I'm an STM32 user, so I use the SWD connector. (not yet this project).

I am not aware of the "SWO" pin in that configuration....

The "discovery" boards also have the target NRST signal connected on the programming connector. there is a spot available for that -> do it.... (possibly mark as optional... )

The "discovery" boards have a 6-pin connector and I have been unable to find what the 6th was ever intended for.
I have now repurposed that for the "boot0" signal. With a pulldown to GND and a switch to VCC I can then boot the target into bootloader mode.

I suggest to mark one of the further free pins for that purpose. Then, I'd be able to have two of the clamps, one on my programmer and one to a two-switch-board that allows me to hit the boot, and reset buttons to make the target switch to bootloader mode. (=DFU over USB for most of my targets).

Hi again.

I suggest 2 more pinouts - combination of SWD + UART + FactoryMode detection and same for PIC. They're already presented in my libs. I use these combos in my projects, so I can get debugger, console and FactoryMode detection from single connector speeding up the manufacturing process and making R&D space cleaner. FactoryMode is just GPIO with pull-up, that is pulled down by the programming clip and makes the device enter init / test sequence after boot automatically.

Here are my pinouts:
PIN | SWD | PIC
1 - VCC - VCC
2 - GND - GND
3 - NRST - MRST
4 - SWDIO - PGD
5 - SWCLK - PGC
6 - RXD - RXD
7 - GND/FM -GND/FM
8 - TXD - TXD

I've designed this connector into a PCB, I think it's quite a neat solution - thank you for designing it.

Unfortunately the SOIC clip I used was a bit flimsy, and so never made very good contact with the PCB, and before long the plastic guide pins broke.

I was wondering therefore if you had the model number of a SOIC clip that was easy to modify, and worked reliably?

Nice work!

If, instead of the 5 holes after the contact pin, you fill one of them (for example, the 2nd one to the top) and cut the matching plastic part on the grip, you have revert connection protection.
It's even probably cheaper since that make one less hole to drill on the PCB.

This worth it if you intend to reprogram your device few month later and don't want to spend time figuring out the pinout of the SOICbite's connection, or if you have (unskilled) workers programming board in series.

what is the purpose of the holes? seems they get get in the way or u have to use thin traces? Just wondering why they are there in the first place?

Hey,

nice work! Whats the max. resaonable trace width to route between the inner pads?

I'm creating SOICbit footprints and components for https://librepcb.org/, in order to connect an ARM based board to an SWD debug probe (JLink).

JLink offers a (non-standard?) 5V pin on the connector (power from USB) that can be turned on via software: https://www.segger.com/products/debug-probes/j-link/technology/interface-description/#swd-connector-pinout Do you think we could put it on pin 2 or 3, to get some kind of standardization?

Thanks for putting this out there :).

I'm putting this into my personal KiCAD library, which I plan on throwing up on github, but I would feel weird putting it up without some sort of attribution, or confirmation that that is okay with you in some way.

It's really not necessary but I do suggest putting a license file somewhere, may I suggest BSD, or the WTFPL.

Do you have thoughts on generating converter boards that go from the SOIC clip standard IDC connectors to protocol specific interfaces eg. typical USB/UART module pinout?

I'm aware that this might move SOICbite away from a more connection agnostic solution, but the table of suggested connection pins in the Readme was really good from my point of view, it lets me just take the connection I want to add, and route to a pad arrangement that already exists, instead of having to create new ones each time, eg. xkcd.com/927.

I've forked a repo at DefProc/SOICbite with a USB converter I'm trying as an example.

Made a copy in Diptrace. Could push sch and pcb libs if you'd like them here.

I'm finally trying to implement this footprint on one of my board, and I'm wondering what is the best expected distance from board edge ?

In the connector footprint, there's an inverted T on the bottom, but I just wanted to be sure it's meaning it's supposed to match the board edge:

Can you confirm ?

Hello and thank you for this nice project.

I'm thinking of doing a flex PCB for running some UART lines out of a device case but
I'm stuck on how to later connect to the extension from a UART-USB adapter.

This looks promising as it is dead easy and does not require me to place a new connector (SMT or TH), hell I could reuse any SOIC clip lying around.

Let's say we use OSH Park and work along some limits.

Obviously we need solder mask on both side of the PCB but that is covered

Soldermask Type | SMOBC (Soldermask Over Bare Copper), both sides

We also need two layers of copper and that is also covered.

So the real question is if we can actually find a SOIC connector that can clamp this
extremely thin PCB as the stackup is the following.

Which would yeld a total of 10.8 mil ( ~ 0.25 mm).

So maybe just one side would be available at a time, the other covered by some thickening to make the connection more reliable as otherwise the film might just slip out all the times.

We also have some restrictions on how far from the edge copper should be as the process
of cutting it's different from rigid material.

Board Edge Keepout Copper within 15mil (0.381mm) of a board edge will be fabbed, but may be damaged during shipping, copper within 6mil (0.152mm) of a board edge will be removed prior to fabrication

So overall if the connector is not placed very close to the border it should survive the fabrication process and the SOIC clip can still reach it.

I'm interested in the drill holes as for flex pcb there are different limits and the wear and tear might be different.

Is there any other physical requirement for this connector that I haven't gone over?