SEPIC CONVERTER FOR SUPPLY HIGH SPEED IC
Luca , 02-14-2017, 02:11 AM
Good morning to every one, 3 days ago i've finished a specific design for a customer that need a particular power supply for two boards and i've implemented a non isolated SEPIC converter and works like a champ, at least in my lab, now i'm waiting the response from customer and certification institute.
Tipically i've notice that SEPIC it's a lot more clean in waveform than other DCDC converter, my question is very simple.
Anyone have tested SEPIC for supply for example, a processor or FPGA?
Tipically i've notice that SEPIC it's a lot more clean in waveform than other DCDC converter, my question is very simple.
Anyone have tested SEPIC for supply for example, a processor or FPGA?
robertferanec , 02-14-2017, 03:55 PM
Good question. I am curious to see the answers too.
Luca , 02-15-2017, 02:12 AM
You mean from the customer or certification institute? o both? or what? 
mairomaster , 02-15-2017, 02:49 AM
Hi Luca,
I've never used a SEPIC converter. Apart from providing a clean output, what other advantages do you think they have compared to the conventional DC-DC converters/micro modules? Which brand did you use?
I've never used a SEPIC converter. Apart from providing a clean output, what other advantages do you think they have compared to the conventional DC-DC converters/micro modules? Which brand did you use?
Luca , 02-15-2017, 03:24 AM
I've used a tiny controller from TI. The part number is LM3481.
When i say clean waveform i mean not only at output voltage but also at the switching node and input of the converter.
The switching node voltage are really clean, like in ZVS or somenthing like that... the result in waveform are very close to that tecniques.
Also the input current are very clean and "continuos" instead the typical waveform of buck converter (for example) that seems much more to a sawtooth.
The main advatanges, for my opinion, are 3:
1 - When the converter is in off state, or inihibit state (or somenthing to similar), the output are "isolated" from the input by a capacitor and then the voltage can go to "real zero" also if
the first start of the converter begins, for example, from a pre-biased voltage.
2- SEPIC can regulate the output at fixed voltage, for example 3.3V, while the input voltage can vary from less to more than input voltage. So basically can work in boost or buck mode, depends on output voltage in relationship to the input voltage.
3- the best "ratio of work" for the power mosfet, inductor and loop stability for the response of the converter occurs at 0.37 of duty cycle, good value for magnetic parts design like sepic inductor or input/output filter.
When i say clean waveform i mean not only at output voltage but also at the switching node and input of the converter.
The switching node voltage are really clean, like in ZVS or somenthing like that... the result in waveform are very close to that tecniques.
Also the input current are very clean and "continuos" instead the typical waveform of buck converter (for example) that seems much more to a sawtooth.
The main advatanges, for my opinion, are 3:
1 - When the converter is in off state, or inihibit state (or somenthing to similar), the output are "isolated" from the input by a capacitor and then the voltage can go to "real zero" also if
the first start of the converter begins, for example, from a pre-biased voltage.
2- SEPIC can regulate the output at fixed voltage, for example 3.3V, while the input voltage can vary from less to more than input voltage. So basically can work in boost or buck mode, depends on output voltage in relationship to the input voltage.
3- the best "ratio of work" for the power mosfet, inductor and loop stability for the response of the converter occurs at 0.37 of duty cycle, good value for magnetic parts design like sepic inductor or input/output filter.
Use our interactive Discord forum to reply or ask new questions.