TPS923652: FAQ to calculate Comp network outdated?

Ah i just found gM(ea) , its in section 6.5
And in section 8.2.1.2.5 you find a reference to the RDAMP Component. Maybe you should update the faq.

Hi Markus,

Thanks for your good suggestion! We will check and make it more clear. Thank you!

BRs

Lucia

I found another area that is not quite clear:
On page 22, the input capacitance is calculated. But the Text and the Equation don't fit together.

The text says: Use Equation 6 to calculate the input ripple voltage, where ESRCIN is the ESR of input capacitor, and KDR is the derating coefficient of ceramic capacitance at the applied DC voltage.

But Equation 6 doesnt use ESR or KDR.

Equation 16 on Page 27, for the Buck-Boost design does seem to fit. So is Equation 6 wrong?

Also if i redo the calculation for Equation 6 with the values in the example, i get 1uF as a result not 33uF.

And in both sample designs you are using electrolytic capacitors, which makes sense. But why is there a KDR factor in the formula, i believe this irrelevant for electrolytic capacitors?

Hi,

Our expert will update you next week

Thanks, really looking forward to an answer, thats the only open question in my current project.

Hi Markus,

Thanks for your good findings! I believe it will be a typo in the datasheet. Actually, we usually don't consider the KDR factor for electrolytic capacitor. But for ceramic capacitor, it should be considered. And yes, you can find both electrolytic capacitor and ceramic capacitor as the input cap with EVM board. Thank you!

BRs

Lucia

Hi Lucia,

thanks, but i still dont get it. I understand of course, that there is a mix of electrolytic and ceramic capacitors. But i can't understand the calculations.

On page 22, Equation 6 is:
VIN ripple = IL ripple / (8 × CIN × fSW)

Changed, to get CIN it is:
CIN = IL ripple / (8 × VIN ripple × fSW)

The values in the sample design are:
IL ripple:1,6A
VIN Ripple=0,4V
fsw=500.000

CIn = 1,6/(8*0,4*500.000)=0,000001F= 1uF

The next line in the datasheet says:In this design, a 33µF, 25V electrolytic capacitor, a 1µF, 25V X7R ceramic capacitor and a 0.1µF, 100V X7R ceramic capacitor are chosen, yielding around 400mV input ripple voltage.

Why?! How do you go from 1uF to 34,01uf? Specially because the largest part is the electrolytic, without any DC Bias derating. I would understand the choice of capacitors, if the calculation gives 30uF for example. But 34 times higher, without any explanation?

I need to find the right input caps for my design with
IL ripple:0,75A
VIN Ripple=0,4V
fsw=1.200.000

if i do the calc, i get 195nF, and this can't be true.

Dear Markus,

Pls kindly let me check your question and get back to you soon. Thank you!

BRs

Lucia

Any news?

Hi Markus,

Sorry for late due to my bandwidth limitation and really thanks for your patience. I believe it's a typo. And just as you said, there is almost no derating effect for the electrolytic capacitor but we need to consider the derating effect with ceramic capacitor. Thanks for let us know! 

Thank you!

BRs

Lucia

Hi Lucia.

This doesn’t answer the question.

How do i calculate the input capacitance?

Please check the calculations on page 22 and tell me if you get 33uf (or close) as a result.

I can‘t wait any longer, so i used the „standard“ equation for input capacitance for boost converters:

Cin=(Iin*D)/(Vin Ripple*fsw)

But please check this formula, either i completely missed something, or its just misleading and wrong.

Thanks,

Markus

Hi Markus,

Let me check the Boost version input capacitor equation and I will get back to you by today. Thank you for your patience!

BRs

Lucia

Thanks a lot!

Hi Markus,

Your welcome. Have a nice day!

BRs

Lucia