Wednesday, 2019-07-03

*** tpb has joined #tomu00:00
*** AmosSam has left #tomu00:01
*** AmosSam has joined #tomu00:01
*** AmosSam has left #tomu02:15
*** AmosSam has joined #tomu02:22
*** xkapastel has quit IRC03:05
*** AmosSam has left #tomu03:34
*** AmosSam has joined #tomu03:42
*** rohitksingh_work has joined #tomu05:17
*** EmilKarlson has joined #tomu06:45
EmilKarlsonthe crowd supply happened, but the order functionality on the page is not closed, does this mean it's still actually open for orders?06:53
xobsEmil Karlson: it is! With Crowd Supply, the "Back It" button turns into a "Preorder" button after the campaign.06:54
EmilKarlsonwould usb-c connector on tomu be simple?07:11
*** rohitksingh has joined #tomu07:12
xobsThere are ideas of how a USB-C connector might work, but no concrete plans.  The form factor is the big unknown.07:17
*** TomKeddie has quit IRC08:26
*** TomKeddie has joined #tomu08:39
tntThis took way longer than I care to admit :p09:06
xobstnt: fancy!  FreeCAD 3D drawing of the RGB LED?09:08
TomKeddiexobs I'm digging into the flash interface a little, I wanted to check the sizes, the schematic says 128Mbit for EVT and DVT but the python code says 16Mbit (2Mbyte) thanks.09:16
xobsTomKeddie: PVT/DVT have 16 Mbit, as does Hacker09:17
TomKeddieIf you have a moment, I had another one about FBM_OFFSET, perhaps you can clarify for me why FBM_OFFSET uses 0x1a000 but this doesn't correspond to any of the multiboot locations?  I realise they are somewhat unrelated but I would have expected them to be aligned just for simplicity?09:20
xobsThe idea is that FBM_OFFSET is where the FoBootMain program lives.  It uses the same CPU as FoBootRecovery.  That is, the ICE40 doesn't do a  SB_WARMBOOT at all.09:22
TomKeddieOk, thanks that makes sense, they're unrelated.  On Arty I'll try to make sensible equivalent locations, just need to think it through. The fpga image is size is obviously a lot bigger but so is the flash.09:25
*** futarisIRCcloud has quit IRC09:38
*** futarisIRCcloud has joined #tomu10:01
*** awe00 has joined #tomu10:07
xobsTomKeddie: this might be useful to copy into the repo somewhere:
tpbTitle: Fomu / ice40UP5K Partition Scheme - Google Sheets (at
*** AmosSam has left #tomu10:12
*** AmosSam has joined #tomu10:12
TomKeddiexobs: that's awesome data - I sent you a PR with a new .md file, I'll try to translate the spreadsheet into the .md tomorrow.10:15
TomKeddieI also sent a PR fixing direct address mapping (well I think it does, I still don't have direct access working on Arty but I'm on it)10:15
xobsTomKeddie: Thanks!  I'll take a look at that PR, too.10:16
xobsIt would be good to get some of the more advanced tools to look at valentyusb.10:16
xobsI just discovered that v1.8.1 works just fine on my laptop, and indeed on lots of other devices, but fails on about 5% of packets on my desktop.10:17
xobsRe-running with a different seed solves it, but that's still worrisome.10:17
TomKeddieI have PCBs for openviszla but they'll just be pcbs for some time I suspect.10:20
TomKeddieI've also done a pmod for the openvizla front end (usb phy) that could be useful, again just pcbs at this point.  openvisla is written in migen though....10:21
tpbTitle: prj-pmod/hardware/pmod-usb-analyser-openvizla at master · TomKeddie/prj-pmod · GitHub (at
TomKeddieseed change implies unconstrained timing requirements I guess?10:23
xobsUsually.  Either that or the timing input into nextpnr isn't accurate.10:25
TomKeddiePerhaps synth the hdl with the lattice tools (or is that a sin?)10:25
xobsNo, that's a really good idea.  It's always good to have more views into a problem.10:26
TomKeddieThere is also the aliexpress usb 2.0 tools mentioned in the teardown talk 4h40m into
xobsTomKeddie: It doesn't fit with the Lattice tools.10:42
*** xkapastel has joined #tomu10:49
TomKeddieIs is close, perhaps you can drop the non essential stuff like the leds/touch etc?  drop the rom/ram size?10:50
TomKeddieOne thing that occurred to me is that the laptop is likely to have a higher quality signal path (all pcb) to the usb controller, on the desktop you have connectors and cable transitions.  Perhaps its worth using a passive cable with the laptop to see if that also produces errors?10:52
TomKeddie(I need to crash, will let you know if I have more thoughts)10:53
xobsTomKeddie: Goodnight!11:12
*** rohitksingh_work has quit IRC11:47
*** rohitksingh has quit IRC11:50
*** alexhw has joined #tomu12:02
*** mifune has quit IRC12:48
*** rohitksingh has joined #tomu12:56
*** xkapastel has quit IRC12:57
*** rohitksingh has quit IRC13:16
*** AmosSam has left #tomu13:29
*** AmosSam has joined #tomu13:31
*** futarisIRCcloud has quit IRC13:38
*** mifune has joined #tomu14:50
*** emeb has joined #tomu17:05
*** AmosSam has left #tomu17:22
*** AmosSam has joined #tomu17:27
*** rohitksingh has joined #tomu17:38
*** awe00 has quit IRC17:50
*** rohitksingh has quit IRC18:33
*** rohitksingh has joined #tomu18:43
*** awe00 has joined #tomu19:37
mithroTomKeddie: If you think having a constructed OpenVizla would mean more hacking on ValentyUSB, then I'll send you one (and you can send me back it if you even construct your own)19:50
tntAlso ...   if you have a usb-> serial adapter and any ice40 you could use that.19:52
tpbTitle: GitHub - smunaut/iua: ice40 USB Analyzer (at
tntIt's FS and not HS but that's enough for ValentyUSB.19:52
mithroxobs: " Resource component uart_0__tx uses physical pin C11, but it is already used by resource component uart_0__rx that was requested earlier"20:23
*** rohitksingh_ has joined #tomu20:27
*** rohitksingh_ has quit IRC20:40
*** rohitksingh has quit IRC20:40
*** futarisIRCcloud has joined #tomu22:04
*** awe00 has quit IRC22:17
*** awe00 has joined #tomu22:38

Generated by 2.13.1 by Marius Gedminas - find it at!