Charging connectors

[CrazyIvan]

But that point about not being able to charge and use the USB3 port at the same time is a very good counter point. That causes a bit of trouble for a docking station.

Hm? Why?


All phones I had so far could connect via Micro USB to the PC and charge at the same time?

USB host always provides power and USB device always receives it so if it's being charged then it's device.

Hub won't help as it needs pyra USB to be in host mode, so yes, USB3 in dock will be unusable

Can µUSB ever (with a hub/special adaptor, or otherwise) be used to charge the device acting as logical host?

unless it's made according to USB Power Delivery specification,no.

But if pyra would support this specification then yes with special usb pd hub

Edit: USB PD defines additional negotiation protocol to choose appropriate voltage (up to 20V) so pyra will need additional chip to handle this protocol but I don't know if it's possible, maybe there are only full USB controllers that support PD.

If USB3 controller is in the SoC than it could mean added complexity and costs for a board only to be able to use USB3 while charging.

Maybe it's better to make USB3 for full ports and USB2 for microUSB assuming its controller supports OTG?

Edit 2:  Yes, USB controllers are implemented in OMAP5. I guess we need gta04's comment here on feasibility of USB PD implementation in pyra

 

[Neelix]

If it's in use as OTG then unless it happens to be connected to a USB3 port it's only getting 500mA max.

Making it work in slave mode with an alternate power source that can provide more than 500ma is trivial.

Define trivial. If it's in slave mode the power lines are connected to the host USB device How do you then also hook an alternate power source into the circuit when the physical connector is already occupied?

- Neelix

 

[Binky]

USB host always provides power and USB device always receives it so

http://www.usb.org/developers/powerdelivery/

Power direction is no longer fixed. This enables the product with the power (Host or Peripheral) to provide the power.

 

[CrazyIvan]

USB host always provides power and USB device always receives it so

http://www.usb.org/developers/powerdelivery/

Power direction is no longer fixed. This enables the product with the power (Host or Peripheral) to provide the power.

Assuming USB controller conforms this specification but I'm sure it is not, see my previous post

 

[WizardStan]

Define trivial. If it's in slave mode the power lines are connected to the host USB device How do you then also hook an alternate power source into the circuit when the physical connector is already occupied?

Don't connect them to the host device, just connect them to a 5v 2A power supply. The device doesn't (or shouldn't) actually care whether the plugged in device is drawing power or not.If it does care then it is slightly less trivial to connect in parallel with some diodes: same voltage, device still detects something drawing power, but the bulk of the power comes from the external source.

Power direction is no longer fixed. This enables the product with the power (Host or Peripheral) to provide the power.

Well, that would solve that particular problem.

Maybe it's better to make USB3 for full ports and USB2 for microUSB assuming its controller supports OTG?

It doesn't. The USB3.0 is the OTG port and the only way to get a slave mode.

 

[Neelix]

Define trivial. If it's in slave mode the power lines are connected to the host USB device How do you then also hook an alternate power source into the circuit when the physical connector is already occupied?

Don't connect them to the host device, just connect them to a 5v 2A power supply. The device doesn't (or shouldn't) actually care whether the plugged in device is drawing power or not.If it does care then it is slightly less trivial to connect in parallel with some diodes: same voltage, device still detects something drawing power, but the bulk of the power comes from the external source.

How? Is there a standard cable that lets me do this or are you expecting me to hack one together? Having to hack a custom cable removes it from the realm of trivial IMHO.

- Neelix

 

[CrazyIvan]

http://www.prnewswire.com/news-releases/displaylink-and-rohm-demonstrate-usb-power-delivery-solutions-at-ces-2014-239077341.html

USB PD is only recently demonstrated and uses USB controller with USB PD support. The possibility for pyra to support USB PD is quite bleak unless there is some kind of add-on solution and it can be bought

Edit  http://www.techpowerup.com/196105/etron-announces-ej888-usb-power-delivery-controller.html Maybe this is what we need but the description is quite vague, it's not exactly clear if it's just USB PD controller or USB PD+USB 3.0

I couldn't find anything about this chip on company's site

 

[Galaxis]

But that point about not being able to charge and use the USB3 port at the same time is a very good counter point.

Hm? Why?
All phones I had so far could connect via Micro USB to the PC and charge at the same time?

I assumed the primary use for an USB3 port would be to connect some external storage. And then it would kinda suck to not be able to charge the Pyra when an external hard disk is connected via USB3...


Or is the plan to use the USB3 port just as docking connector (with the docking station including an USB3 hub and charging functionality towards the Pyra)?

 

[Granitehead]

Or is the plan to use the USB3 port just as docking connector (with the docking station including an USB3 hub and charging functionality towards the Pyra)?

It's not sure whether that is possible.

 

[notaz]

If it's in use as OTG then unless it happens to be connected to a USB3 port it's only getting 500mA max.

Making it work in slave mode with an alternate power source that can provide more than 500ma is trivial.

Doing that properly on pandora is actually difficult. The twl/tps chip has no ability to detect anything and it all has to be done in software, so you have to enumerate the device to find out how much it can provide. However the kernel drivers are abstracted away from each other and the charging driver has no way to know about USB subsystem, unless some driver hacking is done. USB chargers are their own story and there are several types of them and need to be detected somehow. None of this is actually implemented on current pandora, it just tries draw ~500mA and violates the spec which forbids taking more than 50 (or was it 100?) mA before negotiation.

So it's unclear what we'll have on Pyra, maybe something that can at least detect chargers in hardware, but probably not..

 

[fusion_power]

My 1000mA charger can barely push the Pandora Battery over 80% and usualy stops at 70% until I reconnect it. I doubt the battery will charge better with only 500 or less mA from an ordinary USB connector.  h34r:   OK, it may be an old Battery but the Amps still can make a difference while charging. The recommended value of the Pandora Charger may be 2000mA if I remmeber correctly. So I would support a normal barrel form jack for an efficient charging of the Pyra.

 

[WizardStan]

Is there a standard cable that lets me do this or are you expecting me to hack one together?

Any standard USB Y connector would do it. Plug the "data" end of the Y into your PC and the non-data into any USB wall charger. This is effectively how those USB powered hard drives work, by powering off of two USB ports, except we could plug one into a wall charger for even more juice. Which one is which would need a small amount of experimentation since when using on a PC you usually plug both into USB sockets and don't care which is the extra, but that shouldn't be hard to figure out.And I'd expect that such a setup would be built right into a docking station.

Doing that properly on pandora is actually difficult.

Difficult to do automatically. It's pretty easy to just change the numbers and tell it to draw 1A+ over USB, just don't then turn around and plug it into a PC without turning it back down.I meant more in the hardware sense anyway, physically connecting multiple power sources so that it can draw more power than a single USB port can provide.

 

[TrashyMG]

My 1000mA charger can barely push the Pandora Battery over 80% and usualy stops at 70% until I reconnect it. I doubt the battery will charge better with only 500 or less mA from an ordinary USB connector.  h34r:   OK, it may be an old Battery but the Amps still can make a difference while charging. The recommended value of the Pandora Charger may be 2000mA if I remmeber correctly. So I would support a normal barrel form jack for an efficient charging of the Pyra.

I charge via USB 2.0 port on my laptop with 500mA limit to 100% all the time on all of my Pandora(s). I also use a 1 Amp wall charger often too.

 

[Granitehead]

It also depends on the computer, mine is able to power one of those cheap chinese Android sticks over USB which doesn't run from my 1.2A phone charger. So the computer must provide more than 1.2A on that port, probably close to the 2A peak that the Android stick wants.

 

[Natsu]

I charge via USB 2.0 port on my laptop with 500mA limit to 100% all the time on all of my Pandora(s). I also use a 1 Amp wall charger often too.

I was just going to ask can it charge by USB 2.0.

 

[Neelix]

Yes, but not quickly, and not if you are using any high power applications on the device at the time.

Lets say I was running a game in DOSbox....  (which tends to run the CPU at or near max) and noticed the power was running low.  If I connected the USB2 cable with the intention of charging it while I continue my game I'd very likely run it flat instead.

- Neelix

 

[slaeshjag]

And this danger will probably continue all the way up to 1.2 A charging current for the Pyra.

 

[WizardStan]

I was just going to ask can it charge by USB 2.0.

It cannot use the USB2.0 ports of the OMAP5 if that's what you're asking, the only OTG port is the USB3 one.That being said, you can of course plug it into a USB2 host on your PC and get the minimum charging, as was mentioned.

I have no idea about what kind of PMIC they're planning on putting in this time around, the Pandora could not automatically detect whether the USB charger was plugged into a PC or a wall charger but maybe the Pyra will be able to, especially since they're planning to remove the barrel charger.

 

[moxie]

I very much like my barrel jack. Just wanted to have that for the record.

 

[_jr_]

I'd say these are my requirements: (1) being able to charge the device without additional special hardware and/or kernel hacks while having something connected to the USB 3/OTG port (in both host and slave mode) and (2) being able to actually charge (at least slowly) the device using the included charger and default firmware while running CPU/GPU at 100% and (3) the charging port survives at least several thousand insertion/removal cycles.

If these can be satisfied with a single micro USB port, I won't complain. But I'd rather sacrifice a second full size USB A port to get a dedicated charging port than one of the above requirements.