I’ve been rather interested in how USB-C “does its thing” when it comes with USB-C PD and DRD (dual-role devices). Being rather late to the USB-CC party, I’ve not had any experience with how it works at a low level, but I thought I’d get some commercial adapters that did what I wanted and examine how they (might) work.
This post looks at a T Tersely USB-C OTG Adapter that has 24W USB-C PD charging pass-through capability. How does this work, I wondered? (Oh, and is their brand Tersely (which I think would be more sensible) or T Tersely (as they’ve called their shop)?).
The Item
The unit arrived in a sky-blue and white box. Nothing fancy, very generic.
They guarantee 100% satisfaction with a motto of “Enhance Your Daily Life”. It seems their products are likely to be rebadges of other products available elsewhere.
At a first glance, this “dongle” bears a passing resemblance to an Apple product, perhaps their “camera adapter”. But unlike Apple products, the wire is thin and stiff, the casing is thin and hollow-feeling and there is little weight in the product.
The USB-A socket is a USB 3.0 type with its blue tongue and USB 3.0 SuperSpeed contacts visible. A USB-C socket provides charger power from a compatible USB-C PD charger.
But curiously, looking down the hole of the captive USB-C cable’s connector shows that it is missing the contacts for the SuperSpeed pairs. As a result, this is almost certainly a USB 2.0 product and the listing does allude to this.
Semi-Destructive Teardown
What’s inside? Getting my side-cutters, I made quick work of the plastic casing with only some minor damage to the PCB and USB-C socket that didn’t entirely put the unit out of commission. It is noted that the USB-C socket is, like many of the recent Chinese cheap products, made of very thin metal that bends at the slightest nudge. I wouldn’t have confidence that the shell would serve to make good contact or retain plugs in the long-run.
The board is marked HZF-OTG B-2_v1.1 dated 11th April 2022. It seems to have solder pads for wires to be tacked down for the SuperSpeed pairs, but I’d have to wager that such a termination on a 2-layer 0.8mm PCB would not be satisfactory for the signal integrity of USB 3.0 signals. Perhaps running this was causing enough instability and returns that they simply didn’t fit USB 3.0 wires, but didn’t change the design and left the USB 3.0 connector as it might create the impression of a faster adapter.
On the underside, we can see the clearly marked pads where the captive wires are tack-soldered to the surface. This arrangement is perhaps cheap and sufficient, but not ideal for signal integrity either – whatever happened to termination of the shield, or the twists of the D+/D- wires?
Looking closer at the connector area, it seems an unmarked chip orchestrates most of the action, with a set of different MOSFETs (2N7002 N-ch, BM3415E P-ch) in charge of switching VBUS from the internal supply to the external supply.
The bottom half with the SOT23 and inductor appears to be a switchmode converter based around some chip marked CWAQ to derive power for the mystery IC. This is perhaps necessary as VBUS can change depending on what USB PD profile is negotiated.
What is this IC?
This had me wondering, what is the mystery IC and how is it connected. It wasn’t marked on the top-side, so I removed it …
… to find it is marked 072714 2217H1E. This isn’t particularly useful and suggests to me that the chip has a date code of Week 17 of 2022. But it is likely a 16-pin microcontroller of some description that has been programmed at the factory.
I looked closely at the board and also noticed that the majority of the pins are not-connected, further supporting the notion that this is a microcontroller of some sort. Interestingly, it doesn’t use a bus driver (e.g. FUSB302) and seemingly interfaces to CC pins through a resistor.
Tracing lines, it seems it interfaces to two CC lines from the charger, with only one CC line out to the device via a series 12-ohm resistor. It controls the MOSFET VBUS switches, while also seemingly measuring the VIN from the charger through a voltage divider (maybe to know when to generate the power supply ready message). There is a pin connected to a capacitor – I wonder if this is a bootstrap supply or a hold-up supply to bridge over VBUS switching events when power could be lost for a short period.
Indeed, it is a bit of a mystery to me – but maybe probing this board while it’s in use will provide some more answers.
Conclusion
This product feels a bit cheap, light and flimsy. The USB-A port does not operate at USB 3.0 rates in spite of being a USB 3.0 port with SuperSpeed contacts and has a very weak metal shell which does not bode well for its longevity. Internally, it seems to be based around a mystery pre-programmed microcontroller to perform the PD negotiations between the charger and host, with an assortment of MOSFETs to perform VBUS switching and a switch-mode power supply for itself.
