LM5158: Misleading soft start computation

Hi Clement,

Thanks for using the E2E forum.

I will come back to you with the answer after consulting the team.

Best Regards,

Hassan 

Hi Clement,

The excel calculator will calculate the SS time from 20V to 45V. You need to select 20V as an input in the calculator. 

Kindly, share your excel calculator with the selected parameters. Thanks

Best Regards,

Hassan

Hassan,

I am not sure to understand you, do you mean that if I put 28V as input voltage, the soft start computation is from 28V to 45V (and not from 0V to 45V) ?

This would somewhat makes sense considering the diode.

Attached is my file.

P45V LM5157-58_Excel_Quickstart_Calculator_for_Boost_Converter_Design_V1_0_0.xlsx

Best regards,

Clément

Hi Clement,

You have selected very big output cap 750uF. This cap value is causing the quick start calculator to recommend you a very large cap for SS.

Moreover, SS time is influence by the minimum input voltage. Check page 17.

https://www.ti.com/lit/ds/symlink/lm5158.pdf?ts=1743758113139&ref_url=https%253A%252F%252Fwww.ti.com%252Fproduct%252FLM5158

Best Regards,

Hassan 

Hi Hassan,

I am not sure we are fully understanding ourselves.

1/ I told you we have a big capactive load on this power supply because we need to keep it above 15V during at least 50ms when the power supply input is lost. This value corresponds to a 10uF at least just right on the output of the LM5158, a big 680uF to serve as energy tank, and few other caps on the input of the secondary power supplies ahead.

2/ I put that value in the file because you told me in another thread that I needed to put in the file the whole capacitive load on that supply to compute the compensation.

3/ With my first post, my question was raising a question related to the fact that I think the computation file considers, for the Css suggestion, a Vout that would be raising from 0V to the Vout. While it's not the case for me considering the fact that I enable the boost after some time and thus as the input power supply is already stable, the Vout is already at that power supply value and the raise is not from 0V. But I am not 100% sure as the computations are hidden and the datasheet does not state how to dimension the Css to avoid triggering the hiccup mode overload at startup due to the big capacitor.

I hope it's more clear and we can end up with a solution to my question.

EDIT:

Ok my bad the formula is in the app note "How to Design a Boost Converter Using LM5157x, LM5158x".

In there it's using Vload, so correct me if I am wrong but in our assumption, the capacitor will already be loaded to the initial voltage on the input and thus the Vload in the formula when starting the boost will be Vload - Vin ?

Clément

Hi Clement,

Sorry, for the misunderstanding.

The formula given in the app note will calculate the required Css depending on the Cout, but your SS time depends on the input voltage.

After calculating the required Css, you can calculate the actual Tss with the following formula.

tss = Css / Iss * (1- Vin/Vload)

In your case, the recommended Css is 1.125uF at Cout of 680uF and Vload =45 and Iload =0.3A. Whereas, the actual Tss is 42.5ms because your Vin = 28V and Vload = 45V. 

Best Regards,

Hassan 

Hassan,

I am not fully aligned with your answer.

The soft-start capacitor recommandation is supposed to be there to avoid triggering the component temperature protection with the overshoot induced by the load upon start-up of the component.

So if the load is already at 20V, the overshoot will be lower/last longer so the time required for that temperature protection triggering avoidance should obviously be smaller.

Best regards,

Clément