USB C PD for General Purpose Use?

A little research shows USB PD V3.0 can only deliver 3 amps at 12 volts.

Doing some digging:

USB PD 1.0 supports
10W (5V, 2A)
18W (12V, 1.5A)
36W (12V, 3A)
60W (12V, 5A) <-------
60W (20V, 3A)
100W (20V, 5A)

USB PD 2 and PD 3 support
0.5 – 15W (0.1 – 3.0A)
15 – 27W (9V, 1.67 – 3.0A)
27 – 45W (15V, 1.8 – 3.0A)
45 – 100W (20V, 2.25 – 3.0A) [3.0 – 5A with rated cable]


So it looks like certain PD V1 could support what I want to do. Alternatively, I could use a PD "trigger" rated at 20 volts and a buck converter to get the current I need at 12 volts.

I'd suggest sticking with 20V, your setup will be compatible with any PD version then.
It also gives you a bit more current at 12V after the converter than the 20V input current.

Edit - and charge a 12V battery or lithium pack from the same supply, for backup?

Yeah, I think the 20 volt PD trigger and a buck regulator is the way to go. This would be easy enough to test with a couple modules, and designing a dedicated board with the trigger and buck converter to reduce the size wouldn't seem to be too difficult.

Definitely agree that designing for a standard output in the current version of the USB PD spec rather than an unusual configuration in an obsolete spec is the way to go. Ads for chargers aren't always clear in what voltage ranges are supported; if you're lucky, there's a picture of the enclosure where the options are listed.

All this PD stuff is starting to make sense now.

Looking at 3 PD chargers I have:

Lenovo computer:
20V – 3.25A (65 watts)
15V – 3.0A
9V – 2.0A
5V – 2.0A

iBook computer:
20V – 1.5A
15V – 2.0A
9V – 3.0A
5V – 3.0A

Generic charger:
20V – 2.0A
5.2V – 2.0A

It's my understanding with the trigger devices, if the requested voltage isn't available, you get the next lower available option; with these three chargers, if you wanted 12 volts, you'd only get 9 volts or even only 5 volts.

When I first saw the USB C PD voltages above 5 volts, I thought they were likely chosen as the commonest supply voltages for devices using 2, 3 & 4 cell battery packs.

eg. Most laptops use 15 - 16V if they have three cell or 19 - 20V for four cell series type batteries.

You're probably right, since the driving force was computers and peripherals. I suspect nobody thought about 100 watt or even 200 watt chargers when the concept was created. Or that it would become the generic power option for DC devices.

Things I have learned so far.

USB C PD chargers only supply a few fixed voltages:

5V - always, and one or more of:

9V
15V
20V

12V may be available on older generations of chargers, but it is no longer required and isn't common on newer chargers.

The charger output voltage is negotiated with the connected device. If the device's requested voltage isn't available, the next lower one is supplied. The connected device may or may not be able to make use of the lower voltage.


For my application of running a 12 volt CPAP, I am using a USB PD 20 volt "trigger" module, which handles the negotiation with the PD charger, and a DC-DC buck converter to drop the 20 volts (or possibly 15 volts) from the PD charger to 12 volts for the CPAP.

The trigger module is surprisingly tiny (which really shouldn't have been a surprise since it's not much bigger than the USB C connector!); it's easy to see how it can be built into a standard size USB C cable. The small blue board in the picture is the trigger module.




The red board below is the buck converter. It's huge compared to the trigger module, but it's only about 1" × 2".



Testing will be coming.

As a reminder, from a linear voltage regulator, you can't get more current out than is going in. The excess voltage gets converted to heat.

A switching regulator is far more efficient. The output power will be roughly 85% of the input power.

For a linear regulator, with Vin = 20, Vout = 12, and current = 3 amps,

Power dissipated by the regulator:

P = (Vin - Vout) × A = 24 watts

This is power wasted as heat.

For a switching regulator:

(Vin × Ain) × 0.85 = Vout × Aout

Ain = (Vout × Aout)/(0.85 × Vin)

Ain't = (12 × 3)/(0.85 × 20) = 2.1 amps

42 watts input ==> 36 watts output (and 6 watts heat) with a switcher compared to
60 watts input ==> 36 watts output (and 24 watts heat) with a linear regulator.

My first experiments weren't entirely successful. The DC-DC converter rated to handle 5 amps won't handle the startup current of my 3 amp CPAP.

I tested the DC-DC converter under load. At about 2.5 amps, the chip got hot pretty quickly, and the output voltage dropped substantially. Adding small stick-on heatsinks to the top of the chip and back side of pcb didn't improve things much, and with a 2.1 amp draw, the chip was already pretty warm after a couple minutes.

So my quest is a better 20v – 12v DC-DC converter that will handle 3+ amps for 8 hours continuously.

Y0u might try looking at automotive suppliers?, cars are 12V, trucks are 24V, so it's a VERY common requirement dropping 24V down to 12V to feed radios and CB's etc.

Something like this perhaps?

Thanks Nigel. I've been looking at some similar units (looks like that particular one isn't available in the US) but adding a 24v – 12v adapter that's as big as the CPAP's AC adapter defeats my goal of a smaller/lighter one adapter goal for travel.

The project has really clarified USB PD chargers for me, so it's not a wasted effort whatever I find.

I've used some of this style; they seem to work OK at 5A or so.

Good call on the DC-DC converter module! Success running my CPAP with a 20 volt trigger module and the linked DC-DC converter.

My CPAP is rated to draw 3 amps. I tested the DC-DC converter with a 3 amp load and the inductor was uncomfortably hot. But at whatever load my CPAP actually draws, the DC-DC converter runs cool and the CPAP reliably starts.

A little packaging, and this will be a good solution.