Tuesday, 2019-03-19

*** tpb has joined #tomu00:00
*** nrossi has quit IRC00:35
*** shalzz has quit IRC00:35
*** xobs has quit IRC00:35
*** ptotter[m] has quit IRC00:35
*** shalzz has joined #tomu00:42
*** xobs has joined #tomu01:15
*** nrossi has joined #tomu01:15
*** ptotter[m] has joined #tomu01:15
*** xkapastel has quit IRC01:41
*** Farscrap has quit IRC02:02
*** futarisIRCcloud has joined #tomu03:13
xobsOkay, back in the office.  Today's goals are to (1) validate the test jig and DVT boards, and (2) get SPI working so I can finish up a first draft of Foboot.04:23
*** rohitksingh_work has joined #tomu04:38
*** futarisIRCcloud has quit IRC05:23
*** futarisIRCcloud has joined #tomu06:10
*** davidlt has quit IRC06:50
*** davidlt has joined #tomu06:53
xobsECO1: The factory put the crystals on rotated 180 degrees.07:41
tntEC ?07:42
xobsI mean, they put them on according to the datasheet, which appears to be wrong?  The datasheet seems to indicate that when you can read the text, pin 1 is in the upper-right corner.07:43
xobsEngineering Change Order.  Factory bug report.07:43
*** xkapastel has joined #tomu07:57
*** futarisIRCcloud has quit IRC08:39
xobsHas anyone here gotten the LED to work on a Fomu Hacker Board?08:55
tntAny reason you powered it from 2.5v instead of 3v3 ?  That's a bit low given the Vf of G/B leds.09:01
xobsLuke designed the Hacker Board to be built on a 4/4 process. He did that simply because he couldn't route the 3v3 line to the LEDs :)09:02
xobsOn the DVT (and PVT, eventually) boards, they're using 3v3.09:02
tntgetting rid of 2.5v all together would maybe have freedup some space :p09:03
tntAnyway, was this why you were asking or do you have other trouble with the led ?09:04
tnt(I don't have a tomu so never tested, just curious)09:04
tnts/tomu/fomu/ ...09:04
xobsI'm asking because I'm not able to drive the LED at all, and I'm wondering if anyone has tried it with the WLCSP part.  I'm measuring the pads, and they're just not... padding.09:05
tntare you using the SG_RGBA core ?09:07
xobsI'm using SB_RGBA_DRV.  I've also tried using the pins as GPIO, which didn't work either.09:08
xobsIt seems as though pins C5, B5, and A5 just don't work.09:09
tnt:/  I never tried using them as gpio on the sg48 either. I know they're open-drain output only but not sure if there isn't any bug in nextpnr when handling that. But they sure work fine with SB_RGBA_DRV.09:13
tntDid you try a design with icecube ? To make sure that package doesn't need some other enable bit somewhere that was missed in icestorm ?09:14
xobsThat's what I've noticed with the sg48.  The same code works fine on an EVT board, but on DVT I don't get anything.  I've tried sending the same signal to the touchpads, where I'm able to measure it there.  Nothing wrong with it there.09:15
xobsI just did a design with icecube.  What a strange program. It didn't work either, but it correctly shows that pins A5, B5, and C5 are all routed to the LED.09:37
xobsOkay, I forgot about the "open drain" bit. You're right, tnt, the wires are working. The LED is not.  Now why would that be...09:51
tntwrong pinout / soldered wrong way ... the fact that the pad with the notch in it is pad 4 and not pad 1 sure doesn't help :p10:03
xobsThat's my best guess.  Though it looks correct under the microscope, according to the drawing.10:06
tntyeah, just looked at the gerbers and the picture on the website and looks fine to me as well.10:07
tntdo you see the 2v5 rail ?10:08
xobsAre you looking at the DVT gerbers?10:08
tpbTitle: fomu-hardware/hardware/releases/dvt1e at dvt · im-tomu/fomu-hardware · GitHub (at github.com)10:09
tntlooking at https://github.com/im-tomu/fomu-hardware/tree/master/hacker/releases/v0.0-19-g154fecc10:09
tpbTitle: fomu-hardware/hacker/releases/v0.0-19-g154fecc at master · im-tomu/fomu-hardware · GitHub (at github.com)10:09
xobsThat's the original Hacker Board that mithro did, based on Luke's gerbers.10:10
tntyeah, I thought that's what you asked about10:10
xobsOriginally, yes.  Because I'm quite sure nobody else has tried the RGB LEDs on the DVT boards. :)10:11
xobsI've managed to program a DVT board with a blinky program, so now I'm trying to troubleshoot the LED.10:11
xobsIt's the same device, so they ought to behave the same way.10:11
xobsThe EVT boards used a footprint-compatibie alternative that worked (CL505S), but for some reason these CREE UHD1110 devices just won't light up.10:12
*** davidlt has quit IRC10:19
tntDoes the dvt have blind vias ?10:21
xobsVery much so.  But the only blind via there is 3.3V, which I've measured at the LED point and it's working fine.10:22
*** awe00 has joined #tomu10:30
*** rohitksingh_work has quit IRC12:28
xobsI do see that I swapped pins 1 and 2, which, oops. But that shouldn't affect it because, as you say, the key is on pin 4, and the LED looks oriented properly.12:32
tntDo you have spare leds ? to try and pwer them externally ?12:32
xobsI don't think I have the manual dexterity to hit these pads.12:35
*** xkapastel has quit IRC13:06
*** rohitksingh has joined #tomu13:37
*** xkapastel has joined #tomu13:59
*** rohitksingh has quit IRC14:27
*** rohitksingh has joined #tomu14:44
mithroxobs: FYI - I would accept Fomu related projects in GSoC under SymbiFlow or TimVideos projects...15:48
xobsmithro: that's good to know!15:50
mithroxobs: Could use some help promoting it15:51
xobsmithro: by the way, have you tested out the RGB LED on the hacker board?15:51
mithroxobs: Unsure, I'm pretty sure I have seen it light up...15:51
*** xkapastel has quit IRC16:06
rohitksinghxobs: I've tested it for days :p ...works fine, only issue seems to be uneven brightness16:10
xobsrohitksingh: That's good to know!  Thanks.  Now I need to figure out what the difference is.16:11
rohitksinghFrom the schematic, I can probably guess the reason for uneven brightness...2.5V being dropped by different amount by each color channel and sinking to FPGA IOs?16:13
xobsrohitksingh: yeah, 2.5V is actually under-spec for green and blue.16:14
rohitksinghthat should be the reason then!16:14
xobsOn DVT boards I'll actually need to set the current to super low on the red channel to keep it within tolerance.16:14
rohitksinghyup, but how? extra resistor?16:15
xobsI'll have to look into that Thursday, though.  Tomorrow's going to be a USB-kinda day.16:15
rohitksinghah okay16:15
xobsrohitksingh: by setting the "current level" on the SB_RGBA_DRIVER block.16:16
xobsRGB2_CURRENT possibly. I don't remember which one is tied to Red.16:16
rohitksinghoh...is that an ice40 specific thing? :o Wasn't aware of that16:17
xobsYep!  It even has a fancier control block with registers and a "heartbeat" setting.16:17
xobsSomething to play with later.16:17
rohitksingh oh that's nice! I'll look into it16:18
*** awe00 has quit IRC16:53
*** AmosSam has left #tomu17:16
*** AmosSam has joined #tomu17:16
*** rohitksingh has quit IRC18:38
*** im-tomu has quit IRC18:49
*** im-tomu has joined #tomu18:50
*** im-tomu has left #tomu19:01
*** im-tomu has joined #tomu19:01
*** awe00 has joined #tomu20:44
*** Jmac_ has joined #tomu21:25
*** Jmac_ has quit IRC21:32
*** davidlt_ has joined #tomu22:03
*** davidlt_ is now known as davidlt22:08
*** Farscrap has joined #tomu22:36
*** awe00 has quit IRC22:50

Generated by irclog2html.py 2.13.1 by Marius Gedminas - find it at mg.pov.lt!