LMZM33603: Cff for variable output; also output caps

Stephen Fenwick

Part Number: LMZM33603

I want to make the output of the LMZM33603 dynamically-adjustable from about 5V to 12V, with a 24V supply. However, I'm not sure what to do about Cff. Making it dynamically adjustable is not really practical (unless it's simple...).

Can I put in a Cff to support the lowest voltage that I'd want to generate? What will that do to the transient response for the higher voltages that would not otherwise need a Cff?

The nominal output load is about 1.5A.

Also, for Cout, to make the 5V to 12V range work I'm looking at 3x 33uF 25V parts, like the C3216X5R1E336M160AC. What do you think of that?

Thanks!

over 5 years ago

0 Jimmy Hua over 5 years ago

TI__Mastermind 47060 points

Hi Stephen,

Table 1 in the LMZM33603 datasheet provides an optimized Cff value for each output voltage application. The Cff is recommended for optimal application for each output voltage. I'd have to check some simulations to see what the effect of Cff will do on the higher output voltage Vout=12V. Let me get back to you once I get that information.

Also in your application, you would need to design your Cout for 5V since this requires more output capacitance. You'd also need to consider the C3216X5R1E336M160AC DC bias effects of the capacitor at higher output voltages. At 12.5V output, the effective capacitance will be 4.127uF so we can estimate 12Vout will result in an effective capacitance of 4.5uF. This means you will have an effective Cout of 13.5uF for 12Vout which is not enough based on the Table 1. I'd recommend a 4 x 22uF(GRM32ER71E226KE15L) 25V configuration since that will have lower derating.

Regards,

Jimmy

0 Stephen Fenwick over 5 years ago in reply to Jimmy Hua

Intellectual 260 points

Thanks for the prompt reply. I agree about the caps and was headed that way (4x 22uf, 25V) due to the global MLCC parts shortage. I was planning to use an X5R since that seems to be more readily available.

We are trying this with a dev kit with no Cff; we can add the 5V Cff to the 12V circuit to see what happens. I'm okay with some loss in efficiency.

Thanks!

Steve

0 Jimmy Hua over 5 years ago in reply to Stephen Fenwick

TI__Mastermind 47060 points

Hi Steve,

I've done a loop response for different Vout cases given a 100pF Cff. PM is greater tan 45deg which indicates a stable design.

Regards,

Jimmy

0 Stephen Fenwick over 5 years ago in reply to Jimmy Hua

Intellectual 260 points

Thanks! Also, this is only being used to drive 12VDC case fans, so there's really no transients to worry about.

0 Stephen Fenwick over 5 years ago in reply to Jimmy Hua

Intellectual 260 points

BTW, what was the switching frequency for these simulations?

0 Jimmy Hua over 5 years ago in reply to Stephen Fenwick

TI__Mastermind 47060 points

That was taken at 450kHz. I went back and retook the simulation at 500kHz to be consistent with the datasheet and got almost identicalphase margin.

0 Stephen Fenwick over 5 years ago in reply to Jimmy Hua

Intellectual 260 points

Ok. I was planning to run at 900kHz, which is listed as being ok for 2.3V to 13.5V. I'd rather have some margin at the top end in case there's any anomalies with the control mechanism.

0 Jimmy Hua over 5 years ago in reply to Stephen Fenwick

TI__Mastermind 47060 points

Hi Stephen,

For completeness I ran my simulation again at 900kHz and don't see any issues with the loop response. I think it is good to have some margin at the top end for the output capacitor case. Below is a snippet of the simulation results. All output voltage options satisfy PM>45deg.

Regards,

Jimmy

Power management

Power management forum

LMZM33603: Cff for variable output; also output caps