DC coupling capacitor value in the ZETA converter

Dmil

Intellectual 390 points

Other Parts Discussed in Thread: LM3478

Hello!

slyt372.pdf suggests to calculate Cc value using the output ripple voltage, see p.3, Eq. 10b.

The ZETA is similar to SEPIC. LM3478.pdf, eq(53) suggests to calculate Cdc using the energy balance.

The second approach gives a much lower capacitor value.

Why slyt372.pdf doesn't use the same approach?

over 5 years ago

0 Jeff F over 5 years ago

TI__Guru*** 144000 points

Hi Dmitriy,

The Zeta coupling capacitor sees the same input current ripple as a regular buck converter. To minimize input current ripple and therefore voltage ripple, and minimize EMI, you need a bigger coupling capacitor for the zeta compared to the sepic.

0 Dmil over 5 years ago in reply to Jeff F

Intellectual 390 points

Hi Jeff,

Thank you for your answer.

Are there limitations if coupled inductors are used, like for SEPIC?
See www.eetimes.com/document.asp

0 Jeff F over 5 years ago in reply to Dmil

TI__Guru*** 144000 points

Same benefits of coupled inductor for Zeta as with SEPIC. Only limitation that I am aware of is difficulty in finding an off the shelf coupled inductor that fits your application, as Robert Kollman mentioned in his power tip.

0 Dmil over 5 years ago in reply to Jeff F

Intellectual 390 points

Hi Jeff,
So there is no need to use high leakage inductors? Tightly coupled inductors can be also used without doubts?

0 Jeff F over 5 years ago in reply to Dmil

TI__Guru*** 144000 points

I haven't confirmed but I would assume that the same benefits for loosely coupled inductors for SESEPIC Converter Benefits from Leakage Inductance final.pdfPICs as explained in the attached, also apply to ZETA.

0 Dmil over 5 years ago in reply to Jeff F

Intellectual 390 points

Thank you for the doc.
Yesterday I was trying to simulate the ZETA with tightly coupled inductors and different coupling capacitor values. I haven't noticed the same problems like with SEPIC even if I used a small capacitor value.
I would appreciate if you confirm that behaviour.

0 Jeff F over 5 years ago in reply to Dmil

TI__Guru*** 144000 points

Unfortunately, I can't confirm that. I didn't do more work with ZETA beyond publishing that basic app note. ZETA is more expensive than SEPIC because you have to have such a high voltage rated power FET.

0 Dmil over 5 years ago in reply to Jeff F

Intellectual 390 points

Hi Jeff,

my simulation shows that Cc affects on output current ripple, not for input. The lower Cc the higher output current ripple.

0 Dmil over 5 years ago in reply to Dmil

Intellectual 390 points

also the simulation shows for uncoupled inductors:
Cc affects on output voltage.
Vin = 12V, L1 = L2 = 100uH, Cout = 10uF, F = 250kHz, DC=0.5, Rload = 10 Ohm
Cc = 1uF -> Vout around 12V
Cc = 1nF -> Vout around 9.5V

For coupled inductors (K = 1) Cc affects on output current ripple, not on the output voltage.

So there is a minimum acceptable value of the coupling capacitor.

0 Dmil over 5 years ago in reply to Dmil

Intellectual 390 points

simulation results:
- it is very dangerous to use tightly coupled inductors, like MSD1260 in slyt372.pdf
In this case the transformer just connect the input to the output.
- you should avoid a Cc-Lleakage resonance. slyt372.pdf does not take it into account.
- Cc can be calculated using LM3478.pdf, Eq(53)

Power management

Power management forum

DC coupling capacitor value in the ZETA converter