Tryingto build a CT-VPD using CCG2

[XenGi]

I wanna build a Charge-Through VCONN Powered Device (CT-VPD).
I went throw the product catalog of infineon and other vendors and figured that the CCG2 is probably the right chip for my use case. I put a picture of my current PCB design as an attachment. I would appreciate your feedback on that design.

You can find some info on CT-VPDs from the USB-IF here, if you want to read up on the topic: https://www.usb.org/document-library/ctvpds-and-making-your-own-usb-cr-thingamajig

For a bit more context, my design is a USB keyboard attachment for a smartphone. The smartphone does support USB-PD 3.0 and can charge with more than 5.0V on VBUS. The CCG2 chip should enable the smartphone connected to the upstream port of my PCB, to power the keyboards micro controller via VCONN and still be charged with 5.0V or more over VBUS when a capable USB-PD charger is connected to the Charge-Through port of my PCB. You can see a good overview of the ports involved on page 10 of the PDF from the USB-IF. My keyboard PCB pretty much follows the same concept as the audio dongle that is shown there.

Because this setup is quite a new approach there is no application note in the datasheet of the CCG2 for it yet. Because of that I'm not sure how to properly design the PD circuitry. I was hoping that you could take a look at my design and give me some feedback if the design should work like that or if I need some adjustments.

I hope I can create a working design with the CCG2 chip or I would need to make a redesign with another vendor. I'm already looking for alternatives but the support engineers there are a bit slow to answer.

If you got any questions, just shoot them at me and I'll try to answer them.

[Wang_Fred]

Hello，

Unfortunately, we have no such Charge-Through VCONN Powered Device reference design you could refer to directly. CCG2 could be VCONN powered, however it's old product and not promoted for new design any more. For CCGx devices, there is rarely pure VCONN powered application other than EMCA.

After checking the function diagram, since it's a charge-through device, the downstreaming port should be able to sink power from external charger then pass to the upstreaming port to charge smartphone, right? If it is, you need 2 PD controllers or 1 PD controller covering 2 ports, not just 1 CCG2 that sending SOP' to host. In other words, it's not just a simple application from my perspective. 

Best Regards,

Fred

[XenGi]

---

@Wang_Fred wrote:

After checking the function diagram, since it's a charge-through device, the downstreaming port should be able to sink power from external charger then pass to the upstreaming port to charge smartphone, right? If it is, you need 2 PD controllers or 1 PD controller covering 2 ports, not just 1 CCG2 that sending SOP' to host. In other words, it's not just a simple application from my perspective. 

---

Are you sure about this? It's not yet clear to me in which scenario 2 controllers are needed. I would've connected VCONN1 from the CCG2 to the phone and VCONN2 to the charger. Then CC1 from the charger routed through to the phone with an additional connection to CC1 on CCG2. This is also how it is shown in the image in my first post.
So basically the same setup as with a single chip EMCA cable. The keyboard could then get power from the VDDD pin.

Did I get this right?

AFAIK the only reason to use 2 CCG2 in an EMCA cable is to save on cost for a copper line in the cable. Cost saving like this is not a priority for me. I would rather reduce component count.

[Wang_Fred]

As per USB Type-C Spec, the definition of CT-VPD is different from EMCA. The VCONN may be removed by source after source has determined the cable’s current carrying capacity, but for CT-VPD, VCONN must be always there. By now I don't see CCGx series has reference application for CT-VPD.

Below is the capture of DRP to CT-VDP model example defined in Spec.

[XenGi]

Why is this accepted as the solution? I still don't have a solution. This image just describes how it should work. Not how to do it with a component from Infineon.

[Wang_Fred · Accepted solution]

Sorry for the misunderstanding. As stated in several previous replies, IFX has no such CT-VDP design solution that only VCONN powered.

[XenGi]

The config software has the option to create a Charge Thought Downstream device based on CCG2. Examples are a HDMI dongle or legacy USB device, which would be my use case. Doesn't this enable VCONN forever? The CCG3 also supports a Charge Through Upstream device mode. One of them should solve my problem, right?

For the CCG3 it has 2 interesting settings. One is "VCONN retain on" which says: "Set to yes to leave VCONN supply on even if the EMCA does not require it". The other is "VCONN required" which says: "Whether VCONN is required for device operation".
IMHO, setting both to yes should give me permanent VCONN right?

[Wang_Fred]

Hello,

The CCG2 and CCG3 application you mentioned use SOP PD message, instead of SOP'. So it's beyond the scope of CT-VPD defined in Type-C Spec. These applications mainly get power from VBUS with a whole PD contract.

Do you just want to be single VCONN powered?

[XenGi]

Yes, I would like to only get VCONN power and let VBUS be connected only between upstream device and charger.