2
\$\begingroup\$

I need to power three devices/loads from a single DC source (a 48V at 200A battery). The buck converter used is this.

The three loads are:

  1. RFID reader (24V DC @ 2.1A)
  2. Rasberry Pi 4 (5V DC @ 3A)
  3. DC fan (5V @ 800mA)

Which option is more efficient in terms of power?

Option 1:

enter image description here

Option 2:

enter image description here

Block Diagram of power supply: enter image description here

\$\endgroup\$
10
  • 2
    \$\begingroup\$ I don't think there is a 'general' answer. It depends on the efficiencies of the specific converters you buy/ build. \$\endgroup\$ – brhans Dec 3 '20 at 14:12
  • \$\begingroup\$ @brhans sorry I added the buck converter to be used. It has an efficiency of 95%. \$\endgroup\$ – Joey Dec 3 '20 at 14:15
  • 3
    \$\begingroup\$ Option 1 provides power transfer at 90% (95% of 95%) efficiency to Pi and Fan. While Option 2 provides 95% efficiency with every connected device. \$\endgroup\$ – Mitu Raj Dec 3 '20 at 15:04
  • 4
    \$\begingroup\$ The DC/DC converters you link to have a maximum Vin of 32V and 38V respectively; neither should be used with an input voltage of 48V. Also, in calculations note that efficiency often depends on Vin - Vout. \$\endgroup\$ – ocrdu Dec 3 '20 at 15:10
  • 1
    \$\begingroup\$ If it is an option, using a 24V battery would make everything cheaper, smaller, and more efficient. \$\endgroup\$ – ocrdu Dec 3 '20 at 15:54
3
\$\begingroup\$

In the ideal world, doing everything in one stage would result in higher efficiency as the losses are not compounded.

However high ratio converters (eg 48 > 5) require very small duty cycles (about 10% in this case) and tend to be more expensive and lower efficiency, Option 1 does the 5V output in two steps so doesn't require such high ratios (2:1 and 5:1).

This does have the downside that the input converter (48 > 24) has to be able to provide the power for both the 24V and the 5V loads as well as lower efficiency as the 5V power has to pass through both converters.

For example with 90% efficient converters it will require 15.8/0.9 = 17.5W at the input of the 24>5 converter and 19.5W at the input of the 48V converter to power the 15.8W of the RaspberryPi and fan. If driven with a single-stage converter only 17.5W would be required.

Also, take into consideration that the 48V input from a battery will vary significantly. Depending upon how it is being charged it could vary from ~60V down to ~42V.

There may also be transients that need to be tolerated such as when other loads are switched on or off. These transients may be significantly more than 60V. Filtering inductors, capacitors and surge limiters may be required at the input.

\$\endgroup\$
5
  • 1
    \$\begingroup\$ Wouldn't a capacitor at each load inputs help with the large transients? As well as a fuse for addtional saftey? \$\endgroup\$ – Joey Dec 3 '20 at 15:29
  • 2
    \$\begingroup\$ @Joey - Yes. A capacitor at the input would be required anyway to avoid undesirable effects of the inductance of the wiring to the battery. \$\endgroup\$ – Kevin White Dec 3 '20 at 15:38
  • \$\begingroup\$ Just to be sure will a capacitor at the input of each buck converter only be required or do I need to also have to add a capacitor at each of the inputs of the loads as well in addition to the capacitor at the input of each buck converter? This is for Option 2. \$\endgroup\$ – Joey Dec 3 '20 at 15:42
  • 1
    \$\begingroup\$ You have to be careful about adding too much capacitance to SMPS outputs; read it's datasheet for limitations. \$\endgroup\$ – rdtsc Dec 3 '20 at 15:53
  • \$\begingroup\$ @rdtsc The power supply already has a built-in EMI filter and a capacitor at its output builtin. So there is no need for additional filters. See picture in post. Thank you! \$\endgroup\$ – Joey Dec 3 '20 at 16:04

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.