Each new model of RPi "pushes the envelope" in terms of the amount of power it draws over a USB electrical interface. The RPi 4 was spec'd to require up to 15 watts: 3 Amps at 5 Volts. Three (3) Amps was at or near the "ragged edge" of the current-carrying capacity for the power wiring in a USB-C cable - typically 22 or 24 AWG. Now, the RPi 5 is spec'd to require 5 Amps.

How do they do this? Are they violating USB "standards", or have they somehow compromised the performance of their new RPi 5 for the sake of retaining a USB power interface?

1 Answer 1


AFAIK: No USB standards violations have been committed - since the RPi4 (⊙⊙).

Instead, it is the standards themselves which have changed. Specifically, the latest addition to the line of "Official" power supplies is the "Raspberry Pi 27W USB-C Power Supply", which provides up to 5A current at +5.1V. This supply may have been accommodated by the new-ish Power Delivery (PD) additions to the USB standard that were announced back in 2019.

You might wonder what's in this new PD standard - the details, iow. To answer that, there is a "Developer's" Briefing, and there is a specification document - embedded in this .zip file. These are both interesting documents - for different reasons. The briefing comes across as a product of fevered minds plotting to take over the world. The specification document is 1,113 pages, including 7 full pages (double column) listing the "Contributors". The companies listed as "Contributors" give you clues that this is all motivated by money. Now when you look at the USB cable you bought from Amazin' for $8-$10 with the official "logo" stamped on it, you can understand what's going on.

Lest readers become confused wrt my agenda, let me say that as a career engineer, I do not consider myself a Luddite - but I know the smell of greed also. Instead of forking over $12 for the new PD supply, I personally would be just as pleased with a simple barrel connector that brings 12V (or 5V) to the RPi. Anyway - back to business.

After perusing the 1,113-page specification, I found nothing that fully answered the question. I did find a section titled "4.4 Power and Ground" in yet another document (the "USB Type-C Cable & Connector Specification", see REF #1 below) that details limits on the "maximum allowable cable IR drop". This is useful, but still doesn't answer the question. But the "product brief" on the new 27W supply discloses the following:

Cable: 1.2m 17AWG, white or black

Wow... 17AWG. Consulting a wire table we see that 17AWG has a resistance of 16.60992 ohms/km. This gives us a total wire resistance for the 1.2m power supply of:

Rwire = 1.2m * 1km/1000m * 16.60992 = 0.01993190 ≅ 0.02 ohms

Which yields a voltage drop of ~ 0.1 V at 5A, but this excludes contact resistance. Per para of the document at REF 1, that is limited by specification to 50 mΩ. With a 'captive' cable, we double this figure to obtain the total contact resistance (supply & GND return) of 100 mΩ. This gives us:

Vdrop = ( Rwire + Rcontact ) * 5 Amps = 0.12 * 5 = 0.6 V

And since the new 27W supply is rated at 5.1V, that leaves us just a wee bit shy of the 4.63V "low voltage threshold" imposed by the PMIC:

5.1V - 0.6V = 4.5V

So - that is how they (RPi.com) "did it". The two 17AWG wires helped, but they could not turn the trick with that amount of contact resistance. And so we're asked to pay $12 for a 27W supply that may not quite avoid Low Voltage warnings at full load, and still operates on "the ragged edge". This supply does have some added features, such as different operating voltages, but these features provide no additional utility for the RPi 5. I cannot help but wonder what motivates RPi.com to cling to their "power via USB" policy; seems to me that barrel connectors offer both cost and performance advantages.


  1. Section 4.4 of the document "USB Type-C 2.3 Release 202310 2/USB Type-C Spec R2.3 - October 2023.pdf", which is embedded in this document bundle
  • Cannot agree with you more. The USB delivery of power to RPi5 is nonsense. If they at least cared to use Power delivery or Quick charge protocol, so we can use the gazzilion throwaway chargers from mobile phones. No care for enviroment and usability for end user, just pure greed in my opinion. Hopefully it will be running no problem with 5.1V running through GPIO, like i power all my PIs. Commented Dec 12, 2023 at 14:56

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.