Different results for FRA using LTpowerCAD and LTspice, Specifically LT8643S and LT8390

Hi Levandovsky ,

You can find more information about the FRA feature using the help file. Just type in FRA and all related topic will show up.

FRA is a new feature of LTspice and thus app notes about this feature are still limited.

with regards to the difference between LTpowerCAD and LTspice result. It depends, LTpowerCAD is limited to CCM mode of operation. also, it can be likely that there is a bug on the LTspice model.

What I can do on the other hand is to consult the part owner and let them answer the question for you.

Thank you.

Regards,

PaulDaria 

Hi Levandovsky ,

You can find more information about the FRA feature using the help file. Just type in FRA and all related topic will show up.

FRA is a new feature of LTspice and thus app notes about this feature are still limited.

with regards to the difference between LTpowerCAD and LTspice result. It depends, LTpowerCAD is limited to CCM mode of operation. also, it can be likely that there is a bug on the LTspice model.

What I can do on the other hand is to consult the part owner and let them answer the question for you.

Thank you.

Regards,

PaulDaria 

Hi

I use the same components value and load for LTpowerCAD and LTspice and mode of operation is the same.

I think there must be no significant differences between two designs.  

Probably you right with your assumption about problem with LT8643S model.
But what model is wrong, LTpowerCAD or LTspice?

I more\less successfully tested LT8608S circuit. BW is 14.13 kHz for LTpowerCAD and 12.94 kHz for LTspice.

LT8608S_FRA.zip

Also I test my LT8390 design and have also different results...

More global question is

LTpowerCAD vs. LTspice. Which of them can be trusted?

Hi Levandovsky ,

Someone is already working with LT8643S Bode plot mismatch issue of LTspice and LTpowerCAD. With regard to which to trust better, I would suggest LTpowerCAD,. Simply because it's came in first and have already higher reliability over the LTspice FRA tool. We are still figuring out what might have caused the plot mismatch.

Thanks.

Regards,

PaulDaria 

Hi

The same problem with LT8390....

Hi Levandovsky,

Kindly attach your design circuits. Our internal team is investigating this matter.


As for the documentation of the FRA tool, you may also try checking the HELP Tool.

LT8390_FRA.zip

Hi Levandovsky ,

Thank you for providing these files, we will investigate the cause. We already have a few hints what might be missing. But to give you a conclusion, it is likely that the LTspice model is missing something and must be fixed so what I said earlier about who to trust still holds true which is LTpowerCAD.

A fix on the LTspice model of LT8643S is coming out soon. We will let you know once it becomes available.

We appreciate your help in ensuring the highest quality of our LTspice models.

Regards,

 PaulDaria  

Hi PaulDaria

Any news?

Hi Levandovsky ,

We are on some finishing work to update the LT8643S model.
We are now able to match the LTpowerCAD gain/phase plot.



We will also be updating the models for the LT8608S and LT8390.

For your question:LTpowerCAD vs. LTspice. Which of them can be trusted?

LTpowerCAD simulation results have been demonstrated to be matching actual hardware test results of various ADI parts.
The new LTspice FRA tool is also a good tool, well-designed to make measurements following the industry-accepted control loop theory.
It just so happened that the model you encountered, the LT8643S has a missing internal parameter that affected its results in the frequency domain analysis. Nonetheless, the LT8643S model is still representing the actual device in time-domain analysis.