards

 

  • 0
    TI__Expert 8130 points

    Hi Tevhid,

    Thanks for using PSPICE for TI. Unfortunately, the simulation runs as the note states on the file. I have taken note of this review and forwarded to our internal team for feedback. Thanks again for using TI simulation tools.

    Best regards,
    Alejandro

  • 0
    TI__Expert 5880 points

    Hello Tevhid,

    Please find the attached project where I have loosened the tolerances (ABSTOL, RELTOL, ITL2, and Autocovergence enabled). Please be aware that you may have lower accuracy for your results. You can read about the specific simulation settings under Help -> Documentation. On the left side of the help window pop-up, click the menu (3 lines in a circle) to open the reference guides. The attached project runs in ~15 mins, although you are welcome to continue modifying to reduce the time. I know this is not competitive with LTspice's simulation time, but we are actively working to improve the PSpice for TI simulation engine. Future releases of the tool will include improvements to simulation speed.

    Thank you,

    Jackie

    LMR50410Y3F-Q1_TRANS-loosened.zip

  • 0
    Prodigy 50 points

    Dear Jackie

    Thank you for the file. I have already done my simulation with LMR50410 but i will do for after parts selected from TI for example LM393LV or TVL2376 etc. If I spend 40 minutes for every single simulation, Pspice for TI becomes useless. As you said LTSpice is so fast comparing Pspice for TI. I have been used Proteus, Tina, LTspice for a long time and I faced many problem with simulation software but speed is not one of them. 40 minutes is unbelievable. I hope you can reduce simulation time soon because we cannot use Tina any more. When I would like to define new macro in Tina with what TI provides as psice lib, it failes to define.

    https://e2e.ti.com/support/tools/simulation-hardware-system-design-tools-group/sim-hw-system-design/f/simulation-hardware-system-design-tools-forum/1009868/tina-new-macro-definitian-problem-and-pspice-for-ti-existing-model-doesn-t-work

    Therefore we have to use Pspice for TI efectively.

    Note: I made a simulation for OPA2376 in both Tina and Pspice for TI. Pspice for TI has exceptional accuracy which is very good. It works as what datasheet explains. Tina doesn't give the accuracy and manipulate the designer. I did the same simulation and OPA2376 input is 3V while it is powered from 3.3V. Tina output is 3V only. Pspice output is (3V+2.48mV) which is what datasheet says. Pspice for TI is good but some issues need to develop. For example cursors... When a graph is zoomed in cursor disappers. You have to zoom out first and bring the cursor where you are going to zoom in and zoom in again. I couldn't find a way to bring cursor in zoomed view.. maybe there is a way... I made a simple simulation to see input offset behaivour with a fuction of power supply voltage and input voltage and it is 1 ms simulation. It is done in few seconds. It is not complex maybe that is the reason it was completed very fast with exceptional accuracy.

    Best Regards