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.

UCC21750: Transient Simulation with imported model in Keysight ADS

Part Number: UCC21750

Hello,

for weeks I have tried to run a transient simulation with the UCC21750 in Keysight ADS, but I am not able to get it running.

First I tried to import the model, but ADS was not able to translate the PSPICE model .MODEL IXGA8N100 NIGBT.

Followed this thread I commented the line. Now I was able to translate the model, but I am either not able to get a DC-convergence or unexpected simulation results.

I already ordered several UCC21750 to get a physical test on the tesbench running (this is working quite good and as expected), but I am still not able to do the simulation.

Do you have any possibility to assist me with your model in Keysight ADS? Or could you offer an example workspace for your model for Keysight ADS?

Attached you can find my ADS workspace with a most basic configuration.

TI_Test_wrk.7z

Thank you very much in advance.

Kind Regards,

Valentin

  • Hello Valentin, Thanks very much for reaching out.

    I don't think the team are trained to support Keysight ADS. I suggest you work with Keisight ADS and debug on the errors which pops up.

    We can surely visit the errors together when it comes to the spice model itself.

    Wei

  • Hello Wei,

    thank you for your feedback.

    I am already in close contact with the Keysight ADS support, but as they do have very little knowledge about the structure of your TI model, it is very difficult to solve the issue on the ADS side without getting any support from TI side.

    The model is running now for some us, but the results are not at all as expected. The current output of the driver is not exceeding some nAs and the voltage output just shows a short spike. (image 1 shows the circuit used for simulation, image 2 shows the simulation results. Blue is the IN+ signal, red the output voltage).

    Is this related to the omission of the IXGA8N100 NIGBT model in the main model, as suggested by

    UCC21710-Q1: Model conversion error - Power management forum - Power management - TI E2E support for...

    e2e.ti.com
    Part Number: UCC21710-Q1 Other Parts Discussed in Thread: UCC21710 Hi, I was trying to convert UCC21710-Q1 into Tina TI model , but I am getting same error of

    I look forward to your reply.

    Kind regards,

    Valentin

  • Valentin,

    We can try to help but it is difficult, we only support the models with TINA and Orcad and we don't have a way to verify your simulation.

    I have one question, how are you relating the Secondary supply VDD/VEE referenced to COM, to the low Side, GND? At least in other simulators, you need to tie them together e.g. by huge resistor or V source / nodebreaker. I can't see this in your schematic.

    best

    dimitri

  • Hello Dimitri,

    thank you for your message.

    I tied them together now via 1GOhm resistor, but the result remains unchanged.

    Kind regards,
    Valentin

  • Valentin,

    Try by VDC source to connect COM to ground.

    Also, are you using the original unaltered model, or did you follow the guide to remove the IGBT references?

  • Hello Dimitri,

    I also tried now to connect COM via a 10V VDC-source to GND. This results in a shift of the output signal and no voltage spike is visible anymore:

    Yes, as written above I removed the IXGA8N100 NIGBT, as the translation of the model was not possible including this NIGBT model.

    Kind regards,
    Valentin

  • Valentin, though i think this is unlikely to be the problem,

    Can you change your PWM to range from -1 to 5V instead of 1 to 4V, i ask this because we had another question where this is the solution, and it seemed to be a nuance with the model.

    Also, please check on RDY/FLT pin voltages as well please.

    Even though it looks like drive must be getting some input to secondary side, lets try this regardless.

    Best

    Dimitri

  • Hello Dimitri,

    thank you for your reply.

    I changed the PWM voltage, but the result remains unchanged.


    I checked RDY/FLT and ENable/Reset Pin:

    They all reach to round about 5V. So I guess they are ok?

    Kind regards,

    Valentin 

  • Valentin, that would be fine.

    I will build your schematic and try to corroborate the issues youre having in PSPICE simulator, but suspect that there just might be an issue with using the models in Keysight ADS.

    In case we can't solve the issue with this model in Keysight ADS, is it possible you could simulate in PSPICE/Orcad ? Do you have access to those tools? Alternately, we have tool like LTSPICE called TINA which is free and our models work without issue in TINA?

    I ask this because our support for unsupported is very limited. We can't debug them in Keysights simulator so since the basic debug steps the most I can really do is ensure that your schematic works OK in the supported simulator and recommend you to use a supported simulator.

    Best

    Dimitri

  • Hello Dimitri,

    thank you for your reply.

    Yes I am also guessing the issue is the converted model. Can you tell me what the IXGA8N100 NIGBT is used for, and if it is maybe possible to substitute/simplify this part?

    Sadly I do not have access to PSPICE/Orcad. Yes, I have used TINA before which is fine, but not  suitable for my application.

    The reason i use ADS is to calculate my parasitics of my PCB to simulate with the final design data.

    Kind regards,
    Valentin

  • Valentin,

    The IGBT is a holdover from the test setup of the model, but is not actually used in simulation.

    I really apologize, but I can't do much more to help you with ADS. Like i said, we dont have this program and we the PSPICE model only supported in Tina / PSPICE (orcad, allegro),


    I am not sure how ADS gives the PCB parasitics, but if it possible to back-annotate it to a different tool this might be the best option. TI will also have a free PSPICE solution (Capture, but with a limit on plots) that will work flawlessly with our models, soon.

    Best

    Dimitri

  • Hello Dimitri,

    thank you for your feedback.


    Ok, so it seems like kind of impossible to use the model in ADS.

    So I will stop trying and skip using the model.

    Kind Regards,

    Valentin

  • Yes, Valentin. And apologies we couldn't provide more support on this!