[sdiy] USB-C experience

Didrik Madheden nitro2k01 at gmail.com
Sat Sep 30 20:49:26 CEST 2023


I come from a different world when it comes to electronics design, and this
might come off as a bit harsh, but my reaction is, well d'uh.

Re: 1) they're meant for automated soldering, so manual assembly was never
a design priority. Design and order a board from a manufacturer with
assembly service like JLCPCB. This could even be a breakout board with
castellated vias around the edge so you could mount it to a 0.1" proto
board if that's what you want to do.

2) likewise this is less of a problem if you're designing with a BOM that
tracks the schenatic. But I do agree that it's an annoying aspect.

3) Well, yeah. It's the headline feature of the spec, and the digital part
of the spec is designed around making this work. This should be a fully
expected problem.

4) To me this comes across as very obvious to me. How else would you do it,
considering the intended use case? The multiple gnd and Vcc pin are there,
apart from symmetry, for redundancy. If they were separated into individual
wires, loose contact one pin would not increase total contact resistance,
but also wire resistance since one whole wire is now out of commission and
not just one pin.

It was some time since I looked at the spec, and while I'm not sure it's
specifically called out that gnd and Vcc share the same wires, I'm pretty
sure they're consistently referred to as a single signal, which at least
heavily implies this fact.

And btw, another reduction in conductor count: only one of DP1 and DP2 (the
center pins) are wired up in a typical cable. This is because a port is
only expected to have a single USB 2.0 differential pair, and the extra
pins in the connector are only there for redundant rotational symmetry.

So the number is down to 16 pins. I believe (not sure now) that the
standard mandates that the gnd pin should be distinct from the shield, so
you might be ok to try using the shield as your ground and gnd as a 16th
signal if you're ok with the risk of bad contact. This risk is nonzero
since the shield is not gold plated, doesn't have a spring contact and is
just intended to absorb interference.

/Didrik

On Fri, 29 Sep 2023, 22:36 Sean Ellis via Synth-diy, <
synth-diy at synth-diy.org> wrote:

> (I made this post on wigglers but I'd like to share here. I'd love some
> feedback from people who might have opinions.
> https://www.modwiggler.com/forum/viewtopic.php?p=4036980  )
>
>
>
>
> I've spent the last few weeks investigating buss cables for running
> between cabinets and I'll save people the trouble with USB-C. On paper it
> could be handy being so small and so many lines, for CV it would work but
> there are some showstopper issues.
>
> Firstly the connectors are SMT only, and even then virtually all
> connectors have to go through an oven or hot air to make proper joints. You
> can't physically get under them with an iron so are rather unfriendly for
> even prototyping.
>
> Secondly there are approximately 9000 different footprints with little in
> the way of a standard. Identical looking parts are off by 0.25mm and don't
> fit.
>
> Third the reversible nature is a waste for us. You need to find a solution
> for keying them to make use of the full cable. (design specific...)
>
> Lastly, it's 24 pins! On paper 24 pins.....in reality it's 18 or less. All
> the cables I've tested have *merged the power and ground lines within the
> cable*. This is sadly not written in the spec so I wasted money on boards
> that didn't account for that. I guess you could make your own cables using
> the male plugin boards but what's the point.
>
> Stick to HDMI or another type even if only carrying CV. Wish they'd
> mention somewhere in the spec that the power pins are merged as a rule.
>
>
> (There's more in the thread about the cable issue)
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at synth-diy.org
> http://synth-diy.org/mailman/listinfo/synth-diy
> Selling or trading? Use marketplace at synth-diy.org
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/20230930/12ad0064/attachment.htm>


More information about the Synth-diy mailing list