| 00:00.25 | *** join/#neo900 atk (~Arch-TK@ircpuzzles/staff/Arch-TK) |
| 00:01.09 | Joerg-Neo900 | https://www.we-online.de/web/en/index.php/show/media/04_leiterplatte/2013_1/webinare_1/hdi/02072013_Webinar_HDI_DesignGuide_1_EN.pdf seems useful |
| 00:02.15 | houkime | they even have a piece when D+ and D- are run in parallel on different layers |
| 00:03.32 | houkime | and crossing different numbers of vias |
| 00:03.56 | Joerg-Neo900 | note that GTA04 is not exactly a benchmark though |
| 00:04.34 | Joerg-Neo900 | I don't even know if USB2.0 speed been tested for signal integrity |
| 00:05.08 | Joerg-Neo900 | if we can do better than GTA04 we for sure would want to |
| 00:07.34 | Joerg-Neo900 | ideally you verify the layout of such critical traces with a vector network analyzer |
| 00:10.12 | Joerg-Neo900 | but in the end it's just a question of "does it work, or doesn't it?" |
| 00:12.34 | houkime | Well, my problem with design guides now is that they rarely account for extreme situations. |
| 00:13.47 | Joerg-Neo900 | hmm. at least that Wuerth design guide webinar has a quite luxury dogbone BGA example on page 29. If only we could do this, we wouldn't need in-pad vias |
| 00:13.59 | Joerg-Neo900 | but no way for 0.4 pitch |
| 00:19.15 | houkime | Joerg, do you have a zap gun? |
| 00:19.39 | Joerg-Neo900 | what for :-D |
| 00:23.46 | *** join/#neo900 infobot (ibot@rikers.org) |
| 00:23.46 | *** topic/#neo900 is http://neo900.org | CCCAMP15 lightning talks at http://neo900.org/stuff/cccamp15/ - major: http://neo900.org/stuff/cccamp15/ccc2015talk/neo900-wpwrak_CCC2015.webm | conversations are logged to http://infobot.rikers.org/%23neo900/ and http://irclog.whitequark.org/neo900 | Latest news: http://neo900.org/news/joerg-pia-2018-03-01 |
| 00:23.46 | *** mode/#neo900 [+v infobot] by ChanServ |
| 00:30.41 | houkime | should we consider B2B connectors as possible ESD entrypoints. That's the important question |
| 00:30.49 | houkime | because if we do |
| 00:31.00 | houkime | some esds should be actually doubles |
| 00:31.07 | houkime | doubled |
| 00:31.42 | Joerg-Neo900 | http://susepaste.org/35332748 https://www.we-online.de/web/en/leiterplatten/webinare/archiv/microvia_hdi_webinar/teil_ii_1/webinar_archiv_18.php |
| 00:32.25 | Joerg-Neo900 | houkime: nope, B2B are always between PCBs and thus I consider then sufficiently "shielded" |
| 00:32.35 | Joerg-Neo900 | them* |
| 00:33.19 | houkime | yes, but it is a tinkerphone. user WILL disassemble it |
| 00:33.30 | houkime | and if he does, ZAP |
| 00:33.56 | houkime | can ruin everything |
| 00:35.43 | houkime | This is sort of reliability scenarios that haunt me. |
| 00:45.11 | Joerg-Neo900 | when user disassembles the device, the B2B are still reasonably protected by mere sequencing of signals in connector. We can't take care to make a disassembled device as robust against environment hazard incl massive ESD as the assembled device |
| 00:47.19 | Joerg-Neo900 | nice http://susepaste.org/60071649 |
| 00:52.00 | Joerg-Neo900 | then we also have level shifters on many of the signals |
| 00:54.32 | Joerg-Neo900 | what bothers me most is actually the OTG-USB (pogopins) |
| 00:54.56 | houkime | tests are needed for pogos |
| 00:55.06 | houkime | do a small testboard |
| 00:55.20 | houkime | you can't probably tell otherwise |
| 00:55.31 | Joerg-Neo900 | ESD is very hard to test |
| 00:56.52 | Joerg-Neo900 | actually I think usually nobody tests it, unless it's MIL standard or somesuch |
| 00:57.17 | Joerg-Neo900 | you do your best during design and then trust in what you did |
| 00:57.35 | houkime | ESD maybe not, but the whole concept of usb pogos in terms of signal integruty. |
| 00:57.43 | Joerg-Neo900 | maybe run a few zap tests against the complete device |
| 00:58.04 | houkime | if they fail signal integruity no esd tests on pogos are needed. |
| 00:58.23 | houkime | *integrity |
| 00:58.47 | Joerg-Neo900 | when HB USB signal integrity doesn't pan out for USB2.0, we simply rate it USB1.1 only, unless we can fix some issue after proto-v2 |
| 00:59.34 | Joerg-Neo900 | proto_v2 is exactly to test all this. And to show off something tangible |
| 01:00.16 | Joerg-Neo900 | we won't do single-item tests for every detail, it's too expensive |
| 01:00.22 | houkime | proto_v2 doesn't include pogos. |
| 01:00.30 | houkime | It is in your documents |
| 01:00.31 | Joerg-Neo900 | err? |
| 01:00.53 | Joerg-Neo900 | you lost me |
| 01:01.13 | Joerg-Neo900 | prot_v2 definitely includes hackerbus |
| 01:01.25 | houkime | it includes hackerbus |
| 01:02.21 | houkime | But it doesn't include usb for hackerbus |
| 01:02.39 | houkime | let me recheck this] |
| 01:02.48 | Joerg-Neo900 | and we will use a tiny pcb with two pogopins and a USB cable soldered to them, to test HB USB in proto_v2 |
| 01:04.11 | Joerg-Neo900 | then we'll see if USB2.0 flies or not |
| 01:06.32 | Joerg-Neo900 | maybe we talk different topics? there are actually TWO usb channels available on HB: the OTG and a dedicated one |
| 01:08.04 | Joerg-Neo900 | the latter one is subject to feasibility, it needs quite a few additional BOM |
| 01:10.08 | Joerg-Neo900 | don't worry about such stuff yet, it's a nice topic to work on when the huge majority of design is finalized |
| 01:10.43 | houkime | i guess i know what's going on |
| 01:10.51 | houkime | rechecking the schematics |
| 01:11.12 | houkime | wait a sec |
| 01:13.55 | houkime | oh, ok. they are actually included as testpoints |
| 01:13.57 | houkime | ok |
| 01:14.05 | Joerg-Neo900 | you got USB_HB_DP/M and OTG_DP/M |
| 01:14.06 | *** join/#neo900 Kabouik- (~kabouik@236.34.200.37.customer.cdi.no) |
| 01:14.43 | Joerg-Neo900 | the latter as "testpoints" though they actually are pogopin landing pads |
| 01:15.53 | Joerg-Neo900 | http://susepaste.org/32537385 |
| 01:18.38 | houkime | they are huge though |
| 01:19.00 | houkime | even if i just connect them as stubs (unconnected) |
| 01:19.28 | houkime | they can detriment signal integrity |
| 01:19.42 | houkime | on the LOWER side of D pair |
| 01:21.04 | Joerg-Neo900 | sorry, you lost me. What's huge? |
| 01:21.14 | houkime | landing pads |
| 01:21.17 | houkime | they are huge |
| 01:21.26 | houkime | and they are conductive |
| 01:21.42 | Joerg-Neo900 | not larger than the average solder joint in a USB-A jack |
| 01:22.23 | Joerg-Neo900 | please don't worry about such stuff right now too much |
| 01:27.12 | Joerg-Neo900 | USB specs say 90 Ohm differential impedance +-15% is completely ok for cables *up to 5m* |
| 01:29.27 | Joerg-Neo900 | and watching a standard through-hole USB-A (or -B) receptacle, you'll notice that it's almost impossible to layout impedance matched traces that connect to the pins of those, so small discontinuities are perfectly ok too |
| 01:32.03 | houkime | I know about acceptability of small discontinuities. I just think bout how to connect these things in parralel to a pair, about echoes and such. |
| 01:34.28 | houkime | they are a low frequency resonator |
| 01:34.32 | houkime | basically |
| 01:35.56 | Joerg-Neo900 | a pad of less than 2mm diameter is no low freq resonator |
| 01:38.31 | houkime | gonna check freqs of USB 3 and compare to characteristic freqs of this stuff. |
| 01:38.38 | houkime | just to be safe |
| 01:38.59 | Joerg-Neo900 | the two pads are 5.8mm apart from edge to edge, I suggest running the tuned pair through the center between then and add two ~2mm stub traces from the pair to the two pads |
| 01:39.12 | Joerg-Neo900 | we have no USB3 |
| 01:40.26 | Joerg-Neo900 | we also could move the two pogo pads closer together, I guess the center distance of 7.6mm been picked rather arbitrarily |
| 01:42.48 | Joerg-Neo900 | btw N900 has a ZSB D+/- testpad pair in battery bay - no idea of the layout for that |
| 01:42.52 | Joerg-Neo900 | USB* |
| 01:43.35 | Joerg-Neo900 | basically I just moved those testpads from battery bay to HB |
| 01:49.54 | Joerg-Neo900 | if you feel like, you may move the pogo landing pad that's closer to case edge towards the other pad, away from case adge. Keep in mind however that user needs to solder the pogopins to a PCB and they need a certain distance between their possibly large footprints |
| 01:51.25 | houkime | possible pogo places are determined by BOB--lower connectors now |
| 01:51.36 | Joerg-Neo900 | seems the one we used for reference has a diameter of 1.83mm |
| 01:52.27 | Joerg-Neo900 | http://susepaste.org/64494183 |
| 01:54.12 | Joerg-Neo900 | so you can't reasonably move them closer together than maybe 2.2mm |
| 01:54.42 | Joerg-Neo900 | feel free to do that, if you see need and opportunity for it |
| 02:11.01 | *** join/#neo900 sulpher1 (~sulpher@200116b860347b00f2def1fffe10229a.dip.versatel-1u1.de) |
| 02:51.27 | houkime | gonna sleep. |
| 03:03.19 | *** join/#neo900 wpwrak (~werner@74-26-17-190.fibertel.com.ar) |
| 03:03.19 | *** mode/#neo900 [+v wpwrak] by ChanServ |
| 03:50.59 | *** join/#neo900 knttl (~knttl@p200300D7B3D49F005604A6FFFECEDAF9.dip0.t-ipconnect.de) |
| 03:53.51 | *** join/#neo900 pagurus (~user@i5387ED04.versanet.de) |
| 04:44.08 | *** join/#neo900 ArturShaik (~artur-sha@212.112.100.88) |
| 05:16.17 | *** join/#neo900 freemangordon (~ivo@46.249.74.23) |
| 06:04.07 | *** join/#neo900 bemyak (~bemyak@195.26.185.118) |
| 06:36.16 | *** join/#neo900 Pali (~pali@Maemo/community/contributor/Pali) |
| 06:59.09 | *** join/#neo900 jonsger (~Thunderbi@200116b82a5d0d006a05cafffe0f7407.dip.versatel-1u1.de) |
| 08:45.49 | *** join/#neo900 qwazix (~qwazix@Maemo/community/ex-council/qwazix) |
| 08:54.15 | *** join/#neo900 Kabouik (~kabouik@158.38.1.70) |
| 08:55.11 | *** join/#neo900 Kabouik (~kabouik@158.38.1.70) |
| 09:17.46 | *** join/#neo900 ArturShaik (~artur-sha@212.112.100.88) |
| 09:23.22 | *** join/#neo900 qwazix (~qwazix@Maemo/community/ex-council/qwazix) |
| 09:23.31 | enyc | Joerg-Neo900: i note, the usb-testpads in the battery-bay, seemingly bypass the l/c/whatever protection/etc circuitry under one of the tin-cans nearer the usb-socket... |
| 09:23.39 | enyc | [on n900] |
| 09:24.20 | enyc | I have n900 with replaced usb connector, usb was working at some point, not worked in ages ... someday i'll take apart and investigate agin... been changing batteries in external charger etc. |
| 09:32.58 | *** join/#neo900 houkime (~houkime@109.252.86.188) |
| 10:02.47 | enyc | houkime: that was a short sleep?? |
| 10:03.31 | houkime | normal one. |
| 10:03.42 | houkime | 7 hours all ok |
| 10:03.53 | houkime | well maybe 6 hrs |
| 10:03.59 | houkime | anyway it worked |
| 10:05.31 | houkime | dreamed about some strange things though. Teleporters, hybernation chambers and amphibian schoolbuses. |
| 10:06.25 | houkime | and now a connector is at 19 degrees and I think it strikes the balance |
| 10:06.55 | houkime | (well not really, it is just that you can't really plug it other way) |
| 10:07.44 | bencoh | huhu |
| 10:35.50 | *** join/#neo900 atk (~Arch-TK@ircpuzzles/staff/Arch-TK) |
| 11:30.30 | *** join/#neo900 jonwil (~jonwil@27-33-80-219.tpgi.com.au) |
| 12:04.09 | *** join/#neo900 wiewo (~wowie@fossencdi.org) |
| 12:13.57 | *** join/#neo900 trx (ns-team@cable-188-2-25-100.dynamic.sbb.rs) |
| 12:13.57 | *** join/#neo900 trx (ns-team@devbin/founder/trx) |
| 12:15.39 | houkime | nope, you can't run a D pair between pogos. pogos are too close to the edge and D pair just won't fit when run between the edge and a pogo |
| 12:16.34 | houkime | oh |
| 12:16.48 | houkime | that'ts not what you meant |
| 12:17.39 | houkime | you wanna do symmetrical as it seems |
| 13:00.47 | *** join/#neo900 Chris_ (~Chris@p4FCA0779.dip0.t-ipconnect.de) |
| 14:07.53 | *** join/#neo900 houkime (~houkime@109-252-86-188.nat.spd-mgts.ru) |
| 14:11.01 | *** join/#neo900 xmn (~xman@user-0cdft2n.cable.mindspring.com) |
| 15:02.51 | *** join/#neo900 Kabouik (~kabouik@158.38.1.70) |
| 15:17.00 | *** join/#neo900 Kabouik (~kabouik@236.34.200.37.customer.cdi.no) |
| 15:23.11 | *** join/#neo900 Kabouik- (~kabouik@236.34.200.37.customer.cdi.no) |
| 15:25.04 | *** join/#neo900 Kabouik_ (~kabouik@115.ip-91-134-138.eu) |
| 16:48.01 | *** join/#neo900 drrty2 (drrty@gateway/vpn/privateinternetaccess/drrty) |
| 16:57.16 | *** join/#neo900 wicket64 (~wicket@dynamic-186-154-117-54.dynamic.etb.net.co) |
| 17:10.17 | *** join/#neo900 jonsger (~Thunderbi@200116b82a5d0d006a05cafffe0f7407.dip.versatel-1u1.de) |
| 17:53.07 | *** join/#neo900 jkepler (~joel@2a01:e35:2f37:400:698f:5fe2:196e:8d27) |
| 18:25.58 | *** join/#neo900 illwieckz (~illwieckz@unvanquished/developer/illwieckz) |
| 20:16.46 | *** join/#neo900 galiven (~galiven@52.119.112.90) |
| 20:18.55 | *** join/#neo900 vdv (~vdv@x4e315dde.dyn.telefonica.de) |
| 20:28.06 | *** join/#neo900 houkime (~houkime@109-252-86-188.nat.spd-mgts.ru) |
| 20:32.59 | *** join/#neo900 wiewo (~wowie@fossencdi.org) |
| 20:48.13 | *** join/#neo900 xes (~xes@unaffiliated/xes) |
| 21:30.48 | *** join/#neo900 jkepler (~joel@2a01:e35:2f37:400:bac7:21b:b552:1653) |
| 22:21.00 | *** join/#neo900 neo900 (~office@neo900/coreteam/joerg) |
| 22:21.00 | *** mode/#neo900 [+v neo900] by ChanServ |
| 22:21.02 | *** join/#neo900 DocScrutinizer05 (~saturn@openmoko/engineers/joerg) |
| 22:21.02 | *** mode/#neo900 [+v DocScrutinizer05] by ChanServ |
| 22:27.05 | *** join/#neo900 chomwitt (~chomwitt@2a02:587:dc01:ed00:3875:c5cc:7222:5c83) |
| 22:41.04 | *** join/#neo900 threebar (~user@unaffiliated/threebar) |