This page covers yet another just-a-chip-on-board project: a home-made SMD-module containing an FT230X USB-to-serial bridge.
I use the FT230X quite often. Instead of re-doing the same layout and schematic over and over again, I decided to make a versatile SMD-module containing the chip itself and all necessary glue-components (caps etc.).
The module contains a USB micro-B connector, and can be reflowed onto the edge of a carrier-PCB to make it talk USB.
Schematic as follows - P101 corresponds to the 8 SMD-pads:
Not much to say here, so let's talk about different voltages a bit.
Bus-voltage "VBUS" is (always) provided by the USB-host. That is, the module is always bus-powered. VBUS is made available through a module-pad.
I/O voltage "VTARG" (up to 3.3 V) can either be provided by the target / carrier-board,or by a low-power on-chip LDO, available as "VCC_3V3". (In the latter case, VTARG and VCC_3V3 pads can be connected together.) The LDO is able to power smaller targets (approx 50 mA IIRC), which is convenient.
From left to right: components (silkscreen), front copper, and back copper:
SMD-pads are arranged in 2 rows, with one row sitting right underneath the USB-connector for mechanical strength.
Here's a nice 3D-view.
3D is the new 2D, folks.
When panelised using break-tabs, It makes sense to keep mousebites off the side with the USB-connector, since the connector hooks around the PCB-edge a tiny bit:
Single PCB is shown below. I'm still evaluating (...) USB micro-B connectors instead of my previous giant of choice - the USB-B - and I like it quite a bit. Feels solid, and supposedly survives a lot of mating-cycles. We'll see.
Module fitted onto a baseboard is shown here:
I happened to use a 130 um stencil for this. Module turns out reasonably level; works OK, and I can't see any issues under the microscope.
We'll see how this turns out - only time and more PCBs will tell.
Delivered to you by Vim, GNU Make, MultiMarkdown, bozohttpd, NetBSD, and 1 human.