| + cobra (~cobra@user/Cobra) | 00:11 | |
| - xktr (QUIT: Quit: leaving) (~xktr@user/xktr) | 00:33 | |
| + reform26428 (~lifning@97.126.153.150) | 00:53 | |
| * reform26428 -> lifning | 00:54 | |
| lifning | hello! i was changing up the keyboard layout on my pocket reform and was nervous about making a mistake, so i made a safety blanket for myself that might be useful to others :) https://cybre.tech/lifning/pocket-reform/commit/40050f0314 | 00:57 |
|---|---|---|
| + lifning_ (~lifning@user/lifning) | 00:59 | |
| - lifning (QUIT: Quit: Leaving) (~lifning@97.126.153.150) | 00:59 | |
| * lifning_ -> lifning | 00:59 | |
| + xktr (~xktr@user/xktr) | 01:07 | |
| - colinsane (QUIT: Quit: bye) (~colinunin@97-113-64-230.tukw.qwest.net) | 01:33 | |
| + colinsane (~colinunin@97-113-64-230.tukw.qwest.net) | 01:38 | |
| chartreuse | Aha found where I put the screws for my bottom lid, time to put it back properly | 02:47 |
| - chartreuse (QUIT: Remote host closed the connection) (~chartreus@S0106908d78501d1d.cg.shawcable.net) | 02:48 | |
| violet | i remember i lost one of the standoffs for my iMX8MQ heatsink for like a year and then it just randomly showed up one day | 02:49 |
| - nsc (QUIT: Ping timeout: 264 seconds) (~nicolas@4-99-142-46.pool.kielnet.net) | 03:08 | |
| + nsc (~nicolas@205-99-142-46.pool.kielnet.net) | 03:09 | |
| + chartreuse (~chartreus@S0106908d78501d1d.cg.shawcable.net) | 03:40 | |
| + tretinha_ (3a571d9f43@2a03:6000:1812:100::1151) | 03:49 | |
| + noam_ (81879d1ffa@2a03:6000:1812:100::dfc) | 03:49 | |
| + romi_ (bd30729973@user/romi) | 03:49 | |
| - tretinha (QUIT: Read error: Connection reset by peer) (3a571d9f43@2a03:6000:1812:100::1151) | 03:49 | |
| * tretinha_ -> tretinha | 03:49 | |
| - romi (QUIT: Read error: Connection reset by peer) (bd30729973@user/romi) | 03:49 | |
| * romi_ -> romi | 03:50 | |
| - noam (QUIT: Read error: Connection reset by peer) (81879d1ffa@2a03:6000:1812:100::dfc) | 03:50 | |
| - theesm (QUIT: Ping timeout: 268 seconds) (~theesm@2a01:4f9:c011:adc3::) | 03:50 | |
| + murphnj (~murph@user/murphnj) | 03:55 | |
| chartreuse | So did a couple simple mods to the built in speakers on mine and they no longer buzz/resonate at high volume levels (>80%). Added a small patch of gaffer tape behind each driver enclosure to make sure there's no room for it to vibrate between the case and 3d printed bracket. Also used some soft adhesive open cell foam to make a gasket around the driver to the front lcd bezel. I suspect the padding | 04:47 |
| chartreuse | behind had the most effect, but I didn't try them independently. | 04:47 |
| chartreuse | But now I can have music going at 100% volume without any annoying buzz | 04:47 |
| chartreuse | I can even push a bit beyond 100% without that resonance, though that starts having clipping distortion of course | 04:48 |
| chartreuse | But the difference between being effectively limited to 75-80% and 100% is quite significant in loudness | 04:49 |
| - lifning (QUIT: Remote host closed the connection) (~lifning@user/lifning) | 05:02 | |
| - mtm (QUIT: Ping timeout: 240 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 06:08 | |
| chartreuse | Also noticed with the headphone detection patch, it's not related to the chip going into shutdown/powersave as after a while of playback it'll swap to headphone too, and it's not releated to linux but the WM8960 | 06:09 |
| chartreuse | Since I have it setup with the gpio input not configured to the linux driver. So somehow the WM8960 is getting confused about jack detect, like somehow the physical pullup is acting weird on it or something | 06:10 |
| chartreuse | Actually just noticed something there in the circuit. The jack detect is on HP-R which has a 47k pulldown on it, while the pullup from the jack detect is also 47k, so when the headphones aren't plugged in you'd expect 1.65v on that pin. I wonder if the pull up needs to be weakened some so the 47k pulldown on the headphone lines can have it lower for a logic low | 06:12 |
| chartreuse | The example schematics for jack detect on the wm8960 page show a jack with a ground pin switch, rather than the tip switch that the Reform uses | 06:15 |
| chartreuse | The input is shown with a schmidt trigger input so I suspect with 50% vcc on there it's just enough to flick from one side to the other | 06:16 |
| chartreuse | Input LOW from the data sheet should be 0.3×DBVDD, DBVDD is 3.3v so it needs to be below 1v on that pin, but we have 1.65 right in the middle between low 0.3x and high 0.7x | 06:18 |
| chartreuse | If the 47k pulldowns are kept that means the pullup needs to be at least 108k rather than 47k | 06:19 |
| chartreuse | Or the pull downs changed to ~20k | 06:19 |
| chartreuse | I don't know if I have any resistors that big in SMD, but at least R143 is not hidden below my bodged in bypass caps | 06:23 |
| chartreuse | er 144 | 06:23 |
| chartreuse | 142 is though (barely) so probably simplist to try increasing R144 to ~150k or such and see if that fixes the jack detect | 06:26 |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 06:31 | |
| * robin_ -> robin | 07:33 | |
| - akira (QUIT: Ping timeout: 264 seconds) (~akira@37.4.230.225) | 07:42 | |
| + akira (~akira@2a01:599:a2b:ed6c:56f9:e9a7:43d4:bc2a) | 07:43 | |
| - akira (QUIT: Read error: Connection reset by peer) (~akira@2a01:599:a2b:ed6c:56f9:e9a7:43d4:bc2a) | 07:53 | |
| + akira (~akira@37.4.230.225) | 07:53 | |
| + op_4 (~tslil@user/op-4/x-9116473) | 09:27 | |
| josch | chartreuse: but if it is a hardware problem, why does it work in 9front as sigrid reports? | 09:39 |
| + gustav28 (~gustav@c-2f35524e.019-141-67626730.bbcust.telenor.se) | 10:02 | |
| - mtm (QUIT: Ping timeout: 260 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 10:35 | |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 10:49 | |
| - mtm (QUIT: Ping timeout: 264 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 10:54 | |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 11:07 | |
| - mtm (QUIT: Ping timeout: 268 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 11:15 | |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 11:35 | |
| + andreas-e (~Andreas@2001:861:c4:f2f0::c64) | 11:44 | |
| - andreas-e (QUIT: Client Quit) (~Andreas@2001:861:c4:f2f0::c64) | 11:44 | |
| - mtm (QUIT: Ping timeout: 268 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 11:48 | |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 12:00 | |
| - colinsane (QUIT: Ping timeout: 255 seconds) (~colinunin@97-113-64-230.tukw.qwest.net) | 12:06 | |
| + theesm (~theesm@2a01:4f9:c011:adc3::) | 12:14 | |
| - mtm (QUIT: Ping timeout: 264 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 12:18 | |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 12:24 | |
| - mtm (QUIT: Ping timeout: 268 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 12:39 | |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 12:50 | |
| - mtm (QUIT: Ping timeout: 272 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 13:00 | |
| + colinsane (~colinunin@97-113-64-230.tukw.qwest.net) | 13:00 | |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 13:08 | |
| - mtm (QUIT: Ping timeout: 260 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 13:14 | |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 13:20 | |
| - mtm (QUIT: Ping timeout: 264 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 13:25 | |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 13:38 | |
| - mtm (QUIT: Ping timeout: 246 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 13:43 | |
| sigrid | josch: chartreuse: the ultimate test would be to boot 9front on the same machine to make sure :D | 13:50 |
| sigrid | but yeah, it works for me | 13:50 |
| sigrid | i did not do any hw mods either | 13:51 |
| josch | sigrid: 9front only supports the imx8mq and not the a311d, right? | 13:54 |
| sigrid | yep | 13:56 |
| josch | bummer -- i should've tried this out before giving my imx8mq away :( | 13:58 |
| - op_4 (QUIT: Remote host closed the connection) (~tslil@user/op-4/x-9116473) | 14:33 | |
| + op_4 (~tslil@user/op-4/x-9116473) | 14:36 | |
| chartreuse | Dunno, maybe mine is just marginal at that voltage. | 15:22 |
| chartreuse | Is there a pre-made 9front image for cf card (ideally one that also contains a media player since I'm not the most familiar with the plan 9 ecosystem? | 15:23 |
| chartreuse | I will try bodging the hardware mod in though to at least eliminate that as a problem, since it doesn't seem right to have 1.65v on a detection pin for low | 15:23 |
| chartreuse | Alright downloading 9front, I'll give it a test in a bit before swapping the resistor | 15:53 |
| sigrid | chartreuse: yes. there is a usb image | 15:56 |
| sigrid | the player is zuke | 15:57 |
| sigrid | but i guess you could also just "play file.mp3" for a test | 15:58 |
| sigrid | to see what the system thinks about jack: cat /dev/audiostat | 15:59 |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 16:05 | |
| chartreuse | Should be able to do the jack detect on the A311D I suspect, just need to find the right register values to set the pin to gpio and then assign it to headphone detect | 16:25 |
| chartreuse | Not conclusive yet, but the first time I ran cat /dev/audiostat after booting it was showing HP 1 even though I didn't have headphones on, though plugging them in kept it at 1, then went to 0 when removing it. Now just waiting to see if it'll flip | 16:45 |
| josch | minute: i found an even better way to test the disk images in qemu -- with graphical user interface! | 17:19 |
| josch | minute: here is the first boot-up of a pocket reform system image for a311d: https://mister-muffin.de/p/Wt07.png | 17:20 |
| josch | minute: and since this is vnc i can interact with it: https://mister-muffin.de/p/gjRW.png | 17:21 |
| josch | i think this is superior to my current solution of starting everything inside podman because this way, a full boot with systemd is performed and instead of running reform-setup-wizard in podman explictly, this tests whether the scripts in the system image are indeed set up such that the setup wizard starts on boot | 17:22 |
| josch | minute: this way we can even do clicks in sway and test the setup: https://mister-muffin.de/p/8hh7.png | 17:25 |
| josch | the qemu magic is -device virtio-gpu-pci,edid=on,xres=1024,yres=768 -vnc :94,share=force-shared | 17:26 |
| chartreuse | Well it hasn't flipped yet, even after I figured out how to mount a linux partition with a mp3 file on it. It showing 1 that first boot is kinda odd and might hint that there still is a bit of hardware oddness, but it's much more stable than linux is at not flipping | 17:28 |
| [tj] | ACTION tries to click the close button in josch's screenshot | 17:34 |
| [tj] | josch: that is really cool, are the qemu models good enought to test hardware specifics image boot or only desktop? | 17:35 |
| josch | [tj]: no hardware specifics at all | 18:03 |
| sigrid | chartreuse: re it being 1 at boot, was it a complete power shutdown before you booted into 9front? | 18:33 |
| sigrid | ie with power rails going off and on | 18:33 |
| chartreuse | Yes, I did a poweroff from linux, then from the circle menu did a power on | 19:32 |
| chartreuse | Even though I'm not sure on the resistor values, the bug seems not to be present in 9front though, so something is off with the linux config. Possibly a register config or such | 19:32 |
| chartreuse | That poweroff one also seems to affect linux too, just had that after booting back into it. So likely something with the initial state of the wm8960 | 19:37 |
| chartreuse | Right now I have linux setup without the gpio (so just the wm8960 doing it) and I notice that it plays for maybe a minute, then has a studder every couple seconds, before eventually flopping over to the headphone | 19:38 |
| chartreuse | Which is why it feels like it's related to the pulldowns. Could also be something with what mode the wm8960 is being driven in such how the headphone outputs are driven when not selected | 19:39 |
| chartreuse | There is a register for discharging the DC blocking caps on the headphones, is 9front doing that when it swaps from headphones to speakers? | 19:45 |
| chartreuse | Which would basically add a 400 ohm pull down on those pins which would help keep it in speaker | 19:46 |
| chartreuse | The linux driver doesn't look like that ever gets enabled, the only code with DISOP is disabling it for the lowest power state | 19:49 |
| chartreuse | Alright replaced R144 with a bodged on small through hole 220k resistor. I've got some 0603 ones in my digikey cart if this does work. | 20:43 |
| josch | how do you order at digikey without either insane shipping costs or without filling a huuuuge cart? | 20:45 |
| josch | asking for a friend :) | 20:45 |
| sigrid | chartreuse: 9front doesn't do anything with headphones interrupt | 20:45 |
| sigrid | it's just there | 20:45 |
| chartreuse | Hmm, I wonder if there's some other register differences, do you have a list of the default register values it sets? Maybe a source link to that file? | 20:45 |
| chartreuse | With the weaker pull-up resistor I'm not getting that swapping behaviour yet. Just played a 5 minute song without it cutting out | 20:46 |
| sigrid | wm8960 spec has all default values listed | 20:47 |
| sigrid | and in 9front driver I specified in the comments what it does exactly | 20:48 |
| chartreuse | Alright, might just need to compare that with what the linux driver does, since something must be different. Though changing R144 for 220k does seem to have also fixed the issue | 20:50 |
| - nsc (QUIT: Ping timeout: 255 seconds) (~nicolas@205-99-142-46.pool.kielnet.net) | 20:59 | |
| - gustav28 (QUIT: Quit: Quit) (~gustav@c-2f35524e.019-141-67626730.bbcust.telenor.se) | 21:02 | |
| minute | josch: idk about digikey but at mouser you "only" need 50 eur | 21:16 |
| minute | josch: maybe an alternative https://www.segor.de//#/versand/versand-in-deutschland | 21:17 |
| chartreuse | Digikey canada at least it's 100 CAD for free shipping, otherwise 8 CAD | 21:17 |
| minute | josch: what kind of part are you looking for? | 21:18 |
| minute | josch: very cool @ qemu | 21:19 |
| + gustav28 (~gustav@c-2f35524e.019-141-67626730.bbcust.telenor.se) | 21:19 | |
| minute | chartreuse: very interesting @ speaker mod, would you share some pix? | 21:20 |
| chartreuse | I put a picture of the foam gasket in a community post, though didn't take a picture of the bit of gaffer tape behind the speaker | 21:21 |
| chartreuse | Or are you meaning the resistor swap? | 21:22 |
| minute | chartreuse: ah, this is what i meant @ foam + gaffer. didn't see the picture when i checked earlier | 21:44 |
| josch | minute: just a general question as there have been situations in the past (like for the pocket reform speaker) where i was unable to find a source that didn't have a 50 EUR min-buy or similar | 21:51 |
| josch | minute: about qemu+vnc, the real solution is called "openqa" which essenitally is doing the same thing my script does (run qemu, connect via vnc, take screenshots, do OCR, do mouse clicks) but in professional. That softare is used for the debian-installer for continous integration and i want to use it for the reform as well. Here is an example of openqa installing cinnamon on aarch64 debian: | 21:53 |
| josch | https://openqa.debian.net/tests/280556 (wait a bit for the screenshots to appear) | 21:53 |
| josch | i'm talking to the admins and they want to give me access to the infrastructure for the purpose of MNT Reform integration into debian | 21:54 |
| minute | josch: oh nice that's sophisticated looking | 22:01 |
| - akira (QUIT: Read error: Connection reset by peer) (~akira@37.4.230.225) | 22:10 | |
| + akira (~akira@37.4.230.225) | 22:11 | |
| - gustav28 (QUIT: Quit: Quit) (~gustav@c-2f35524e.019-141-67626730.bbcust.telenor.se) | 22:15 | |
| josch | minute: oh segor.de looks good! thank you for the tip! I wish we had an electronics component store in my home town :) | 22:18 |
| + nsc (~nicolas@205-99-142-46.pool.kielnet.net) | 22:19 | |
| - akira (QUIT: Ping timeout: 264 seconds) (~akira@37.4.230.225) | 22:20 | |
| + akira (~akira@2a01:599:a0d:36c2:a577:fef4:bcc6:6fb) | 22:21 | |
| - akira (QUIT: Read error: Connection reset by peer) (~akira@2a01:599:a0d:36c2:a577:fef4:bcc6:6fb) | 22:22 | |
| + akira (~akira@37.4.230.225) | 22:22 | |
| chartreuse | minute: The tape part is just a tiny 10x10mm or so square put into the back of the lid right behind the speaker. | 22:38 |
| minute | chartreuse: ok! | 22:44 |
| chartreuse | byns on the forums just said they used some thin adhesive foam for the back mounting too, really just needs to be something thin with a bit of give. Likely is mostly just taking out any slack in the mounting, while also adding a bit of dampening | 22:45 |
| minute | ok, we'll try that | 22:57 |
| minute | personally i have the dual boxed speaker mod in my reform but always interested in how to make the standard model better | 22:57 |
| + jacobk (~quassel@2600:1700:9e1e:7800:b4a1:ccf2:3d91:1e21) | 23:23 | |
| - mjw (QUIT: Ping timeout: 264 seconds) (~mjw@gnu.wildebeest.org) | 23:43 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
