1
\$\begingroup\$

I have been designing a drone flight controller and currently have been using LDOs to regulate the voltage to 5V.

As I have started to use higher voltage batteries the LDOs have started to overheat. I know that switching regulators are so much more efficient which is definitely a plus with battery-operated devices, but heat is not my only concern when considering which power supply to use.

When the drone's motors are on, the voltage supply becomes very noisy, and I have RF ICs on the board that are incredibly sensitive to power supply noise. Would a switching regulator be better for this application, or would it generally perform worse?

TLDR: When comparing LDOs and switching voltage regulators, which are generally the better performing option in terms of output voltage noise when considering an already noisy power supply?

\$\endgroup\$
  • \$\begingroup\$ Search internet on "ldo vs smps" ... \$\endgroup\$ – Huisman Jul 10 '19 at 6:28
  • \$\begingroup\$ The heat comes from wasted energy. The more you waste, the shorter your flight time. \$\endgroup\$ – JRE Jul 10 '19 at 8:01
3
\$\begingroup\$

When comparing LDO's and Switching Voltage Regulators, which are generally the better performing option in terms of output voltage noise when considering an already noisy power supply.

As you have probably found searching internet, a LDO has generally less ripple/noise/EMI than a SMPS.

heat is not my only concern when considering which power supply

You could distribute the heat using resistors in front of the LDO. Adding capacitors to form RC filters help to filter "an already noisy power supply" at the same time.
For example:

schematic

You can also preceed the circuit above with a SMPS to gain better efficiency. You probably don't need the RC filters.

When the drone's motors are on, the voltage supply becomes very noisy

But addressing the cause of the problem rather than patching the effect of the problem would be a better approach.

\$\endgroup\$
1
\$\begingroup\$

You can use both. Use the DC-DC to drop the rail down to a voltage the LDO can use. You get the best of both worlds: DC-DC efficiency, and low noise for your RF stuff.

This is a really common design approach, that is, sub-regulating a local supply from a bulk DC-DC supply to create a rail with special requirements. The TI NMOS ‘cap-free’ LDOs are particularly good for RF applications (TPS731xx, 2xx, etc.)

As far as motor noise corrupting the supply, a way to mitigate that is to make separate runs to the battery for the motor drives and to your radio board. That is, isolate each power on its own loop so that ground/power noise from the motors doesn’t find its way into the rest of your system.

You can also apply some common-mode filtering to the individual power runs to stop the noise from coupling between them. More on that idea here: Noisy Voltage Rails (Vcc, Gnd) - Noise Isolation for Automotive Circuit

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

A switch mode regulator by its nature of 'switching' generates noise. You can kind of think of it like this: An LDO produces waste heat to drop the voltage while a SMPS produces waste noise.

Regardless of which you use, you generally want to do something about minimising the noise into them. While your LDO will produce less noise for a given input, it may not do all that much when it comes to reducing existing noise.

In a drone I would really be looking strongly at the SMPS, simply because you're looking at an extra 50% or more efficiency over the LDO. Noise can be filtered out and you already have noise problems that need addressing regardless of which you use.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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