These are chat archives for nightscout/intend-to-bolus

12th
Jun 2018
katie disimone
@Kdisimone
Jun 12 2018 01:37
@cluckj @scottleibrand got the serial working just fine on mac. I did the piBakery OTG serial recipe like shown in the docs (merged now I think).
Kathryns-iMac:~ iMac4K$ cd /dev
Kathryns-iMac:dev iMac4K$ ls *cu*
cu.Bluetooth-Incoming-Port    cu.usbmodem1411
Kathryns-iMac:dev iMac4K$ screen /dev/cu.usbmodem1411
Scott Leibrand
@scottleibrand
Jun 12 2018 01:38
:+1:
Dana Lewis
@danamlewis
Jun 12 2018 01:58
yup, merged
Jon Cluck
@cluckj
Jun 12 2018 02:18
:clap: woo
davidwagaps
@davidwagaps
Jun 12 2018 02:30
OpenAPS wasn't looping. Manual reboot didn't work. Couldn't connect via wifi, couldn't connect via USB. Notice a little piece had broken off of the Explorer board. Switched in a new Explorer board, still can't connect. Also, Intel/Edison combo (not battery) gets pretty hot (with old & new explorer). What should/could/might I do?
Martin Haeberli
@mhaeberli
Jun 12 2018 02:41
@davidwagaps can you send a picture (showing what broke off)? I had the ‘chip antenna' break off the Edison before
davidwagaps
@davidwagaps
Jun 12 2018 02:51
image.png
henghayan
@henghayan
Jun 12 2018 02:52
if (meal_data.carbs) {
        // if carbs / assumedCarbAbsorptionRate > remainingCATimeMin, raise it         
        // so <= 90g is assumed to take 3h, and 120g=4h
        remainingCATimeMin = Math.max(remainingCATimeMin, meal_data.mealCOB/assumedCarbAbsorptionRate);
        var lastCarbAge = round(( new Date().getTime() - meal_data.lastCarbTime ) / 60000);          
        //console.error(meal_data.lastCarbTime, lastCarbAge);

        fractionCOBAbsorbed = ( meal_data.carbs - meal_data.mealCOB ) / meal_data.carbs;
        remainingCATime = remainingCATimeMin + 1.5 * lastCarbAge/60;
        remainingCATime = round(remainingCATime,1);
        //console.error(fractionCOBAbsorbed, remainingCATimeAdjustment, remainingCATime)
        console.error("Last carbs",lastCarbAge,"minutes ago; remainingCATime:",remainingCATime,"hours;",round(fractionCOBAbsorbed*100)+"% carbs absorbed");
    }
oref0/lib/determine-basal/determine-basal.js line 460
Why did the formula remainingCATime = remainingCATimeMin + 1.5 * lastCarbAge/60 ?
Does that mean that the longer lastCarbAge was, the longer it took to absorb?
LilDucky
@LilDucky
Jun 12 2018 03:06
@davidwagaps that was the chip antenna for the radio that talks to the pump. Look at the doco on openaps for instructions on adding an external wire antenna. It will also improve the radio performance.
davidwagaps
@davidwagaps
Jun 12 2018 03:09
@mhaeberli Suddenly I can connect. Getting "no space left on device" now. I'll look at the docs for that.
Martin Haeberli
@mhaeberli
Jun 12 2018 03:09
K
davidwagaps
@davidwagaps
Jun 12 2018 03:11
@LilDucky Thanks - I already use an external antenna. That piece had broken on one side previously and was askew - the connection failed even with the external antenna until I moved the chip antenna back into place. So I assume now that it was gone, that was the connection issue. But maybe it was the "no space" and the antenna was a red herring. Good thing I didn't throw out the old board!
Scott Leibrand
@scottleibrand
Jun 12 2018 04:38
@henghayan yes, but note that the “remaining” carb absorption there is only for projecting future carb absorption that isn’t being observed yet. When you’ve just entered carbs, we can be pretty sure they’ll hit, and quickly. When it’s been hours, we can’t be so sure that any remaining COB is really still there, or if it is, if it will absorb quickly or slowly. So the longer it’s been since the last carb entry, the more slowly we project any remaining carbs to absorb if they’re not already doing so. Usually, once carb absorption starts picking up, the projection of that observed absorption accounts for all of the COB, so remaining carb time only kicks in if your rate of absorption slows way down (for example due to post-meal activity).
davidwagaps
@davidwagaps
Jun 12 2018 05:27
New Explorer suddenly died. Went back to old one, which now lets me log in via USB, but I can't figure out how to free up space. Got "Failed to start Hostname Service" when I rebooted so I can't download ncdu. Somehow my wifi list disappeared. I think I'll reflash...?
Scott Leibrand
@scottleibrand
Jun 12 2018 05:29
Failed to start Hostname Service is harmless: that's not your problem
the docs page should tell you how to delete enough log files to be able to run logrotate
but if you would rather not fix wifi manually, you can re-run the install script (even without reflashing)
davidwagaps
@davidwagaps
Jun 12 2018 05:34
I couldn't figure out how to delete any files - wouldn't let me download ncdu.
Err http://http.debian.net/debian/ jessie/main ncdu i386 1.10-1
Could not resolve 'http.debian.net'
E: Failed to fetch http://http.debian.net/debian/pool/main/n/ncdu/ncdu_1.10-1_i386.deb Could not resolve 'http.debian.net'
Scott Leibrand
@scottleibrand
Jun 12 2018 05:44
why do you think you need ncdu to delete files?
that's what rm is for
ncdu is just a way to see how big things are. you can also do that with ls -larS (for a given directory) and du
davidwagaps
@davidwagaps
Jun 12 2018 05:49
thx
Any idea why my wifi entries disappeared, or if there's any way to recover them?
Scott Leibrand
@scottleibrand
Jun 12 2018 05:58
what's in your wpa_supplicant.conf now?
davidwagaps
@davidwagaps
Jun 12 2018 06:00
completely blank
cat-wifi does nothing. edit-wifi shows a blank file.
Not even the first line with "ctrl_interface"
davidwagaps
@davidwagaps
Jun 12 2018 06:06
I removed remote_syslog_linux_i386.tar.gz - pump-loop is running again. But rig seems to think it's December 1999 now.
Scott Leibrand
@scottleibrand
Jun 12 2018 06:09
Time will be wrong until you get Internet.
davidwagaps
@davidwagaps
Jun 12 2018 06:15
OK I added the "ctrl_interface" line and my home wifi... how long until it connects would you say?
No wifi yet but somehow grabbed onto the BT tether of the phone which is plugged in across the house!
Adi Barilan
@Adidushi
Jun 12 2018 12:08
Hey there! Anyone know what country and frequency this pump would use? The REF is MMT-522LCAL
I need to buy the explorer HAT and I'm not sure which frequency to get. I don't know where the pump is from since I got it from someone. The only thing I know is that it's a 522 and has an FCC ID if that helps
PeetuHongisto
@PeetuHongisto
Jun 12 2018 12:12
CA means Canada?
LilDucky
@LilDucky
Jun 12 2018 12:13
Is from Canada so use North American frequency 916.5MHz
Adi Barilan
@Adidushi
Jun 12 2018 12:14
Alright thanks!
Jon Cluck
@cluckj
Jun 12 2018 13:19
hi adi
Jon Cluck
@cluckj
Jun 12 2018 13:53
@mwilliams is there anything you want to see added to the menus?
Matthew Williams
@mwilliams
Jun 12 2018 14:57
@cluckj Do they update after every loop? I just looked down and realized my status graph view looks pretty static from an hour ago. A simple screen showing the time, current BG with a trend arrow on the top of the display with a small sparkline on the bottom would be a nice view. Trying to think of improvements for the graph view. But with BG and trend arrow larger so it's easier to view at a glance, especially if you're pulling it out of your pocket
Jon Cluck
@cluckj
Jun 12 2018 14:57
the status graph should be updating with every locally-run loop, but I think that's branch-dependent
Matthew Williams
@mwilliams
Jun 12 2018 14:58
I'm currently running dev
the 0.6.2 release cut off I believe
Jon Cluck
@cluckj
Jun 12 2018 15:02
ah, it's not in dev
Matthew Williams
@mwilliams
Jun 12 2018 15:03
ahh, that'd do it :)
Jon Cluck
@cluckj
Jun 12 2018 15:06
my plan is to make a couple different status screens, then have some menu trigger to switch between them
Matthew Williams
@mwilliams
Jun 12 2018 15:09
Likely a waste of power to keep re-drawing on the OLED, but an always visible indicator if it's looping or not (a spinning animation in the corner or something) and a big X so you can quickly identify if there's a problem to be addressed. But I think in most cases, once you're looping, very few things would go wrong in future loops. Unless you had a row of status icons always displayed (NS connection? Battery level? Looping? BG? And put the menu items in two columns below it or something. And I'm jut spitballing off the top of my head, I think the currently menus are great - addition of the display is so great. I was trying to figure out forever how to work an eInk display on the Edison stack (and still looking at how something like this could integrate into the new stack: https://www.ebay.com/i/272838748699?rt=nc&_trkparms=aid%3D222007%26algo%3DSIM.MBE%26ao%3D2%26asc%3D20160908110712%26meid%3De67dc8487c344dd19172b67eb51cc207%26pid%3D100677%26rk%3D12%26rkt%3D30%26sd%3D192341301927%26itm%3D272838748699 - nice tri-color eink!)
And very cool (re: a few status screens, quick triggering between them)
Jon Cluck
@cluckj
Jun 12 2018 15:13
pimoroni has a pHAT version of that :) it's cool but sold out at adafruit
the icon thing was part of my original plan, but at the size required (5x7 pixels) they're unreadable :laughing:
Matthew Williams
@mwilliams
Jun 12 2018 15:18
yeah, it's incredible what's getting packed on the screen now and how readable you can get a font with so few tiny pixels
Eric
@ecc1
Jun 12 2018 15:19
@mwilliams I've had good results with the waveshare 1.54" square e-ink displays: https://gitter.im/openaps/hardware-dev?at=5b022fba41f54f22e24608c5
Simple to use on the Pi, with its multiple SPI buses, but would be a challenge on an Edison, which only has one (which is used for the radio with Explorer)
Matthew Williams
@mwilliams
Jun 12 2018 15:21
Ohhhh very slick!
Jon Cluck
@cluckj
Jun 12 2018 15:22
I tried to pack as much as possible into it, and I'm learning that readable is relative
katie disimone
@Kdisimone
Jun 12 2018 15:40
Dev branch on piHAT rig. Connects to bt just fine but never actually loops. Appears to still be offline as far as pulling BGs. Any fatal flaws in the dev code still for piHAT on an iPhone tether?
I.e., is this working for anyone else yet?
Matthew Williams
@mwilliams
Jun 12 2018 16:07
I'm running dev branch on piHAT rig, but with Enlite CGM... Looping fine for me but no bt for me.
@Kdisimone ^-----
@cluckj Is the code to refresh the graph on each loop sitting in a single commit that I could cherry-pick to dev to give it a test (or should that code make it into master since that appears to be future 0.6.2? (get it in this PR openaps/oref0#1007 ?)
Martin Haeberli
@mhaeberli
Jun 12 2018 16:11
@Kdisimone running piHAT against a test 722 pump with G5 data via share. looks to be working fine...
does BT also
katie disimone
@Kdisimone
Jun 12 2018 16:12
@mhaeberli using iPhone too?
With current dev?
Jon Cluck
@cluckj
Jun 12 2018 16:14
@mwilliams yes-ish: openaps/oref0@b4d3e59
Jon Cluck
@cluckj
Jun 12 2018 16:20
image.png
Dana Lewis
@danamlewis
Jun 12 2018 16:20
Is that BG min or looping min?
Jon Cluck
@cluckj
Jun 12 2018 16:23
I was thinking BG min, but I'll add labels to that and the clock (which will be "last loop at")
cduarte
@cduarte
Jun 12 2018 16:25
Hi!

I do a temp target 80!
And dont get SMB.
I am checking my log and have this line: SMB disabled (no enableSMB preferences active).
But in my pref have

"enableSMB_with_COB": true,
"enableSMB_with_temptarget": true,
"enableUAM": true,
"allowSMB_with_high_temptarget": true

Why dont receiving smb!?

Martin Haeberli
@mhaeberli
Jun 12 2018 16:29
@kdsimone yes bt with iPhone. With 0.7.0-dev in the last week or so. I haven’t tested against -refactored yet. And, as you know, dev is a bit fluid.
Haven’t tested BT against refactored, to be clear
Dana Lewis
@danamlewis
Jun 12 2018 16:31
@cluckj oh good - was thinking we needed last loop, and definitely need labels cause I thought it was a real clock 😅
Or could make that min ago for looping, too
Jon Cluck
@cluckj
Jun 12 2018 16:31
lol
Dana Lewis
@danamlewis
Jun 12 2018 16:31
Otherwise that makes us do extra math and check the time elsewhere
Jon Cluck
@cluckj
Jun 12 2018 16:34
image.png
Matthew Williams
@mwilliams
Jun 12 2018 16:39
:clap:
Digging that!
Jon Cluck
@cluckj
Jun 12 2018 16:40
the actual look on the HAT will be different because FONTS
Dana Lewis
@danamlewis
Jun 12 2018 16:43
Nice 👍🏼
Martin Haeberli
@mhaeberli
Jun 12 2018 16:43
@cluckj have you thought about the relative physical packaging? I have purchased a number of the e-paper displays you mention, but haven’t yet thought about how it would all fit together with the Explorer Pi HAT, etc
Dana Lewis
@danamlewis
Jun 12 2018 16:44
Note to everyone designing screens: save an image somewhere so when we get several, we can add a section to the Pi docs w examples
Andy Sharrow
@dramageek
Jun 12 2018 16:45
@cduarte If I remember, that message is slightly misleading about the preferences. We need more info to figure out what's going on. What info did you give it to expect an SMB?
NiVZ
@NiVZ78
Jun 12 2018 16:50
@cluckj (see below)
image.png
Jon Cluck
@cluckj
Jun 12 2018 16:53
I want to keep it to as few font sizes as possible :)
davidwagaps
@davidwagaps
Jun 12 2018 16:55
Before running the bootstrap script a message says "WARNING: this script will back up and remove all of your current wifi configs."
Where are the wifi configs backed up? Is it possible to restore them?
Dana Lewis
@danamlewis
Jun 12 2018 16:57
no
@davidwagaps copy that file out of myopenaps if you want to keep it.
davidwagaps
@davidwagaps
Jun 12 2018 16:58
thx
Jon Cluck
@cluckj
Jun 12 2018 17:01
@Kdisimone is the pibakery image letting you login as root over serial?
Jon Cluck
@cluckj
Jun 12 2018 17:28
@danamlewis numerical delta_bg and direction arrow?
katie disimone
@Kdisimone
Jun 12 2018 17:30
@cluckj yes
Jon Cluck
@cluckj
Jun 12 2018 17:32
thanks
Jon Cluck
@cluckj
Jun 12 2018 17:59
image.png
(last loop time is fake at the moment)
Matthew Williams
@mwilliams
Jun 12 2018 18:02
@cluckj Ohhhhhh, ahhhhhhhh :clap:
kallnap
@kallnap
Jun 12 2018 18:02
Looks great
Jon Cluck
@cluckj
Jun 12 2018 18:03
I may have been overly optimistic about the arrows :laughing: I'll have to come up with a special arrow font to render those
katie disimone
@Kdisimone
Jun 12 2018 18:17
@cluckj i take that back…had to login as pi. then when i do sudo bash it gives me a prompt like so…root@celtics:/home/pi#
Martin Haeberli
@mhaeberli
Jun 12 2018 18:17
:-( re burn in ...
Jon Cluck
@cluckj
Jun 12 2018 18:18
okay :) that's what I'm seeing too
katie disimone
@Kdisimone
Jun 12 2018 18:18
i’m still perplexed why it’s not looping
bnep0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
        inet 172.20.10.7  netmask 255.255.255.240  broadcast 172.20.10.15
        ether b8:27:eb:94:33:c1  txqueuelen 1000  (Ethernet)
        RX packets 773  bytes 254574 (248.6 KiB)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 663  bytes 86382 (84.3 KiB)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0
Jon Cluck
@cluckj
Jun 12 2018 18:18
can it see the internets?
katie disimone
@Kdisimone
Jun 12 2018 18:19
yes pings google fine
i’ll watch the loop for longer and see what happens
maybe i just had a bad experience last night
suddenly working
good i guess
Jon Cluck
@cluckj
Jun 12 2018 18:23
nice
katie disimone
@Kdisimone
Jun 12 2018 18:24
btw, for the pi0-hat thing…i think we should recommend just skipping all the intro steps and go straight to bluetoothctl for the bt pairing/tethering setup. my iphone is picking up the rig name without any of those steps.
known issue? Dev branch cuts off the info on the top of the Hat’s status graph screen. (can’t see the info that normally i saw above the graph when i was using 0.7.0-refactor)
Jon Cluck
@cluckj
Jun 12 2018 18:28
did it flip around to the bottom?
katie disimone
@Kdisimone
Jun 12 2018 18:29
nope
IMG_2496.jpg
Jon Cluck
@cluckj
Jun 12 2018 18:31
oh, if it just started looping it may have not populated those files yet
katie disimone
@Kdisimone
Jun 12 2018 18:32
i’ll watch it, but i think last night when it was looping for a while on wifi it never populated either. but i’ll give it 15
Jon Cluck
@cluckj
Jun 12 2018 18:35
okay, if it doesn't, there should be some errors about it in pump-loop.log
katie disimone
@Kdisimone
Jun 12 2018 18:41
hmmm…maybe the screen just plain doesn’t update either. time stamp is 8 minutes old, bg value is old. Pump-loop showing no errors (other than could not parse autotune_data, but i’m not running autotune, so that’s fine)
yup, i have to back out of the graph and then open it again to refresh. but also still not showing the top info
NiVZ
@NiVZ78
Jun 12 2018 18:42
@cluckj What font format can it use?
Jon Cluck
@cluckj
Jun 12 2018 18:42
oh right, the screen update code isn't in dev
@Kdisimone openaps/oref0@b4d3e59
NiVZ
@NiVZ78
Jun 12 2018 18:46
Thanks - will have a read - I have much to learn!
katie disimone
@Kdisimone
Jun 12 2018 18:46
:+1:
Jon Cluck
@cluckj
Jun 12 2018 18:46
that didn't get applied to dev, but I'm thinking it should be?
NiVZ
@NiVZ78
Jun 12 2018 18:51
@cluckj - How do people create custom fonts? Would creating something in www.Fontsctruct.com then using http://noopkat.github.io/oled-js-font-foundry/ to convert to .js work?
Jon Cluck
@cluckj
Jun 12 2018 19:02
@NiVZ78 I think so? I haven't tried making new fonts :)
NiVZ
@NiVZ78
Jun 12 2018 19:04
@cluckj Reading the docs just now - understand the idea - just trying to figure out exactly how the format works. The OLED 5x7 seems to use 4 groups of hex vals. Just need to work out how those represent the pixels in rows or columns
Jon Cluck
@cluckj
Jun 12 2018 19:04
same :laughing:
NiVZ
@NiVZ78
Jun 12 2018 19:06
You've probably already found it but this looks useful
http://meow.noopkat.com/oled-js/
Jon Cluck
@cluckj
Jun 12 2018 19:06
that's what the screen uses ;)
Jon Cluck
@cluckj
Jun 12 2018 19:09
ya, that's a more precise writeString :)
mwilliams @mwilliams patches up his dev checkout with the screen update code
NiVZ
@NiVZ78
Jun 12 2018 19:36
I might have a play with an OLED font simulator in JSBin ;)
http://jsbin.com/jajuveb/edit?html,js,output
(As I've not got any hardware (yet) )
Matthew Williams
@mwilliams
Jun 12 2018 19:43
@cluckj That patch didn't quite get me there, /usr/local/bin/oref0-pump-loop: line 85: /root/src/openaps-menu/scripts/status.js: Permission denied when my loop completed - I just gave it a quick chmod and will see how it goes during the next loop
Jon Cluck
@cluckj
Jun 12 2018 19:56
weird. are you running it as user pi?
Matthew Williams
@mwilliams
Jun 12 2018 20:04
I think it's in pi's cron? Even so, changing permissions didn't help:
/root/src/openaps-menu/scripts/status.js: line 2: use strict: command not found
/root/src/openaps-menu/scripts/status.js: line 4: syntax error near unexpected token `('
/root/src/openaps-menu/scripts/status.js: line 4: `const i2c = require('i2c-bus');'
crontab is under root, not pi on my rig
Jon Cluck
@cluckj
Jun 12 2018 20:25
Hmm
Matthew Williams
@mwilliams
Jun 12 2018 20:30
Not super concerned about it atm, though I do think the screen refresh stuff should be in 0.6.2 if @danamlewis is going to cut a release soon, certainly handy to have and lots of folks getting their PiHAT rigs setup it seems
Dana Lewis
@danamlewis
Jun 12 2018 20:45
I don't want to slow 0.6.2 for visuals, but only if we find any core blockers, since 0.6.1 won't work for Pi installs right now
handy stuff can go in 0.7.0
Matthew Williams
@mwilliams
Jun 12 2018 20:52
Ohhh didn't realize 0.6.1 wasn't working for Pi installs, in that case, :shipit:
Dana Lewis
@danamlewis
Jun 12 2018 20:53
yup. checking for any blockers on 0.6.2 for the various installs (carelink, normal edison, etc.) before releasing it
Scott Leibrand
@scottleibrand
Jun 12 2018 21:02
we might also want to move the (very cool) HAT display development efforts over to https://gitter.im/openaps/hardware-dev to keep ITB free for setup and testing discussions
NiVZ
@NiVZ78
Jun 12 2018 21:03

@cluckj I did these quickly so they might be wrong - just map them to any symbol you want to test:

0x4, 0x6, 0x7F, 0x6, 0x4, // Arrow Up
0x10, 0x30, 0x7F, 0x30, 0x10, // Arrow Down
0x20, 0x10, 0xA, 0x6, 0xE, // Arrow 45 Up
0x2, 0x4, 0x28, 0x30, 0x38, // Arrow 45 Down

Scott Leibrand
@scottleibrand
Jun 12 2018 21:03
I'd be fine with merging a cherry-pick PR to dev if there's something easy we can pull out of 0.7.0-dev to make the 0.6.2 HAT display work better (and someone opens the PR today), but I don't want to hold it up for any more coding work.
@NiVZ78 can you cross-post that over to https://gitter.im/openaps/hardware-dev and continue the font discussion there?
Jon Cluck
@cluckj
Jun 12 2018 21:04
^
NiVZ
@NiVZ78
Jun 12 2018 21:04
Yeah sure
Scott Leibrand
@scottleibrand
Jun 12 2018 21:04
thx
Matthew Williams
@mwilliams
Jun 12 2018 21:04
oh nice, happily checks out hardware-dev
parenthetic
@parenthetic
Jun 12 2018 22:33
Hi all, my wifi has stopped working, and I haven't been able to figure out why or how to fix it. Around the same time, my syslog and kernal.log files blew up with errors (each file was taking up almost 250MB of disk space). After some logrotating and deleting some files, I have been able to free up some space and get my rig running again, but only through a BT tether to my iPhone hotspot. Any thoughts on was to diagnose/fix this issue? Almost seems to me like I need to reflash the OS and start fresh... Any alternatives to that? Thanks!
Martin Haeberli
@mhaeberli
Jun 12 2018 22:36
@parenthetic sounds like you’re on an Edison? Take a picture of the board, including the Edison. The chip antenna broke OFF one of my Edisons, and I suspect its wifi is now broken.
parenthetic
@parenthetic
Jun 12 2018 22:49
@mhaeberli Yup, Edison. Where is the chip antenna on the board?
image.png
Scott Leibrand
@scottleibrand
Jun 12 2018 22:51
that white thing in the top left
Martin Haeberli
@mhaeberli
Jun 12 2018 22:51
top left of your picture
looks a bit displaced
do you have a spare edison?
parenthetic
@parenthetic
Jun 12 2018 22:53
I do have a spare edison. But I doubt this is the issue. My wifi dropped out in the middle of the night on Thursday (I can see this from checking the network log).
Martin Haeberli
@mhaeberli
Jun 12 2018 22:54
assuming you’re in the same place where wifi has worked for a long time, and the wifi you’re (trying to) connect to is still up and running, i myself don’t have an alternate hypothesis… perhaps others do
(and of course if your spare edison isn’t already configured, it would have to be set up)
parenthetic
@parenthetic
Jun 12 2018 22:55
Yeah, that's right. Stable connection up until Thursday night.
Given I may have to start fresh, I might as well build a new rig...
Martin Haeberli
@mhaeberli
Jun 12 2018 22:56
I don’t want to cause a problem where there is none, but if the chip antenna is even a little bit loose, that could be a problem.
if the new edison rig can’t connect to your standard wifi, this would suggest a problem upstream , at the wifi...
parenthetic
@parenthetic
Jun 12 2018 22:57
Is there anything to do about a loose chip antenna? Just push??
Martin Haeberli
@mhaeberli
Jun 12 2018 22:57
ideally needs to be re-soldered by a technician who groks surface mount stuff
(ounce of prevention worth a pound of cure - in my case, I have decided to trash that edison, and wished I had had it in a better case)
parenthetic
@parenthetic
Jun 12 2018 22:57
Yeah, that's what I was thinking, too. But I just tested it on a secondary wifi, and it's still no go. (Plus other devices as home can connect to home wifi.)
Martin Haeberli
@mhaeberli
Jun 12 2018 22:58
I’m leaning to something in the WiFi radio on that specific Edison being implicated; the chip antenna is a part of that chain
(based on your information)
parenthetic
@parenthetic
Jun 12 2018 23:00
Yeah, my case isn't special -- a tic-tac box.
image.png
Kinda just love the irony of it, though...
Martin Haeberli
@mhaeberli
Jun 12 2018 23:00
looks promising - does the chip antenna push against the case?
parenthetic
@parenthetic
Jun 12 2018 23:02
Antenna chip seems to press up against the battery inside the case, but most of the pressure is on the power connection, it seems.
Martin Haeberli
@mhaeberli
Jun 12 2018 23:02
k
parenthetic
@parenthetic
Jun 12 2018 23:03
Thanks for your thoughts on this! 'Tis much appreciated.
Martin Haeberli
@mhaeberli
Jun 12 2018 23:05
:+1: long shot, if you haven’t tested it outside the case, try, with the chip antenna up as shown in your photo, vs next to the battery. there is some radio juju that says it’s better to have antennas a bit further away from conductive plates, like the battery
but i’m not offering much hope
Scott Leibrand
@scottleibrand
Jun 12 2018 23:18
does your wpa_supplicant.conf look correct? are there any errors about it in syslog?
Martin Haeberli
@mhaeberli
Jun 12 2018 23:25
:+1:
parenthetic
@parenthetic
Jun 12 2018 23:45
wpa_supplicant.conf looks good, I believe. I'll search the syslog for errors about it though.