1
\$\begingroup\$

I'm currently studying different approaches to supplying a PCB which needs both 3.3V and 5V from a wide range DC input source (i.e., 6-36V.)

The most common layout I've come across is V_in(6-36) -> buck (3.3) -> boost (5.) Is there a reason for this being the most common? What are the potential drawbacks of V_in (6-36) -> buck(5) -> buck(3.3,) or not putting them in series at all and simply using two buck converters side by side?

\$\endgroup\$
10
  • \$\begingroup\$ To some extent it depends how much current you need on each rail. Which you haven't mentioned. You can make a case for any of those scenarios (and a couple of others) for different requirements. \$\endgroup\$
    – user16324
    Commented Sep 7, 2021 at 16:27
  • 1
    \$\begingroup\$ I would not say most common (far from it actually) but one reason to do that is dropout with a 6V input being unable to support 5V.. \$\endgroup\$
    – DKNguyen
    Commented Sep 7, 2021 at 16:28
  • 1
    \$\begingroup\$ Another reason could be supply sequencing, eg. when the 3V rail must always be on before the 5V rail. \$\endgroup\$ Commented Sep 7, 2021 at 16:33
  • \$\begingroup\$ Thanks a lot, could you elaborate on how different currents affect the choice? In my specific case its going to be in the 100-400mA range for both. But I'm interested in the general case and the reasons for that. \$\endgroup\$
    – paulst
    Commented Sep 7, 2021 at 16:35
  • \$\begingroup\$ Downstream converters load the upstream converters they are connected to. So if you have a 10A supply, you probably want it running through minimal upstream converters, \$\endgroup\$
    – DKNguyen
    Commented Sep 7, 2021 at 16:36

2 Answers 2

1
\$\begingroup\$

If the 5V requires only a small current, using a boost (or even a flying-cap converter) from 3.3V would give wide flexibility for the input voltage. For higher 5V current it could instead be a buck-boost directly from Vin, which would be more efficient than boosting from the 3.3V rail.

One challenge you face is the large input voltage range required. This has a profound effect on the DC-DC: the stepping ratio down to 3.3V varies from about 1.8:1 (55% duty cycle) for 6V to about 11:1 (9% duty cycle) for 36V. Why does this matter? It affects the inductor choice and switching frequency: the higher the stepping ratio, the shorter the minimum on-time becomes.

If you can step down to 5V instead for your bulk voltage, then make 3.3V from that, then these extents become 83% and 14% for 6V and 36V, respectively. This will allow a higher switching frequency. As it is, many DCDC's will run at close to 100% duty, so the lower input is doable even though the Vin-Vout overhead voltage is just 1V (some regulators will even switch to linear mode as Vin gets close to Vout.) Meanwhile your 3.3V gets to be a low-voltage input type, so possibly less expensive and able to run at a higher switching frequency.

\$\endgroup\$
0
\$\begingroup\$

Because the 3.3 V current is so small, I would use a buck regulator to take the input down to 5 V, and a linear regulator to pick off some of the 5 V and turn it into 3.3 V. 0.7 W is not a lot of heat to move.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.