using USB-CC PD power bank with PPS hopefully as a battery, what should I know?

[Simon]

After lots of politics about who's battery to use has gone into orbit we in the design office have decided that maybe the best way to actually get to start on the damn thing is to use a USB-C power bank with power delivery, should keep the world happy. We should be able to get up to 20V which is fine for our application.

But this USB-C PD thing looks like a minefield to me, I'm not even sure how many standards are involved. The PPS protocol could be very useful for us and avoid designing some power electronics into our own product.

So where do I even start? I'm designing a device only (not a host) and I'd like to get as much out of the PD functionality as I can. I see plenty of power banks offered as capable of laptop charging so I suspect that these do implement the PPS protocol?

I'm now on the hunt for a USB-C PD controller IC.

[amwales]

I think there are gotchas to keep them awake

[jbb]

Considered USB-C for a product and learned about some potential snags along the way:
- going above 20 V 3 A requires a special cable, so stick to 60W max if you can
- USB-C ports on a desktop or laptop seem unlikely to go above 5 V 3 A
- there are, allegedly, non-compliant chargers out there which deliver 20 V all the time
- the CC pins (used for PD negotiation) are right next to the Vbus pins and therefore could be subjected to 20V during plug-in :-(

I think you can get USB Power Delivery chips from Texas Instruments and Cypress (and probably many others!) with the right protocol engine built in.

Oh, turns out Cypress was acquired by Infineon…
https://www.infineon.com/cms/en/product/universal-serial-bus/usb-c-power-delivery-controllers/

[SiliconWizard]

USB-C PD is one thing already, but another issue in your case is the power bank. *These* are the real minefield here. Many only comply with PD marginally, and one of the most annoying things with these is that they are almost all made to go to a sleep mode when the drawn current goes below a certain threshold (which of course depends on each vendor/model), and will not wake up unless you actually disconnect the cable and reconnect it.

Really, while the "power bank" thing looks like a good idea on the surface, it's a nightmare waiting to happen if you're going to design a product around one. These are not made for that use case. There are mainly designed for recharging mobile devices. You're in for a lot of... fun, believe me.

[Simon]

I think you can get USB Power Delivery chips from Texas Instruments and Cypress (and probably many others!) with the right protocol engine built in.

Oh, turns out Cypress was acquired by Infineon…
https://www.infineon.com/cms/en/product/universal-serial-bus/usb-c-power-delivery-controllers/

Yes I have a CY4533 kit on order.

[mino-fm]

Really, while the "power bank" thing looks like a good idea on the surface, it's a nightmare waiting to happen if you're going to design a product around one. These are not made for that use case. There are mainly designed for recharging mobile devices. You're in for a lot of... fun, believe me.

Yes and no. You can keep awoken PBs by short current pulses every some seconds. I'm doing this since years and it works with most PBs.
Other circuits can recharge local LiIon cells from time to time or timed by a clock module. Please use translator:
http://mino-elektronik.de/Powerbank/powerbank.htm#pbw4

For higher voltages than 5 V you should switch back to 5 V before go to standby mode.
Ok, you have to pay some attention but it works if desired.

[Njk]

Considered USB-C for a product and learned about some potential snags along the way:

Add one more fact to the list:
-The USB-C connector is tiny and mechanically is not so strong. It's for mobile devices, there is no desktop version.

[Simon]

Yes I'd look for a USB-C connector in some sort of industrial connector shell.