2
\$\begingroup\$

I have a Xiaomi phone that came with a 33W charger. It says it does 5, 9, 11, 12 and 20 volt, so looks like a fairly normal USB-PD charger. Except, it has a type-A socket and came with an A-to-C cable! How is that possible?

It seems to be doing normal PD on the C end, because I successfully charged a laptop at 20V. Both charger and cable have the 5 extra pins from USB 3. Is there a standard for this at all? Does the A-to-C cable have a non-standard pin mapping to make USB-PD work?

Edit: Here's an old eBay listing for an identical charger.

\$\endgroup\$
1
  • \$\begingroup\$ there's many versions of power delivery. \$\endgroup\$
    – Jasen
    Jan 21 '21 at 23:50
7
\$\begingroup\$

tl;dr: USB Power Delivery doesn't require the CC lines of a USB Type-C cable, it can also communicate over the power line.


When most people think of USB Power Delivery, they probably think of two USB Type-C devices talking to each other. However, the Power Delivery standard is actually older than that—revision 1 of USB Power Delivery didn't even mention USB Type-C!

Power Delivery's negotiation (the process where the charger and device agree on the voltage to use) happens over the physical layer, and there are two options for the physical layer:

The first option is, again, what most people think of, and it's using the CC pins of a USB Type-C cable, in yellow below:

Source: https://pinoutguide.com/Slots/usb-type-c_pinout.shtml

The second option (which was actually the only option until Type-C came around) is to use the Vbus line of a non-Type-C cable.

Think about if you were designing this: it has to work with USB 2.0, which only has Vbus, D+, D-, and Ground. Normal USB data communication needs to still work, so you can't use D+/D-...the only option left is Vbus.


But wait! How can you communicate over a power line? We can check the official Power Delivery specification:

Description of BFSK signalling

Apparently, this version of USB-PD used Binary Frequency Shift Keying (BFSK) to modulate the packets over the voltage line. The exact details are probably out of the scope of this post, but if you're interested, I got that screenshot from page 116 of Revision 2 of the USB-PD spec (linked at end).


That explains how USB Power Delivery used to work, but then how is it still compatible with a modern Type-C device, like your laptop? Well, the spec has an answer for that too:

Description of interoperability with BFSK and BMC

(note that here, "BMC" refers to "Biphase Mark Coding"—that's the first option I mentioned before, using the CC lines. The screenshot is page 138 of Revision 2.)

This basically means that, if a device supports communication over CC, it's also able to support the old-fashioned Vbus method for compatibility reasons. They even mention the scenario you're describing: "Note that any system utilizing the USB Type-C connector can see BFSK signaling on Vbus when a suitable USB Type-A/B to USB Type-C adapter is used."

It's not a requirement, so it's possible you could find a device that doesn't support this. It's also not allowed to work the other way around: if a device has a Type-C port, and it supports the BFSK method, then it must also have support for the CC lines.


If you're interested, you can download the official spec from the USB-IF website. For whatever reason, the download seems to only have Revisions 2.0 and 3.0, and I can't find the original Revision 1.0. Revision 2.0 describes both methods though, and is what I used for everything I referenced here.

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

I have a Xiaomi phone that came with a 33W charger. It says it does 5, 9, 11, 12 and 20 volt, so looks like a fairly normal USB-PD charger. Except, it has a type-A socket and came with an A-to-C cable! How is that possible?

Seems pretty apparent to me what they did, they violated the USB spec.

A USB-A port by the USB spec is not to provide more than 5 volts and more than 3 amps. It may be no more than 2.4 amps depending on which spec applies. There's a little bit of room +/- on the 5 volt spec so seeing 5.2 volts will be common, perhaps 5.5 volts might still pass inspection, but 9 volts is just right out. I remember a big deal a few years ago about big names in tech cracking down on nonstandard USB chargers, Amazon and Apple pulling noncompliant products from their stores, Google and Apple threatening to pull any licensing agreements for putting so much as 6 volts on their USB chargers. This is overboard on the breaking of specs like I've never seen.

In looking for some images of this charger on the internet I can't know if the photos I'm seeing are of the same item you have, but what I do see is quite telling. There's no UL listing or other industry safety markings. Nowhere do I see "USB-PD" on the product or in the description, there is no indication that this is a product that even attempts to meet the USB specifications.

It seems to be doing normal PD on the C end, because I successfully charged a laptop at 20V. Both charger and cable have the 5 extra pins from USB 3. Is there a standard for this at all?

No. The USB-A connector was never intended to handle more than 12 watts, 5 volt @ 2.4 amps. It wasn't intended to handle more than 5 watts actually but with some redesign and exploitation of some over engineering in the original design they decided that 12 watts was safe.

Does the A-to-C cable have a non-standard pin mapping to make USB-PD work?

I'd say that is quite likely. Putting voltage on something in the air has the potential to arc. There's ways to minimize this and deal with it if it is unavoidable. Also of concern in a connector that carries power is the heat it will produce from the inherent resistance. One way to lower the heat is to spread out the area of contact. The contacts in a USB-A connector can be only so big and still fit other USB-A plugs. What can also spread out the heat is use more contacts. I have no doubt that they used contacts intended for data to transfer power.

USB-BC, the old "battery charger" spec uses the D+ and D- pins to communicate how much power the connected device was safe to draw. In the pre-BC days USB power supplies would leave these pins unconnected and devices were just expected to behave themselves. Then came all the crazy ways people tried to get more power through USB that got a lot of people upset. Consumers were upset because USB chargers were slow and when they got faster then they were no longer compatible with each other. Some consumers even saw their electronics go up in smoke.

Apple, Google, Amazon, and others were upset because if people can't trust their USB devices from working with each other, and possibly starting on fire, then they had a possible public relations nightmare on their hands. To combat this came USB-BC to push the power allowed to a more acceptable 12 watts, and a new standard on communicating how much power a device could safely draw from a power supply. Not long after came USB-C and USB-PD to allow for more powerful devices to charge from a standardized port.

It just boggles the mind why people go through the effort to engineer these strange and unsafe variations on existing themes. I could have some understanding why this was done in the days before USB-BC and USB-PD. At that time there was a government mandate for phone manufacturers to decide on a common connector for power, this was from China, the EU, or both. The industry chose USB. That lasted about 2 seconds before phone manufacturers started to violate the standard in horribly unsafe ways. The industry moving to USB-C and USB-PD should have removed even the desire to do anything else.

But here we are. Still seeing USB power bricks that can't play by the rules.

If that were my USB power brick then I'd be tempted to smash it to pieces and throw out the bits. That has the potential to damage standards compliant hardware with its non-standard power output. It is not safe and if you choose to keep it then handle it with care. Use it only with the cable and phone of the same brand.

\$\endgroup\$
1
  • \$\begingroup\$ Is it similar with Samsung. I saw USB charger rated for 2 output: 5V and 9V \$\endgroup\$
    – user263983
    Jan 27 '21 at 0:03

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.