00:20 <houkime> needed connector displacement: around 2.9 mm (optimistic).

00:22 <houkime> on-bob connector movability max 1.3 mm OTG-defined (which probably will still ruin sth else)

00:24 <houkime> bob X-axis mobility as a whole - doesn't seem to be very constrained provided that you manage to place flash etc in the right place.

00:28 <houkime> preliminary conlusion - bob form adjustment might be needed to enable movement of Hackerbus etc X+ direction without displacing flash etc.

00:34 <houkime> gonna sleep. bye

01:47 <enyc> him-cesjf: ooh godo luck!

01:47 <enyc> error

01:47 <enyc> houkime I mean, aaah, not here

02:39 <Oksana> Neat : https://www.extremetech.com/gaming/267288-ftc-warns-companies-warranty-void-if-removed-stickers-are-flatly-illegal

11:20 <houkime> Joerg-Neo900: besides camera, flash etc is there a slot in the case that Bob needs to fit in?

11:23 <houkime> if we could just move HB with 2 OTG holes 3 mm and extend bob towards pcbcenter to accomodate this move that would be enough.

11:25 <houkime> probably some trades here and there can diminish movement to 2mm-esque.

15:37 <houkime> it looks like the only concern is the battery

15:46 <houkime> also the design of usb OTG holes itself is concerning. You sure it is ok to run a differential pair like that? 7.5 mm apart?

15:48 <houkime> usually you keep these signals together for them to generate less noise

15:52 <houkime> this is how physics works. + edges limits - edges' influence on surroundings. Like a dipole thing.

15:53 <houkime> You can't make them less sharp because it is a fast speed bus, so you keep sharp edges, but pair them together instead.

15:55 <enyc> houkime: i've certainly seen the USB2.0 wires along side each other a mm or 2 on pcbs

15:55 <enyc> (data wires)

16:23 <houkime> Rechecking guidelines and stuff. Nope, I'm certainly right, at least it is a consensus position. There're also quite stict routing rules. It is rather 0.2 mm depending on material etc. Metacollin used 0.5 (still don't know what material he targeted).

16:24 <houkime> you Do want to separate OTG-D pair from other signals though.

16:25 <houkime> It includes Vbus and OTG ID which are also part of the usb

16:28 <houkime> so you separate two traces of the differential pair from these 2, and yes, even on metacollin's layout this separation is about 1.5 mm.

16:43 <houkime> there's a little quirk here though. The thing is, you need to maintain impedance between traces and air-crossing pins is an electrically different from pcb situation

16:47 <houkime> this further complicates things because electrical properties of air are not constant and depend on humidity and stuff.

16:49 <houkime> Directing differential pair across pogopins looks like quite a bad idea, although I need to calculate stuff to be certain about it.

16:53 <houkime> (and like with lightbeams which partially bounce off any material with n different from air/vacuum, signal gets partially echoed back on any transition between electrically different phases)

17:04 <wpwrak_> to be honest, i never really expected that OTG USB hack to work, so i just tried to make it as electrically inoffensive as possible if not used ;-)

18:10 <houkime> have you considered to build a small testboard to test a pogopin approach?

18:11 <houkime> It would be rather reasonable to test this thing before inclusion to v2?

18:13 <houkime> (oh well, it is not included in v2 anyway. V3?)

18:15 <houkime> Unless it is tested I shall consider pogo geometry undecided and disregard cutouts.

18:18 <houkime> This gives more mobility to connectors, which I will use to try solving usb collision problem.

20:30 <houkime> HB connector datasheet URL (Harwin) is outdated

20:30 <houkime> correct docs are on this page https://www.harwin.com/products/M50-3150442/

20:31 <houkime> *outdated means 404

20:35 <houkime> however aaaah, Harwin says that this connector goes obsolete 15-th June 2018

20:35 <houkime> Connector CON1703

20:37 <houkime> this is in its datasheet

23:57 <houkime> It might be actually possible to cram things together on bob if using tented vias under connectors.

23:57 <houkime> and do some HEAVY pin rearrangement

23:58 <houkime> It might as well produce neeeded 3 mm, although maybe not quite