opulo-inc / lumenpnp

The LumenPnP is an open source pick and place machine.
GNU General Public License v3.0
2.29k stars 310 forks source link

Use USB specific ESD protection package #706

Open warasilapm opened 6 months ago

warasilapm commented 6 months ago

Version Number

Rev5b0

Bugfix or Enhancement

Both?

Description

The USB lines to/from the ESD diode chosen (SRV05-4) is routed a little awkwardly; I think this is a result of the four line configuration of the part chosen.

Suggested Solution

Using something like the USBLC6-2 much simplifies this routing. image

dan-gies commented 6 months ago

I agree that the routing path on the current part is a little awkward.

@warasilapm note that the suggested part should be routed to only one of the two pins (e.g. 3 or 4) and pass under the part to maintain signal integrity. From the datasheet, ST shows one trace of the diff-pair passing under the part, not through it. EDIT applies to power only.

image

warasilapm commented 6 months ago

I agree that the routing path on the current part is a little awkward.

@warasilapm note that the suggested part should be routed to only one of the two pins (e.g. 3 or 4) and pass under the part to maintain signal integrity. From the datasheet, ST shows one trace of the diff-pair passing under the part, not through it.

image

I looked at that part of the datasheet; that is referring to the VBUS line specifically as it continues towards the rest of the board. If anything, it implies the optional diode should follow the ESD IC in the circuit rather than branching off at the connector pin. IIRC the screen cap you took makes sense given a micro-b pinout but I didn't check.

The recommended layout reflects how I show the data lines.

dan-gies commented 6 months ago

@warasilapm you're right! I rushed to conclusions.

warasilapm commented 6 months ago

Image for reference: image

Btw I checked and it definitely did not make sense for micro USB. Who knows what they are thinking with that VBUS trace.

dan-gies commented 6 months ago

For both these parts (existing and new) it's a little weird that they pull the diff-pair so far apart, but I guess the ±Ω tolerance is big enough that it doesn't really matter for FS speeds. For 2.0 I'd think it'd be sub-par, but that's irrelevant here.