EN6337QA SPICE model

[CEHymowitz]

Do you have a SPICE model for the EN6337QA? I am not interested in the spreadsheet as that is not acceptable for Worst case circuit analysis.

[AndrewG_Intel]

Hello @CEHymowitz

Thank you for posting on the Intel® communities.

We understand that you have some inquiries regarding Intel® FPGA part # EN6337QA.

We have a forum for those specific products and questions so we are moving it to the Programmable Devices Forum so it can get answered more quickly.

Best regards,

Andrew G.

Intel Customer Support Technician

[Mostafa_Intel_AE]

Hello @CEHymowitz ,

Attached please find the LTSPICE model for EN6337QA.

Regards,

Mostafa

[CEHymowitz]

Thank you but this is encrypted so I don’t think we can use this model. We need to tolerance it and we need it to be accurate

Also, we have already made a correlated model ourselves. Attached is our test data and correlations.

Is your model as accurate as ours?

Do you have any correlation data?

[CEHymowitz]

Does you model to AC output impedance/frequency domain? Transient models are no where near as useful as average/AC models for the things that designers really care about (stability, PDN/power integrity, ripple)

[Mostafa_Intel_AE]

Hi @CEHymowitz ,

I am sorry,  i can't share un-encrypted model due to the model related confidential information.

I will create the AC model for you, but need some time.

Whatever, the LTSPICE model and the stability tool are accurate to represent EN6337, but the stability tool presented bode plot maybe is not same as what you measure where the present bode plot at the nominal case but when you click the compensate button the tool will check all the other corner case.

 I have enabled the Zout_closed loop model on the attached stability tool, please check IND_Ripple tab.

Please note that EN6337 has a soft inductor and the inductance value will change based on the loading current.

So the LTSPICE model is very good to model the DC analysis and the stability tool will help to adjust the loop to make sure the VR is stable.

Regards,

Mostafa

[CEHymowitz]

there are so many reasons why this is an unacceptable solution. I am happy to talk about this on the phone but these problems are indicative of everything wrong for virtually all power IC suppliers. So i am not picking on you, but this is really bad.

1) the model is junk. see below. If the output impedance is not right NOTHING about the closed loop response will be correct (like step load). The model is not close as you can see by the impedance in the XLS tool. I am not sure how someone is supposed to know not to use this model in AC or transient? You should withdraw it. 

2) it is completely unacceptable to evaluate stability in a separate tool like excel. Stability is dependent on the load impedance (not just the local bulk). you tool will not get the real stability correct and doesn't allow me to assess the PDN response which is what is important.

3) a model that exhibits the correct output impedance is essential to assessing power integrity plane impedance resonances/step load response/ripple. PDN design is critical and the VRM must be part of it. 

4) So you need to include tolerances as the variance is large and you must also model the output impedance correctly over load current and output voltage

5) you can't encrypt the model because you can't simulate the Power IC alone. The PCB must also be included otherwise you will get the wrong answer. There are models in other simulators that need to be simulated along with the power IC. The only way to support all simulators and models in other formats is to either make encrypted models in ADS and PSPICE or make an unencrypted model. We have a free template that we have published that you can use to make and publish unencrypted models. LTspice is a useless tool as you can't include the PCB in the simulation.

[Mostafa_Intel_AE]

Hello @CEHymowitz ,

Thanks a lot for your feedback.

As i can notice that the orange model that you compare with is not from the stability tool that i attached which has a closer impedance to your measurement.

As i explained in my previous response, that the stability tool will check all the circuit variation and corner cases when you click the compensate button.

Attached are an encrypted pspice DC (time domain) and AC(frequency domain) models to simulate EN6337 in PSPICE environment.

Thanks,

Mostafa