USB-C Charging on the PBP?

[Idaho]

When I bought my PBP I also bought an USB to barrel charger but it was of poor quality and it doesnt charge well, so I switched to charging via USB-C and it works just as well, I live in a van or in an off-grid cabin so I use solar power and 12vdc to charge my stuff, I charge from USB plugs with either 1A or 2,1A and it works very well...

[MtnSk8]

Power supply specs from the wiki  "USB-C 5V, 15W PD quickcharge" seems incorrect. If it's fixed at 5V then it's not "PD". 

K, the first time I actually played along with this, the second time I just kept quiet, but now I'm actually getting a bit annoyed...

Without PD USB can deliver up to six unit loads for high power SuperSpeed devices, where the unit load is 150mA on USB 3.0 (SuperSpeed) and 250mA for multi-lane SuperSpeed. That means without PD USB can deliver up to 7.5W. USB PD was developed specifically to standardize the way to deliver more than 7.5W of power, and than includes 15W, the maximum power deliverable without increasing voltage. So technically 15W delivery in PBP is USB PD, compliant with revision 2 or 3 of the standard, though saying "quickcharge" is indeed a bit of a misnomer since that would imply a different proprietary technology by Qualcomm that predates USB PD standard and AFAIK is available only on their Snapdragon platform.

That said, IMNSHO 15W is barely adequate for PBP.

Thanks for the reply, sorry to annoy. I must have misunderstood, this chart seems to suggest that usb-c is capable of 15w without the PD standard and 100w with it.


https://en.wikipedia.org/wiki/USB_hardware#Power

Edit: Video

[moonwalkers]

Power supply specs from the wiki  "USB-C 5V, 15W PD quickcharge" seems incorrect. If it's fixed at 5V then it's not "PD". 

K, the first time I actually played along with this, the second time I just kept quiet, but now I'm actually getting a bit annoyed...

Without PD USB can deliver up to six unit loads for high power SuperSpeed devices, where the unit load is 150mA on USB 3.0 (SuperSpeed) and 250mA for multi-lane SuperSpeed. That means without PD USB can deliver up to 7.5W. USB PD was developed specifically to standardize the way to deliver more than 7.5W of power, and than includes 15W, the maximum power deliverable without increasing voltage. So technically 15W delivery in PBP is USB PD, compliant with revision 2 or 3 of the standard, though saying "quickcharge" is indeed a bit of a misnomer since that would imply a different proprietary technology by Qualcomm that predates USB PD standard and AFAIK is available only on their Snapdragon platform.

That said, IMNSHO 15W is barely adequate for PBP.

Thanks for the reply, sorry to annoy. I must have misunderstood, this chart seems to suggest that usb-c is capable of 15w without the PD standard and 100w with it.


https://en.wikipedia.org/wiki/USB_hardware#Power

Edit: Video

No need to apologize - consulting the USB Type-C and USB Power Delivery specification documents directly from https://www.usb.org/documents looks like we may turn out to be both right and wrong at the same time, since both standards were introduced at the same time and are highly intertwined and somewhat confusing.

Based on USB Type-C doc, it looks like non-PD ports can indeed supply (and I assume draw too, that I haven't found explicit mention of that) up to 3A at 5V without actually implementing USB PD, though the mechanism for that negotiation is outside of typical USB protocol of using discreet unit loads. Non-USBPD ports can still be DRP (dual role port), in that case whether they draw or supply power is determined by the other device connected to them, whether it is sink-only or source-only.

Based on USB PD doc, what determines whether device is a PD device is whether it uses so-called SOP packets to negotiate power flow direction (which device is a sink and which is a source) and a combination of voltage and current after establishing initial USB link. PD-compliance cannot be judged on power alone if it's below 15W, because PD standard explicitly covers low-power uses like USB-C headphones, just like it cannot be judged on power flow direction alone.

So what can we do to verify whether PBP's type C port actually implements USB PD? Well, it looks like with PD (unlike non-PD type C) it should be possible to swap between DFP and UFP (downstream/upstream flow port) roles on a DRP like the one in PBP clearly is, so long as the other device connected to it is also a DRP. So I believe if one has, say, an Android phone with USB PD where they can select that phone should be charging the attached device then if PBP's type C port actually supports USB PD then it should be possible to do both, charge such phone from PBP as well as charge PBP from such phone. With one caveat - the type C cable used to connect the two devices also has to be PD-compliant *sigh*. At the moment I have neither such cable nor such phone handy so I cannot do the test, if someone else does and can verify it for all of us - that would be awesome.

Now, who was it claiming that USB is simpler than barrel plug? :-D

---

Ah, much simpler than the test I proposed:

The Power Delivery Specification, in addition to providing mechanisms to negotiate power also can be used as a side-
band channel for standard and vendor defined messaging. Power Delivery enables alternative modes of operation by
providing the mechanisms to discover, enter and exit Alternate Modes. The specification also enables discovery of
cable capabilities such as supported speeds and current levels.

Since PBP supports DisplayPort alternate mode over the type C port that by necessity means it implements USB Power Delivery specification.

[MtnSk8]

@moonwalkers

K, but you asserted, quite rudely, that anything over 7.5w MUST be PD.

You were totally wrong.

Instead of writing a novel you could just say "oops, sorry for being so wrong and so rude"

Bottom line:
THERE IS NO PD CHARGING ON THE PBP.

Have a nice day!

[wdt]

Personally, I don't mind reading a "wall of text" if there is LOTS of information in it
thanks moonwalkers

[as400]

Guys, no need to fight over this shit.

Truth is that USB-C was designed simplify things while now it's total mess. Look at your discussion. I mean how a normal person can get an idea what she/he should buy.
Whether the charger supports power delivery, the cable supports it and then the device itself. And then there is DisplayPort and Thunderbolt ... Supported by some USB-C devices.

This is a great example how modern IT is going into the bushes.

[moonwalkers]

@moonwalkers

K, but you asserted, quite rudely, that anything over 7.5w MUST be PD.

You were totally wrong.

Instead of writing a novel you could just say "oops, sorry for being so wrong and so rude"

Bottom line:
THERE IS NO PD CHARGING ON THE PBP.

Have a nice day!

My comment about me getting annoyed wasn't directed specifically at you, it was more at the general "WTF says PD but can't charge quickly" attitude. I should've made that clear and apparently I failed to do so, for that I apologize.

And yes, I was wrong in my claim that without PD USB type C cannot deliver more than 7.5W, but I incidentally happened to be correct that just because the port cannot go above 3A@5V doesn't mean it doesn't support PD. You were correct that USB type C can deliver 3A without PD, but incorrect that inability to go beyond 3A@5V means it doesn't support PD. Hence my statement that looks like we may be both correct and incorrect at the same time.

I hope I've clarified my position and we can return to a more civilized discussion.

And to address your last comment:

THERE IS NO PD CHARGING ON THE PBP

According to the specs, it appears there are effectively three different ways to negotiate power delivery:

• standard in-band "unit load" approach, where device requests the number of said unit loads and host controls the power output - until introduction of USB type C also the only standard approach
  
• USB type C @1.5A and @3A, where upstream host provides given fixed power and the downstream device decides how much it will take
  
• USB PD, where devices initially use standard minimal unit load and then re-negotiate both voltage and current.
  

From reading those specs it is my understanding that if PBP is connected to a PD-aware charger through a PD-aware cable then the power delivery between the two devices will be negotiated using PD mechanism but still will not exceed what PBP's charge controller can take, whereas if either the cable or the charger are not PD-aware then the power delivery will happen using either USB type C @1.5A/@3A or the pre-type-C "unit load" approach.

While we now know that PBP does indeed support USB PD (since that is a prerequisite for supporting DisplayPort alternate mode), we can verify your claim of PBP not supporting PD charging using the experiment I proposed above - take an Android phone known to support PD, take a known to support PD USB type C cable, connect those to PBP, and configure the phone to provide power to PBP--if PBP starts charging then there is most definitely PD charging on the PBP. Like I said before, I don't have either the phone or cable like that handy, so if someone else could try that it would be awesome.

[lot378]

The dispute was not that PBP supports PD or not it's that PBP doesn't use it to advantage and is currently under-powered when trying to charge the battery and operate the system. It's good that others are working this out for themselves even if they go down the wrong path to begin. Further understanding may eventually lead to something helpful.

There is a workaround through disconnecting the battery but it's a kludge for the power design issue.

Pine64 probably know about it 100% but so far have been keeping pretty quiet about saying what they'll do.

[moonwalkers]

The dispute was not that PBP supports PD or not it's that PBP doesn't use it to advantage and is currently under-powered when trying to charge the battery and operate the system.

You are correct - the original context of mine and everyone else's "doesn't support PD" was as in "doesn't take full advantage of PD when it comes to charging". And then I misread the reference to the power-specific section of the wiki as a reference to the USB type C port in general, which IS of course PD as without PD DisplayPort alternate mode would be impossible, and that lapse of attention got me all TRIGGERED *facepalm*. That said, despite unintentionally offending MtnSk8, I think indeed something good came out of it - I believe I've gained a bit more understanding of how exactly USB negotiates power and I hope I was able to clarify at least some key aspects of it to others as well.

But back to the power-specific wiki section - from what I understand, it is still correct to talk about the USB-C port being PD even in that context, since it's more about the mechanism of negotiating the current/voltage rather than about the amount of said current and voltage. On the question of "quickcharge" - I do think that is a misnomer, primarily because there is Qualcomm Quick Charge which is a separate and until v4 incompatible with PD protocol. But in defense of whoever put it there - QCv1.0 could allow just 10W, which is only 2/3 of what PBP can take, and while 15W isn't all that quick at the battery capacity we have in PBP in the phone world (most of the consumer mobile uses for ARM SoCs, where all this terminology got established) that would be pretty quick.

[wrzomar]

Is there a simple way to debug USB-C PD on the Pinebook Pro? I've (relatively) recently bought a powerbank with PD output, but it is a simpler and cheaper version (GreenCell PowerPlay20 Model: PBGC03) with max. 18W. There are two ways of charging PBP with it, using USB-C port or USB 3.0 port, but sometimes USB-C (on the powerbank side) doesn't work with PBP, if it happens I can charge using other ports or restart PBP. Sometimes USB-C on the PBP doesn't work at all and restart is needed. Sometimes USB-C charging works (but not from PD port on the powerbank) and the USB-C hub won't work (is not powered up by the PBP) without restart. Is a reboot necessary or is it enough to just reload a kernel module?