60Hz Ripple on input to SMPS

[Jester]

I'm trying to determine the minimum bulk capacitor value for a 60Hz bridge rectified transformer secondary that feeds a dc to dc converter.

I would like to minimize inrush current, so I want to use the smallest acceptable capacitor. The converter only requires 1W and the input voltage range is 12 to 24V. With a 470uF capacitor the input waveform will look something like this.

Will this input ripple be problematic for the converter or will this not be an issue as long as the input voltage dips are kept above 12V?

 

[D.A.(Tony)Stewart]

how about this?


https://webench.ti.com/webench5/pow...130&O1I=2&O1V=5&op_TA=30&application=SWITCHER

 

[ravindragudi]

If the voltage regulator can withstand the ripple it is getting into its input, then there is no need to modify anything. You need to pay close attention on PSRR for this and the efficiency of regulator - if you are worried about it.
The better approach would be as suggested by SunnySkyguy - you can vary the values of the caps and choke to meet the size requirements on your PCB.
You can as well use single choke and a cap after that instead of full PI filter.

 

[treez]

Why do you use a 50hz transformer, just use LNK302 or LNK304 to give output direct.
You can have the voltage dip as low as you like, providing your feedback loop can respond, and you can get enough power through at the lowest voltage.


here I attach simulation so you can see dc bus voltage excursion with different powers of load..it runs in ltspice.

I hope all readers like it....(PS, if you have an article giving control to output transfer fnctionof DCM, current mode buckboost converter, then I appreciate it).

So anyway, have as little capacitance there as you like...remember that many offline flyback led drivers use almost no capacitance there, as they want a good power factor.

Though I don't understand using a mains transformer followed by a switcher...all you need is a switcher. Power.com do many for you. also ti.com, fairchildsemi, etc

 

[Jester]

I apologize for not providing more detail in my original post. This circuit is not directly fed from the MAINS.
- The AC supply is 12-30Vrms (transformer secondary)
- Current available is only 0.1A, so the SMPS needs to have n>80%

The issue that concerns me is the effect of the 60Hz ripple (sawtooth) applied to the input of the converter (even if it stays above the minimum input voltage).

 

[treez]

yes its 12-30vrms secondary, but the primary is connected to the mains 230VAC

 

[Jester]

yes its 12-30vrms secondary, but the primary is connected to the mains 230VAC

The transformer is part of the existing equipment, and this is the power source available. So no access to 120 or 240Vac. The circuit provided also has a specified current limit of 100mA. So the LNK parts are not viable.

 

[treez]

if it is 100mA then the best thing would be to do a standard offline flyback, 230VAC in....it will be much smaller cheaper and efficient than a mains transformer, But if you really want the 50hz txfmr, then a discontinuous mode sepic or buckboost would be fine, and would allow your input voltage to droop to zero at the mains troughs and it wouldnt matter too much, as you could have a big-ish output cap of the sepic or buckboost ..eg LM5118 for buckboost.

 

[Jester]

if it is 100mA then the best thing would be to do a standard offline flyback, 230VAC in....it will be much smaller cheaper and efficient than a mains transformer, But if you really want the 50hz txfmr, then a discontinuous mode sepic or buckboost would be fine, and would allow your input voltage to droop to zero at the mains troughs and it wouldnt matter too much, as you could have a big-ish output cap of the sepic or buckboost ..eg LM5118 for buckboost.

Please reread my last post. I DO NOT HAVE ACCESS TO 240VAC

 

[treez]

my post #8 goes on to suggest a good method for you which isnt using 230VAC.....An LM5118 based buckboost...you re-pasted it in your post #9.

However, you will definitely get more help from more people here if you explain the context of your question..because what you are setting out to do (following a mains txfmr with a switcher) is generally poor engineering practice. People will take one look at your question #1, and wonder of its worth helping you to go "skating on thin ice". -People here don't want to direct you the wrong way down a one way street...unless there's some good reason..context

 

[Jester]

treez,

I agree with your comment about context. I settled on a LM46000. The switcher after the existing transformer was a design input requirement in the specification. BTW I have used the LNK parts in the past, just not an option for this design.

Thanks for your comments, they are much appreciated.

Regards,

 

[schmitt trigger]

Jester;
the crossover frequency in most modern, low cost controllers can be designed to easily exceed 5Khz. Which means that it has enough bandwidth to properly cancel out the 120 Hz ripple. (as long as the valley voltage does not drop below the minimum voltage required for regulation)

 

[Jester]

Jester;
the crossover frequency in most modern, low cost controllers can be designed to easily exceed 5Khz. Which means that it has enough bandwidth to properly cancel out the 120 Hz ripple. (as long as the valley voltage does not drop below the minimum voltage required for regulation)

I checked the crossover frequency for this design, and its 32 kHz, so looks like I'm okay. Thanks for your input.