1
\$\begingroup\$

I am designing a custom PCB and want to power 2 rails: a 3.3V and 5V rail, with a single 250mAh LiPo battery (3.7V) and need the 5V rail to have as stable a voltage as possible.

I understand that a buck-boost converter with the right voltage bias ratings could theoretically power both these rails in isolation, though for this application I am questioning using a single buck-boost to power both the rails simultaneously as I am trying to reduce space on my PCB and improve efficiency. I am fairly new to using DC converters and thus I don't know if it is common practice to do this, and if it is, where to begin with searching for the right buck-boost for this application (bearing in mind I have to make sure the output voltages have as low ripple as possible and the voltage levels don't deviate significantly from the intended level). I need to be able source ~12mA on the 5V line and ~152mA on the 3.3V line.

I have searched for some buck-boost converters and found some I think would work (such as this one) but just want confirmation if it is advised to use them in this way. Alternatively, would it be advised to use a separate voltage regulator and boost converter to achieve this? Thank you!

\$\endgroup\$
1
  • \$\begingroup\$ How stable does the 5V rail need to be? How much ripple is too much? \$\endgroup\$
    – vir
    Nov 19, 2023 at 15:14

4 Answers 4

1
\$\begingroup\$

You can use a linear regulator along with a boost converter to get 3.3V and 5V respectively.

\$\endgroup\$
2
  • 2
    \$\begingroup\$ Given this is a battery powered application, using a linear regulator to get 3.3 V from the nominal 3.7 LiPo can lead to (a) wasted energy in the linear regulator (b) in the later part of the discharge curve the linear regulator might not have sufficient headroom to produce a 3.3 V output. Admittedly I'm not sure how to accurately calculate differences in the maximum runtime from a charged LiPo; it depends upon the efficiency of a linear regulator compared to DC/DC regulators. \$\endgroup\$ Nov 19, 2023 at 12:44
  • \$\begingroup\$ @ChesterGillon, all true, I just want it to be simpler than using two regulators. \$\endgroup\$
    – Shamurov
    Nov 20, 2023 at 13:47
0
\$\begingroup\$

You need two different regulated voltages. So you need two DC/DC regulators, not one.

Therefore, no, a single one can't do that. You might be able to do it with two instances of the same type of regulator. But that has little practical advantage.

Note that there are some ICs that contain multiple regulators in one package, but these very likely were made for a very different target audience than you. (Namely one where compactness trumps ease of use)

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

-You may find a single DC-DC with a dual channel output that can provide you with 5V and 3.3V.

where to begin with searching for the right buck-boost for this application

You can loot it up in mouser.com, digikey etc. Any electronics-selling vendor with appropriate filters to find the IC you need. The IC you linked to your question has its pads below the IC, you might not be able to solder that by hand, just saying.

-Most of the space when you design a DC-DC hookup is taken by the Inductor and the capacitors, so dont worry so much regarding space if you use two DC-DC ICs.

My first recommendation is that you use one DC-DC to ramp the battery voltage to 5V, and then use a LDO regulator to bring the 5V down to 3.3V. Since you need to draw such low currents, using a DC-DC regulator to produce the 3.3V rail doesn't seem like its needed. But you know better, depending on your application.

My second recommendation, since you mention that you want the 5V rail to be as stable as possible, would be to use a DC-DC IC to generate ~7V and then use a LDO regulator to bring 7V voltage to 5V. And again, a second LDO regulator to bring the (7V or 5V) to 3.3V.

Both my recommendations are to use a LDO for the 3.3V, because

  1. you dont need to draw much current so an LDO wont heat up that much.
  2. Using a DC-DC to draw low currents are not so efficient compared to DC-DC hookup for high(er) currents. Check this example plot from MP28167 IC:

MP28167 DC-DC

The lower the current, the lower the efficiency. Especially if the DC-DC IC is a high current-capability one.

Regarding the other considerations you may have:

bearing in mind I have to make sure the output voltages have as low ripple as possible and the voltage levels don't deviate significantly from the intended level

You dont specify the amount of ripple voltage you would be happy with, but here are some techniques to reduce the ripple: -Using my second reccomendation, after you generate these 8V, you can use a capacitance multiplier to filter the se 8V, you will have a voltage drop of ~0,7V due to the transistor of the multiplier, but a much more cleaner DC voltage to use it as the input of your LDOs.
-A more advanced technique is using parallel switching ICs, but I don't think this is what you want to do as a starter for your application.

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

While you may need two regulators for two different voltages, the modern element base offers a few interesting tricks.

E.g. you may get a single-chip switching mode buck+boost regulator that uses a single inductor for 2,3 or more output voltages.

enter image description here (source)

The inductor is usually the bulkiest part of a switching mode regulator. In this topology, the outputs "timeshare" a single inductor.

\$\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.