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

20th
Jan 2017
JELCRAWFORD
@JELCRAWFORD
Jan 20 2017 00:25
Ok - will see if I can find it in the carelink upload. We have not associated it with anything to date. Came in last night at 12:30 last - no suspend, batt good, reservoir good. Kid was sleeping..... Other than running Openaps, nothing was new/different. I'll see though what I find in Carelink - never thought to look at that - had spent some time consulting Google to no avail.... Thx!
Brian Rabinovitz
@bdr1177
Jan 20 2017 00:42
Is there a way to remotely control the edison?
Brian Rabinovitz
@bdr1177
Jan 20 2017 00:48
For example, sending a reboot or shutdown command?
Scott Leibrand
@scottleibrand
Jan 20 2017 00:53
you'd need to log into it via ssh (or usb serial) to do that
I have my edisons configured to launch ssh tunnels to my digitalocean server, and from there I can log into them remotely
there are other ways to accomplish the same thing if you don't have a server
fiberfan
@fiberfan
Jan 20 2017 00:54
Scott, I have duplicated the fresh install from dev problem on a pi3. I used a sd card image from last November that was just the OS. Ran apt-get update and apt-get upgrade before running the steps in the docs to setup openaps. The only thing I did differently was checking out dev before running the install. I get a failure when the loop is running because monitor.iob is empty. I ran into this early this week when I trying to put dev on one of my edisons and just reverted to master. The pi3 is not in use at present and therefore perfect to troubleshoot.
Brian Rabinovitz
@bdr1177
Jan 20 2017 00:55
So I'm just thinking about when my daughter is at school on school wifi and I'm at work. It would be helpful to be able to look at the logs remotely and reboot if necessary.
fiberfan
@fiberfan
Jan 20 2017 00:55
I will have time tomorrow or on the weekend to dig in deeper but I do have some time you could troubleshoot remotely if you want and actually have time for that.
Scott Leibrand
@scottleibrand
Jan 20 2017 00:56
@bdr1177 something like https://openport.io/ might work, or you could do something like https://www.howtoforge.com/reverse-ssh-tunneling or https://raymii.org/s/tutorials/Autossh_persistent_tunnels.html to your own server
@fiberfan sure, let's take a look. can you paste the exact error again?
Brian Rabinovitz
@bdr1177
Jan 20 2017 00:57
@scottleibrand Sounds good. I will check those out.
Scott Leibrand
@scottleibrand
Jan 20 2017 00:58
@bdr1177 there is another service that works similar to openport.io, but I can't remember its name. so doing some more searching might worthwhile if it's not quite suitable for your use case
fiberfan
@fiberfan
Jan 20 2017 00:59
''''
glucose.json newer than pumphistory: Refreshed pumphistory
Could not parse input data: [SyntaxError: /home/pi/js/monitor/iob.json: Unexpected end of input]
Scott Leibrand
@scottleibrand
Jan 20 2017 01:00
ok, and cat /home/pi/js/monitor/iob.json?
maybe also cd ~/js/ and openaps report invoke monitor/iob.json too, and then cat again
fiberfan
@fiberfan
Jan 20 2017 01:02
Nothing when I cat iob.json

iob://text/shell/monitor/iob.json

module.js:485
throw err;
^
SyntaxError: /home/pi/js/settings/profile.json: Unexpected end of input
at Object.parse (native)
at Object.Module._extensions..json (module.js:482:27)
at Module.load (module.js:356:32)
at Function.Module._load (module.js:312:12)
at Module.require (module.js:364:17)
at require (module.js:380:17)
at Object.<anonymous> (/usr/local/lib/node_modules/oref0/bin/oref0-calculate-iob.js:43:22)
at Module._compile (module.js:456:26)
at Object.Module._extensions..js (module.js:474:10)
at Module.load (module.js:356:32)
reporting monitor/iob.json

From openaps report invoke monitor/iob.json
Cat still returns nothing
Scott Leibrand
@scottleibrand
Jan 20 2017 01:07
ok, so next we need to look at profile.json then, since it's complaining about SyntaxError: /home/pi/js/settings/profile.json: Unexpected end of input
so cat settings/profile.json and then openaps report invoke settings/profile.json and do it again
fiberfan
@fiberfan
Jan 20 2017 01:08
profile.json is empty

get-profile://text/shell/settings/profile.json
/usr/local/bin/oref0-get-profile pump_settings.json bg_targets.json insulin_sensitivities.json basal_profile.json [preferences.json] [<carb_ratios.json>] [<temptargets.json>] [--model model.json]

Options:
--model, -m Pump model response [default: false]
--help Show help [boolean]
--exportDefaults Show default preference values [default: false]
--updatePreferences Check for any keys missing from current prefs and add from defaults [default: false]

Unknown argument: autotune
reporting settings/profile.json

When i run the report
profile.json is still empty
Scott Leibrand
@scottleibrand
Jan 20 2017 01:23
ok. so looks like this version of oref0-get-profile is old and doesn't support the --autotune argument
can you do npm list -g oref0 | egrep oref0 ?
fiberfan
@fiberfan
Jan 20 2017 01:24
0.3.3
Scott Leibrand
@scottleibrand
Jan 20 2017 01:27
ok, that is our problem. now to figure out why it isn't upgrading
can you run npm list -g oref0 | egrep oref0@0.3.[6-9] || echo Need to upgrade ?
oh, I see the problem
https://www.npmjs.com/package/oref0 has 0.3.3 as the latest version, so this line doesn't upgrade it past that:
npm list -g oref0 | egrep oref0@0.3.[6-9] || (echo Installing latest oref0 && sudo npm install -g oref0)
ok, I just pushed a fix for you to test
Scott Leibrand
@scottleibrand
Jan 20 2017 01:32
can you cd ~/src/oref0 && git pull to get it?
and then re-run oref0-setup and it should install it correctly this time
fiberfan
@fiberfan
Jan 20 2017 01:33
Just need to clear out crontab and then I will re-run
Scott Leibrand
@scottleibrand
Jan 20 2017 01:36
should also be ok to leave cron as-is
we're not changing anything there, as long as you re-run oref0-setup with the same options as last time, you can skip the cron bits and reuse what you have
fiberfan
@fiberfan
Jan 20 2017 01:47
I was tailing the log and there an error from monitor/iob.json. I then ran openaps invoke report settings/profile.json and then openaps invoke report monitor/iob.json. Both ran without an error and the loop is now running successfully. If you want me to try again and wait a couple of minutes to see if the monitor/iob.json error will fix itself I can.
Aaron Michelson
@Aaybob
Jan 20 2017 01:48
...to_xlsx.py => oref0-autotune-export-to-xlsx.py} | 67 ++++++++++++++++++----
bin/oref0-autotune.sh | 4 +-
bin/oref0-setup.sh | 5 +-
lib/autotune/index.js | 8 +++
lib/iob/total.js | 3 +-
lib/profile/basal.js | 4 ++
lib/profile/index.js | 8 +++
package.json | 5 +-
tests/iob.test.js | 16 ++++--
:+1:
Scott Leibrand
@scottleibrand
Jan 20 2017 01:50
@fiberfan ok, sounds like it fixed it then. LMK if the loop doesn't clear itself up after 10-15 minutes (when it refreshes Settings)
fiberfan
@fiberfan
Jan 20 2017 01:56
I should have time tonight or tomorrow to run from just the OS again to make sure it all works now. (I have been writing software for 20+ years with a good tester(s) for the last 10 and some of good testing practices have rubbed off :smile: )
Scott Leibrand
@scottleibrand
Jan 20 2017 01:58
:+1: thx
maybe you'd like to help whip our unit tests into shape too? :-)
Aaron Michelson
@Aaybob
Jan 20 2017 02:01
Dev install uses Bluez -5.37 from the year before last. Intentionally not 5.43 from December last year?
Scott Leibrand
@scottleibrand
Jan 20 2017 02:02
If someone wants to test bluez 5.43 and can get it to work reliably, I'd be happy to merge a PR updating that
Aaron Michelson
@Aaybob
Jan 20 2017 02:03
will look at ~/src/oref0/bin/oref0-setup.sh
@ryan_anglani_twitter ahhw you're taking care of someone! How did you program cron? With PREFLIGHT?
Eric
@ecc1
Jan 20 2017 02:09
@JELCRAWFORD @scottleibrand looking at @ps2's code, 0x69 seems to be a bolus reminder
Chris Oattes
@cjo20
Jan 20 2017 02:40
Just wrote this: openaps/oref0#338 Which should improve the note text for SensorAlarms we have decoded properly
Aaron Michelson
@Aaybob
Jan 20 2017 03:07
Re-installed dependencies and dev oref0 paired phone with root@red:~# bluetoothctl
[bluetooth]# power on
[bluetooth]# agent on
[bluetooth]# default-agent
[bluetooth]# scan on
Discovery started
[NEW] Device 1C:5C:F2:72:B7:60 DiabeticAaron
[bluetooth]# trust 1C:5C:F2:72:B7:60
[bluetooth]# pair 1C:5C:F2:72:B7:60
Request confirmation
[agent] Confirm passkey 157646 (yes/no): yes
[CHG] Device 1C:5C:F2:72:B7:60 Paired: yes
Pairing successful
Disconnected internet in Edison, hardwired cgm receriver to otg, plugged in large battery to uart
wait forgot to turn off successful BLE
Aaron Michelson
@Aaybob
Jan 20 2017 03:16
wait again forgot to turn off second working unit
R. Dennis Steed
@rdsteed
Jan 20 2017 03:36
Each day PACT Publishing offers a free computer related ebook. Today's (OpenAPS relevant) ebook is "Mastering Embedded Linux Programming" with 20 1/2 hours left to claim: https://www.packtpub.com/packt/offers/free-learning.
Aaron Michelson
@Aaybob
Jan 20 2017 03:39
Thanks. Got it.
Scott Leibrand
@scottleibrand
Jan 20 2017 03:43
nice. thanks @rdsteed - downloaded it. looks like it was written by the guy who did Yocto Linux.
Aaron Michelson
@Aaybob
Jan 20 2017 04:02
Edison dev looping on BLE and off internet with hardwired cgm in charge mode.
JELCRAWFORD
@JELCRAWFORD
Jan 20 2017 04:34
@ecc1 brilliant! Yes, he does have the bolus reminder turned on, and while it doesn't show up all the time, I bet your are absolutely right - Minor mystery solved! Thx!
Scott Leibrand
@scottleibrand
Jan 20 2017 04:35
if you'd like to get decocare updated to report those properly show they show up in NS right, I'm sure there are some folks here who could help you make that PR
I hasn't really ever gotten that deep into the python codebase myself
chrisharris751
@chrisharris751
Jan 20 2017 04:45
Is there a way to turn off Wi-Fi completely on the Edison/explorer rig to save battery?
Aaron Michelson
@Aaybob
Jan 20 2017 06:02

is there more help on this "device"?
usage: openaps-use pebble [-h] USAGE ...
optional arguments:
-h, --help show this help message and exit

Device pebble:

vendor openaps.vendors.process
process - a fake vendor to run arbitrary commands
USAGE Usage Details
shell run a process in a subshell

Scott Leibrand
@scottleibrand
Jan 20 2017 06:04
as far as openapsis concerned, that "device" is nothing more than a shell out to the appropriate command
if you want to figure out what the command is/does, you'll need to look at its --help or man page
openaps device show pebble should show you what the command is, and then you can run that manually on the command line to learn more
Aaron Michelson
@Aaybob
Jan 20 2017 06:05
@chrisharris751 you did a standard install following the docs, then edit /var/wpa_supplicant/wpa_supplicant.conf to take out every network{}
Scott Leibrand
@scottleibrand
Jan 20 2017 06:06
do you know whether that actually shuts off the wifi radio completely?
Aaron Michelson
@Aaybob
Jan 20 2017 06:08
to save battery....so even if not used it may be in an active mode?
Scott Leibrand
@scottleibrand
Jan 20 2017 06:12
I'm not sure
Aaron Michelson
@Aaybob
Jan 20 2017 06:15
There is not much help for setting up PANCREABBLE
Scott Leibrand
@scottleibrand
Jan 20 2017 06:15
Nope, that one hasn't been well documented yet
Aaron Michelson
@Aaybob
Jan 20 2017 06:16
Who designed it?
Scott Leibrand
@scottleibrand
Jan 20 2017 06:21
Dana Lewis
@danamlewis
Jan 20 2017 06:25
Mark is really awesome, and responsive to issues, so logging issues in that repo is best if you have questions or problems with it.
Aaron Michelson
@Aaybob
Jan 20 2017 06:27

Found something:
root@red:~/myopenaps# openaps use pebble shell -h
usage: openaps-use pebble shell [-h] [--not-json-default]
glucose iob basal_profile temp_basal suggested
enacted meal ...

run a process in a subshell

positional arguments:
glucose
iob
basal_profile
temp_basal
suggested
enacted
meal
remainder

optional arguments:
-h, --help show this help message and exit
--not-json-default When the process does not produce json.

Dana Lewis
@danamlewis
Jan 20 2017 06:28
(Related on the list of awesome things he's developed, in addition to pancreabble and urchin, also check out Nightscout osx menubar widget, for those of you with macs)
Scott Leibrand
@scottleibrand
Jan 20 2017 06:28
are you trying to figure out how the pancreabble command works, or how your pebble openaps device is configured?
Aaron Michelson
@Aaybob
Jan 20 2017 06:29
I'm wanting to install it.
Scott Leibrand
@scottleibrand
Jan 20 2017 06:31
are you using oref0-setup from the oref0 dev branch to install it, or doing it manually?
Aaron Michelson
@Aaybob
Jan 20 2017 06:31
ok i figured out how to "new issue" there
Scott Leibrand
@scottleibrand
Jan 20 2017 06:32
openaps/oref0#329 is relevant if you're trying to do it with oref0-setup
Aaron Michelson
@Aaybob
Jan 20 2017 06:32
I saw the dev install put it in but I can't access it or know how to know anything about it, so I tried to follow the document but the commands are just suggestions not commands.
Scott Leibrand
@scottleibrand
Jan 20 2017 06:32
we also need a volunteer to add that info to the pancreabble page in the docs
if you're still having trouble, maybe @Bender1061 can help
chrisharris751
@chrisharris751
Jan 20 2017 06:41
@scottleibrand @Aaybob thanks, I've done that already. It's doesn't actually shut down the radio completely. As I posted the other night, with no network listed in wpa_suppliacnt the Edison shows up as an open network called Chrison any near
Aaron Michelson
@Aaybob
Jan 20 2017 06:41
Where would peb-urchin-status.sh
likely be located?
chrisharris751
@chrisharris751
Jan 20 2017 06:42
Nearby.
At least I'm all the Android devices in my house.
Sorry the Android gitter app and auto correct suck!
Aaron Michelson
@Aaybob
Jan 20 2017 06:45

The command-line equivalent of unchecking the nm-applet's 'Enable Wi-fi' should be

nmcli nm wifi off
To re-enable, use

nmcli nm wifi on
If you need to permanently disable a particular device, then you can do that programatically using the unmanaged-devices feature in /etc/NetworkManager/NetworkManager.conf or (at a lower level) by blacklisting or removing the corresponding driver.

That doesn't work
jmcrawford
@jmcrawford
Jan 20 2017 07:01
There is a discussion on disabling wifi here: https://communities.intel.com/thread/60219
Scott Leibrand
@scottleibrand
Jan 20 2017 07:01
that uses yocto tools though
I wonder if there's a power management setting that would turn it (most of the way) off
maybe try iwconfig wlan0 txpower off ?
       txpower
              For cards supporting multiple transmit powers, sets the transmit power in dBm. If W is the power in
              Watt, the power in dBm is P = 30 + 10.log(W).  If the value is postfixed by mW, it will be automat‐
              ically converted to dBm.
              In addition, on and off enable and disable the radio, and auto and fixed enable and  disable  power
              control (if those features are available).
              Examples :
                   iwconfig eth0 txpower 15
                   iwconfig eth0 txpower 30mW
                   iwconfig eth0 txpower auto
                   iwconfig eth0 txpower off
Aaron Michelson
@Aaybob
Jan 20 2017 07:03
can you experiment with commenting out #auto wlan0 in network interfaces?
jmcrawford
@jmcrawford
Jan 20 2017 07:04
How are you planning to measure the effect, whether it has an impact on power usage or not?
Aaron Michelson
@Aaybob
Jan 20 2017 07:06
Has concern that it shows up on his local network
jmcrawford
@jmcrawford
Jan 20 2017 07:06
I suspect that unconfiguring (or not configuring) wlan0 won't have any effect, the radio is still going to be on. But that iwconfig command looks promising
Aaron Michelson
@Aaybob
Jan 20 2017 07:08
power management should have a sleep mode...
Tim Street
@tim2000s
Jan 20 2017 07:13
@Aaybob I was playing with that last night. I was able to manually add a pancreabble device, and the pebble. The peb-urchin-status.sh is in the dev setup lib folder. I've not been able to get it to do anything though.
Aaron Michelson
@Aaybob
Jan 20 2017 07:16
@tim2000s @Bender1061 says unable to make the script write a .json file. Maybe Pancreabble could send a graph of temp basal in text format.
Aaron Michelson
@Aaybob
Jan 20 2017 13:40
@tim2000s peb-urchin-status.sh is in the dev setup lib folder Where?
Brian Rabinovitz
@bdr1177
Jan 20 2017 13:42
Is there a command or script to check the temperature that the edison is running at?
Aaron Michelson
@Aaybob
Jan 20 2017 13:43
https://github.com/openaps/dev/setup/lib root@red:/dev# ls ::::shm spidev5.1 stderr stdin stdout @tim2000s
Brian Rabinovitz
@bdr1177
Jan 20 2017 13:52
Okay. So is it a tool that I need to install?
Aaron Michelson
@Aaybob
Jan 20 2017 13:55
@bdr1177 I'm not too sure....I tried to follow the discussion at http://unix.stackexchange.com/questions/205525/script-to-log-temperature that discussion uses the command "vcgencmd". Is there a where or which command to know if we have that command?
@bdr1177 What about https://blog.bartbania.com/raspberry_pi/temperature-log-howto/ looks like you have to buy a sensor and solder it in place....
Aaron Michelson
@Aaybob
Jan 20 2017 14:01
using RRD to log disk drive temperatures @bdr1177 http://martybugs.net/linux/hddtemp.cgi
Brian Rabinovitz
@bdr1177
Jan 20 2017 14:02
@Aaybob Well, I found this https://communities.intel.com/message/259293#259293, which says to use cat /sys/class/thermal/thermal_zone3/temp and cat /sys/class/thermal/thermal_zone4/temp and divide by 1000 to get the temperatures of the two cpu cores. In my case, I get 85 Celsius and 86 Celsius, respectively, which seems too hot.
Assuming I'm on the right track, which I'm not sure about.
Aaron Michelson
@Aaybob
Jan 20 2017 14:03
mine says 56000
Brian Rabinovitz
@bdr1177
Jan 20 2017 14:04
That sounds much better. Do you have it in a case or just out in the open?
Aaron Michelson
@Aaybob
Jan 20 2017 14:04
there should be a farenheit option
sitting on the desk with space around it with a cord attached to both connectors and a fully charged battery plugged in
It is looping but my sensor is in two hour warmup mode and it feels particularly cool
Brian Rabinovitz
@bdr1177
Jan 20 2017 14:06
I built a temporary case, so I was trying to see if that would cause it to run too hot. Should have gotten some numbers ahead of time to compare. I'll take it out and let it cool down and check again.
Aaron Michelson
@Aaybob
Jan 20 2017 14:07
you know the operational temperature range?
Brian Rabinovitz
@bdr1177
Jan 20 2017 14:08
trying to look it up now
Specs say operating temperature should be 32 to 104°F (0 to 40°C)
Brian Rabinovitz
@bdr1177
Jan 20 2017 14:14
So I guess I'm running way too hot.
Chris Oattes
@cjo20
Jan 20 2017 14:24
sudo apt-get install lm-sensors, then sensors
Brian Rabinovitz
@bdr1177
Jan 20 2017 14:44
@cjo20 Thanks!
Tim Street
@tim2000s
Jan 20 2017 14:54
Max temperature is 90 C on the Edison. Normally run around 48-50 in my experience.
Brian Rabinovitz
@bdr1177
Jan 20 2017 14:55
Mine seems to be running hotter even when it is just out in the open on the desk.
Chris Oattes
@cjo20
Jan 20 2017 15:04
Mine is usually around 60
Brian Rabinovitz
@bdr1177
Jan 20 2017 15:05
Mine is running 63 right now. I guess that's fine. But it did not like being in a case, even with lots of ventilation.
jmcrawford
@jmcrawford
Jan 20 2017 15:10
Running at 61, inside a case, inside a belt, inside a winter coat, inside a van
Brian Rabinovitz
@bdr1177
Jan 20 2017 15:12
@jmcrawford That's good!
Chris Oattes
@cjo20
Jan 20 2017 15:30
I do wonder whether it'd be better with slightly slower charge rate, just to help manage temperature. It's got to be drawing a couple of watts when charging at the moment
jmcrawford
@jmcrawford
Jan 20 2017 15:36
Can you control the charge rate?
Chris Oattes
@cjo20
Jan 20 2017 15:43
Not easily. It's set by a resistor in hardware.
Brian Rabinovitz
@bdr1177
Jan 20 2017 15:58
The documentation says something about two different Edison models, one being described as "wearable". Is there a way to query the Edison to determine the model?
Chris Oattes
@cjo20
Jan 20 2017 16:08
There is probably a model number somewhere. It refers to the amount of power it puts in to the wifi.
mariano42
@mariano42
Jan 20 2017 17:25
setting up my edison explorer rig... just finished Flashing, and I assume it worked since it pulled up a command line after reboot. I logged in, updated the hostname and passwords, but the only directory I see is otp.bin. I feel like im missing a step.
Aaron Michelson
@Aaybob
Jan 20 2017 17:33
@bdr1177 @tim2000s @cjo20 Several things to respond: 56C when cgm in two hour warmup mode 56C when looping for 1/2 hour. : The charge rate of the battery does not affect the temperature of the microprocessor. : "wearable" and ioT or something. On intel web sites for hardware engineers there are specs available. The specs for the two models are in fact the same document.
Chris Oattes
@cjo20
Jan 20 2017 17:34
Aaybob: the rate of the charge of the battery increases the current flow on the board, which will increase temperature
Aaron Michelson
@Aaybob
Jan 20 2017 17:35
How would that happen?
Temperature is from the conversion of electricity to heat...
Chris Oattes
@cjo20
Jan 20 2017 17:37
For one, lipos don't charge at 5v, so there has to be a voltage conversion done at some point. Even if it's switched mode (which I don't think it is), it isn't 100% efficient
Aaron Michelson
@Aaybob
Jan 20 2017 17:37
current flow through resistance causes heat...
the heat on the board is different than the heat on the micro.
Chris Oattes
@cjo20
Jan 20 2017 17:38
And yes, there is some resistance between the USB plug and the battery, and as P = I^2 * R, if you double the current flow, you quadruple the power draw.
It is different, but when it's in a closed space, there is only a fixed amount of air flow. If the voltage regulator heats the air inside the case, that air will be less efficient at cooling the processor
Aaron Michelson
@Aaybob
Jan 20 2017 17:39
so lets say we increase ambient by 10C. How much heat then does that cause in the microprocessor?
Chris Oattes
@cjo20
Jan 20 2017 17:40
Where do you think the temperature rise while plugged in to USB comes from?
Tim Street
@tim2000s
Jan 20 2017 17:41
Someone was asking about WiFi management earlier. I use this very effectively: https://github.com/TC2013/edison_wifi
Aaron Michelson
@Aaybob
Jan 20 2017 17:41
So your experience is when battery is charging the microprocessor runs hotter?
Tim Street
@tim2000s
Jan 20 2017 17:42
@Aaybob Absolutely
Chris Oattes
@cjo20
Jan 20 2017 17:42
@Aaybob Yes, and AIUI, most people experience that too, especially when it's in a case.
Actually, I'll be more specific. The temperature reported for the Edison CPU by sensors is higher when the battery is charging, especially when it is in an enclosed space.
Aaron Michelson
@Aaybob
Jan 20 2017 17:44
@tim2000s on iPhone I have to turn my hotspot on when I leave home. with hotspot on iPhone cannot use home wifi.
@cjo20 give me a mo to change my thinking
I think I'll listen to inauguration
Chris Oattes
@cjo20
Jan 20 2017 17:46
The reason that hotter air temperatures raises the CPU temperature is the same reason that higher air temperature means that you start sweating. The larger the heat differential between two surfaces, the faster the transfer
Tim Street
@tim2000s
Jan 20 2017 17:46
Yeah - if you set up that edison_wifi script, you can flip over afterwards and the edison will find your iPhone network.
Aaron Michelson
@Aaybob
Jan 20 2017 17:55
@tim2000s rig does that by the order of network{}s in wpa_supplicant. When first in list becomes available it switches from any network lower in the list. That happens within one minute.
I liked the speech.
@cjo20 @tim2000s slowing the charging that is now set a 1000mA would be done by decreasing the supply voltage.
Chris Oattes
@cjo20
Jan 20 2017 18:01
I'm pretty Charging rate is set by a resistor value on the charging IC
supply voltage to the IC won't change the current it tries to charge the IC at
Aaron Michelson
@Aaybob
Jan 20 2017 18:03
Explorer notes, if I remember correctly, which often does not seem to happen, is that there is a voltage divider on the supply voltage to reduce voltage to appropriate LiPo level.
Tim Street
@tim2000s
Jan 20 2017 18:03
I don't find that happening with wpa_supplicant on jubilinux @Aaybob
Aaron Michelson
@Aaybob
Jan 20 2017 18:04
@cjo20 agree that IC supply voltage would not have that effect
Aaron Michelson
@Aaybob
Jan 20 2017 18:13
@tim2000s it is on my rig tested and verrified....what linux version are you using....# uname -r
3.10.17-poky-edison+

lsb_release -a

No LSB modules are available.
Distributor ID: Debian
Description: Debian GNU/Linux 8.7 (jessie)
Release: 8.7
Codename: jessie
Chris Oattes
@cjo20
Jan 20 2017 18:16
@Aaybob: The charging IC charges at 1000/R A, and on the explorer board R is 2000. So it charges at 500mA
Aaron Michelson
@Aaybob
Jan 20 2017 18:17
the resistors are so small I would not attempt de-soldering
Chris Oattes
@cjo20
Jan 20 2017 18:19
That gives a peak power dissipation on the charger IC of 1.4W. The 5-pin SOT-23 version of the IC, which is what seems to be used, has a thermal resistance of 230C/W. So 1.4W would raise the temperature of that IC to 322C. Thankfully, it has thermal limiting so it doesn't reach that temperature. It still gets hot
Aaron Michelson
@Aaybob
Jan 20 2017 18:22
if the chip were perfect it would charge at the same rate regardless what resistance the battery presented.
Chris Oattes
@cjo20
Jan 20 2017 18:23
what?
Aaron Michelson
@Aaybob
Jan 20 2017 18:24
Lets say one battery presents a resistance to the charging current of 47 ohms...
Chris Oattes
@cjo20
Jan 20 2017 18:24
This isn't about battery resistance
Aaron Michelson
@Aaybob
Jan 20 2017 18:24
making the battery heat up
It's not about battery current?
during charging?
Chris Oattes
@cjo20
Jan 20 2017 18:25
Nope. It's about the charging IC, which has to handle the charging current going through it, and the voltage conversion from 5v to the battery voltage
Aaron Michelson
@Aaybob
Jan 20 2017 18:26
"going through it" to the battery?
Chris Oattes
@cjo20
Jan 20 2017 18:26
Yes
Aaron Michelson
@Aaybob
Jan 20 2017 18:26
so its about battery resistance
Chris Oattes
@cjo20
Jan 20 2017 18:26
NO
The amount of power that IC has to dissipate is governed by two things. 1) The voltage drop between the input voltage and the output voltage, 2) The current
Aaron Michelson
@Aaybob
Jan 20 2017 18:26
think about it
Chris Oattes
@cjo20
Jan 20 2017 18:26
it's not about battery resistance. Battery resistance doesn't change the power dissipation of the IC
Aaron Michelson
@Aaybob
Jan 20 2017 18:27
lets take a break
R. Dennis Steed
@rdsteed
Jan 20 2017 18:27
A perfect battery would convert all of the electrical energy it receives to chemical energy not thermal energy.
Chris Oattes
@cjo20
Jan 20 2017 18:28
This isn't about the battery. This is happening internally to an IC before the battery becomes involved
At the input of the IC you have 5V @ 500mA. At the output of the IC you have N volts @ 500mA
say that's 3v, to make the maths easy. That's the voltage that goes the battery to charge it
going in to the IC you have 5V*0.5A = 2.5W
Aaron Michelson
@Aaybob
Jan 20 2017 18:29
uuhh the ic is putting power on the gate of an external transister to control current through the transister?
Chris Oattes
@cjo20
Jan 20 2017 18:29
coming out of the IC, you have 3V*0.5A = 1.5W
1W has dissapeared while travelling through the IC. That has to go somewhere, and it is lost as heat. This heats the IC
R. Dennis Steed
@rdsteed
Jan 20 2017 18:30
An efficient power supply can use switching to have a high (average) charging current and low power dissipation.
Chris Oattes
@cjo20
Jan 20 2017 18:30
Yes, it can. The battery charging IC on the edison isn't switching. It just dumps it as heat.
Aaron Michelson
@Aaybob
Jan 20 2017 18:32
When I connected a power supply to JP1 in reverse polarity explorer did a puff of smoke and U3 got burning hot.
so that transistor will compensate for the resistance the circuit sees in the battery up to circuit limits.
Chris Oattes
@cjo20
Jan 20 2017 18:33
That's the battery charging IC dying. It's getting hot because it's shorted inside, and now it's directly connecting + to ground
what transistor are you talking about?
Aaron Michelson
@Aaybob
Jan 20 2017 18:34
I thought U3 was a transisor
Chris Oattes
@cjo20
Jan 20 2017 18:35
U3 is the battery charger IC
Aaron Michelson
@Aaybob
Jan 20 2017 18:36
U3 is the driver of the battery charge?
Chris Oattes
@cjo20
Jan 20 2017 18:36
U3 is what converts the input from +5V to a voltage the battery can use to charge
Eric
@ecc1
Jan 20 2017 18:38
@chrisharris751 Pretty sure rfkill block will shut off the wifi transceiver (that's what's used to put Linux laptops in "airplane mode"), but I haven't verified that it saves any power.
R. Dennis Steed
@rdsteed
Jan 20 2017 18:55
U3 on the Explorer Board is the MCP73831 . Manufacturers website including specs and application notes: http://www.microchip.com/wwwproducts/en/en024903 . It is, as @cjo20 states, a linear and not a switching controller.
Aaron Michelson
@Aaybob
Jan 20 2017 19:06
@cjo20 In U3 is a pass transistor. Remembering a constant voltage/ constant current source is not possible. The pass transistor is a constant (?) source. It provides the other (?) which is nearly constant using the 2000 ohm resistor. Is that resistor in series or parallel with the pins of JP1? I am not able to find the schematic. Would you point me to it?
@cjo20 but whether series or parallel the resistance of the battery is a factor.
It would be less of a factor if 2000 ohm were in series with it assuming the battery resistance is tens of ohms. It would be less of a factor if the 2000 ohm were in parrallel assuming the battery resistance is kilaohms.
Chris Oattes
@cjo20
Jan 20 2017 19:13
You've had the datasheet linked to you already. I'm not sure why you think there's a "pass transistor", or what one of those is. The 2k ohm resistor is not directly connected to the pins of J1. The IC isn't just a transistor. It monitors the battery voltage, and alters it's output voltage to efficiently charge the battery. That's important for lithium batteries because they can be
dangerous if not charged properly. Part of charging them safely is making sure you don't put too much current in to them. So the board designer looks at what size battery they want to use and tells the IC to limit the current to an appropriate value. This is achieved by passing a constant current through the resistor and using the resulting voltage in comparisons to limit the current
output. In this case, the resistor is 2k ohms, which limits the output current to 500mA. The resistor is not on the direct path between the input and the output.
Aaron Michelson
@Aaybob
Jan 20 2017 19:17
Great! so we're trying to discover how to limit the current flowing into the IC because the greater that current the greater the ambient temperature of the IC and the hotter th microproccessor becomes.
Chris Oattes
@cjo20
Jan 20 2017 19:17
We're not trying to discover anything :P
I know exactly how to change the charging current. You change the value of the resistor.
Aaron Michelson
@Aaybob
Jan 20 2017 19:19
Oh, no the conversation started with a question of how to decrease the current charging the battery because the microprocessor was getting hot when battery was charging. You didn't need to discover anything but I was trying to discover a path forward for the questioner.
Scott Leibrand
@scottleibrand
Jan 20 2017 19:19
do we have any actual evidence that the Explorer board gets "too hot" (as opposed to just "hot") during charging? does anyone have a non-defective board that actually shuts off the Edison due to overheating?
Aaron Michelson
@Aaybob
Jan 20 2017 19:20
I have no evidence of heat. But the questioner was concerned about a temp of 86C
Scott Leibrand
@scottleibrand
Jan 20 2017 19:20
there is obviously a tradeoff between charging speed and board temperature during charging. the original design decision favored faster charging than the Intel mini breakout board performs, for convenience
Chris Oattes
@cjo20
Jan 20 2017 19:20
@scottleibrand I think it's more "too hot" as in "uncomfortable" than "exceeding designed thermal limits" for most people
Scott Leibrand
@scottleibrand
Jan 20 2017 19:20
I believe the cutoff is 90C for shutting down the Edison, FWIW
@cjo20 "uncomfortable" as in "perceived risky" or as in "ouch"?
Brian Rabinovitz
@bdr1177
Jan 20 2017 19:21
Mine was running at 90C during charging today. It did not shut off, but I shut it down to prevent damage.
Scott Leibrand
@scottleibrand
Jan 20 2017 19:21
@bdr1177 what kind of case was it in when charging?
Brian Rabinovitz
@bdr1177
Jan 20 2017 19:21
Plastic case with ventilation.
To be honest, I made the case out of legos, but left lots of space for ventilation.
Scott Leibrand
@scottleibrand
Jan 20 2017 19:22
heh nice
Brian Rabinovitz
@bdr1177
Jan 20 2017 19:22
Not charging now and running at a clean 50C in the case.
Scott Leibrand
@scottleibrand
Jan 20 2017 19:22
I have also noticed the charger IC getting too hot to hold my finger on, so it might be worth opening a github issue on the Explorer board repo
Chris Oattes
@cjo20
Jan 20 2017 19:23
@scottleibrand I've had it charging in the case, taken it out and the charger IC has been too hot to touch
Scott Leibrand
@scottleibrand
Jan 20 2017 19:23
I suspect Morgan could either switch to a more efficient charging IC or reduce the charge rate a bit
Brian Rabinovitz
@bdr1177
Jan 20 2017 19:23
So its an issue with the explorer board, not the edison itself?
Scott Leibrand
@scottleibrand
Jan 20 2017 19:23
if it only occurs during charging, yes
Chris Oattes
@cjo20
Jan 20 2017 19:24
The datasheet expects the 5-pin version of the chip to enter thermal regulation with 500mA charging
Brian Rabinovitz
@bdr1177
Jan 20 2017 19:24
From what I've seen so far, but today was the first day I noticed it so I'm not sure.
Scott Leibrand
@scottleibrand
Jan 20 2017 19:24
FWIW, I also have an Adafruit Powerboost that has a high charging current, and it too gets really hot
@cjo20 can you expand on what that means?
Aaron Michelson
@Aaybob
Jan 20 2017 19:24
Rig on desk was 56C during charging and 61C not charging.
Brian Rabinovitz
@bdr1177
Jan 20 2017 19:24
I gotta run, but I'll try to do more observation later.
Chris Oattes
@cjo20
Jan 20 2017 19:25
@scottleibrand When it gets too hot it decreases the charging current to stop itself melting
Scott Leibrand
@scottleibrand
Jan 20 2017 19:25
how does the thermal regulator setpoint get set?
Aaron Michelson
@Aaybob
Jan 20 2017 19:25
Point to the datasheet?
Chris Oattes
@cjo20
Jan 20 2017 19:26
It's internal to the IC
Aaron Michelson
@Aaybob
Jan 20 2017 19:27
but there's no way to set it .....
Chris Oattes
@cjo20
Jan 20 2017 19:27
blob
Scott Leibrand
@scottleibrand
Jan 20 2017 19:27
that is 4.8 on the one you linked?
anyway, same figure
Chris Oattes
@cjo20
Jan 20 2017 19:28
Yeah, I've got two different revisions from two different sources open at the moment, so it depends on which tab I click on :P
Scott Leibrand
@scottleibrand
Jan 20 2017 19:29
so looks like we can expect that IC to get to at least 100C, depending on where on the curve the equilibrium point is.
Chris Oattes
@cjo20
Jan 20 2017 19:29
http://ww1.microchip.com/downloads/en/DeviceDoc/20001984g.pdf That looks to be the more up-to-date of the two
Scott Leibrand
@scottleibrand
Jan 20 2017 19:30
and it won't completely stop charging until 150C
Chris Oattes
@cjo20
Jan 20 2017 19:30
Yeah, that's what I'd guess as a finger-in-the-air approximation
R. Dennis Steed
@rdsteed
Jan 20 2017 19:31
Definitely in the "ouch" range.
Scott Leibrand
@scottleibrand
Jan 20 2017 19:32
I wonder if there's any way other than more ventilation or lower max charging current to keep the IC from heating up the Edison or anything else beyond safe levels...
Chris Oattes
@cjo20
Jan 20 2017 19:32
heatsink might help
but that'll be very dependent on case design
Scott Leibrand
@scottleibrand
Jan 20 2017 19:32
do they make tiny heatsinks for ICs that small?
Aaron Michelson
@Aaybob
Jan 20 2017 19:33
;would someone direct me to a schematic in https://github.com/EnhancedRadioDevices/915MHzEdisonExplorer/
Chris Oattes
@cjo20
Jan 20 2017 19:34
If they do, I don't think they amount to much more than a pad of metal
The 8-pin version has better heat dissipation, but it's bigger.
Aaron Michelson
@Aaybob
Jan 20 2017 19:35
sheeesh I never scrolled down...
Chris Oattes
@cjo20
Jan 20 2017 19:37
ultimately, that IC is going to have to dissipate the same amount of power for a given charging current.
If it's made better at conducting the heat, it just heats up the surroundings easier, which gives you an opportunity to get it out of the case, but that'll be trial-and-error to see if you can actually remove the heat fast enough
Scott Leibrand
@scottleibrand
Jan 20 2017 19:39
yeah, wondering if it's simpler to just limit the charging current. but I think that'll be a decision for Morgan, so someone should open an issue and link it here (so we can comment) if you want him to consider changing that
Chris Oattes
@cjo20
Jan 20 2017 19:45
blob
That figure suggests that even setting it @500mA, a 1000mAh battery might take about 4 hours to fully charge due to thermal regulation, rather than 2 hours
Aaron Michelson
@Aaybob
Jan 20 2017 20:00
does "too hot" occur at the beginning of battery charge or at the end before turning off.
It's likely at the end or after the beginning....
John Benjamin
@Bender1061
Jan 20 2017 20:02
@scottleibrand when I get home from Vacation I'll document the pancreabble stuff in the docs, I have a bunch of changes I want to make to it as well since I've been using it on my cruise ship all week. Also a few issues with getting the time on the Edison when you have no network. Think I know the problem but right now I've been deleting the stuff in the monitor folder as the brute force solution. But I'm sitting on an all inclusive beach in Cozumel with a BG of 106. OpenAPS for the win!
Aaron Michelson
@Aaybob
Jan 20 2017 20:03
@Bender1061 sweet:
R. Dennis Steed
@rdsteed
Jan 20 2017 20:03
@Aaybob @Bender1061 But not too sweet ;-)
Aaron Michelson
@Aaybob
Jan 20 2017 20:04
I'm sticking with sweet!
Scott Leibrand
@scottleibrand
Jan 20 2017 20:04
@Bender1061 :+1: :clap:
Chris Oattes
@cjo20
Jan 20 2017 20:06
:+1:
Aaron Michelson
@Aaybob
Jan 20 2017 20:06
@scottleibrand @cjo20 We have to ask the manufacturer to change the R2 resistor. We have to tell them the boards are getting to hot. I believe they will change it because the specifications say the 2K ohm resistor was never tested.
Scott Leibrand
@scottleibrand
Jan 20 2017 20:08
by "the manufacturer" do you mean ERD, and Morgan, the board designer? I don't think he has to do anything, but you should open a github issue if you want to ask him to consider addressing the heat issue.
Aaron Michelson
@Aaybob
Jan 20 2017 20:09
There is a company somewhere that receives orders for a production run from some entity. The design files are open source. When someone orders a production run they can simply say "change R2 from 2k ohm to 10k ohm".
Scott Leibrand
@scottleibrand
Jan 20 2017 20:10
the orders are placed by Enhanced Radio Devices, Morgan's company. He also controls the open source repo, so if you open an issue there, he'll see it, and can consider updating the design to help with the heat issues, and use the updated design on future production runs.
Aaron Michelson
@Aaybob
Jan 20 2017 20:11
Will do. Understand no shutdown issues but temperatures of 90C are common.
But people don't charge them while in their pocket.
I'd like it to be a legitimate problem.
Anyway I can say we'd like a bigger safety margin so boards can stay in non ventilated cases during charge.
Aaron Michelson
@Aaybob
Jan 20 2017 20:17
The majority of customers will want the fastest charge.
we can contract a production run
Chris Oattes
@cjo20
Jan 20 2017 20:18
You're getting ahead of yourself
Aaron Michelson
@Aaybob
Jan 20 2017 20:19
you read that internal mental dialog?
Scott Leibrand
@scottleibrand
Jan 20 2017 20:19
lol
no one is going to want to fork the design and contract out a separate ordering process. so our best bet is to work with ERD to get the best balance of charging rate and heat for most users, and then design cases around that
Aaron Michelson
@Aaybob
Jan 20 2017 20:23
They might be able to order a special.
Sandra Keßler
@SandraK82
Jan 20 2017 20:28
maybe someone can help me:mmtune: "868.486", 4, -90 Starting pump-loop at Fri 20 Jan 21:10:44 CET 2017: Radio ok, Listening: .No comms detected Old pumphistory: Error, retrying Listening: .No comms detected mmtune: "868.510", 5, -67 Starting pump-loop at Fri 20 Jan 21:13:04 CET 2017: Radio ok, Listening: .No comms detected Old pumphistory: Error, retrying Starting pump-loop at Fri 20 Jan 21:14:10 CET 2017: Radio ok, Listening: .No comms detected Old pumphistory: Error, retrying Starting pump-loop at Fri 20 Jan 21:15:05 CET 2017: Radio ok, Listening: .No comms detected Old pumphistory: Error, retrying Starting pump-loop at Fri 20 Jan 21:16:14 CET 2017: Radio ok, Listening: .No comms detected Old pumphistory: Error, retrying Listening: .No comms detected mmtune: "868.462", 3, -93 Starting pump-loop at Fri 20 Jan 21:18:35 CET 2017: Radio ok, Listening: .No comms detected Old pumphistory: Error, retrying Starting pump-loop at Fri 20 Jan 21:19:39 CET 2017: Radio ok, Listening: .No comms detected Old pumphistory: Error, retrying Listening: .Starting pump-loop at Fri 20 Jan 21:21:07 CET 2017: Radio ok, Listening: .No comms detected Old pumphistory: Error, retrying Listening: .No comms detected mmtune: "868.534", 5, -59 Starting pump-loop at Fri 20 Jan 21:23:34 CET 2017: Radio ok, Listening: .No comms detected Old pumphistory: Error, retrying Listening: .No comms detected mmtune: "868.582", 5, -75 Starting pump-loop at Fri 20 Jan 21:25:51 CET 2017: Radio ok, Listening: .No comms detected Old pumphistory: Error, retrying Listening: .
sometimes i can recover through shutting down the pi and reboot, but not this time
Aaron Michelson
@Aaybob
Jan 20 2017 20:29
@SandraK82 check the pump battery
Sandra Keßler
@SandraK82
Jan 20 2017 20:29
just changed, reported right before with 1,43V
is only a few days old, new battery same problem
Eric
@ecc1
Jan 20 2017 20:30
and the mmtune responses mean the pump is responding to the GetModel command, I think
can you verify that by running mmtune by hand?
Sandra Keßler
@SandraK82
Jan 20 2017 20:32
its an raspi2 with ti usb (older dev branch for >6U /h) and 522 WW, Stick is flashed with 0.9 usb WW
running mmtune manuell gives same frequencies back
Eric
@ecc1
Jan 20 2017 20:33
so it's not a pump comms problem, it must be whatever is complaining about "old pumphistory" (out of my depth there)
Sandra Keßler
@SandraK82
Jan 20 2017 20:33
a few weeks (maybe two) ago the problem started with 1-3 times a day unable to Connect to pump, lost ti stick or unable to read ti version
now it was 4-10 times a day and now I tried a lot of things to get it running for the last 3h
Aaron Michelson
@Aaybob
Jan 20 2017 20:34
The install process is very easy now. I would go back to installing dependencies and continue from there.
Eric
@ecc1
Jan 20 2017 20:34
that seems like a different problem than this one, since the mmtune is working. can you run one iteration of your loop by hand as well?
Sandra Keßler
@SandraK82
Jan 20 2017 20:35
Im just installing rasberian on a second pi and have 3 fresh flashed ti with 1.0 WW here
Aaron Michelson
@Aaybob
Jan 20 2017 20:36
great! you'll be surprised how easy it is to install , its verrrrry easy.
Sandra Keßler
@SandraK82
Jan 20 2017 20:37
its allready an autoinstall, just older version
Aaron Michelson
@Aaybob
Jan 20 2017 20:37
does the CareLink stick work?
Sandra Keßler
@SandraK82
Jan 20 2017 20:38
@ecc1 how was the install of the dev mmeowlink for >6U/h fix? lost that
ah oki was decocare, my fault
Aaron Michelson
@Aaybob
Jan 20 2017 20:40
it helps to ask questions
Sandra Keßler
@SandraK82
Jan 20 2017 20:41
@Aaybob its not carelink, its ti, i dont have a carelink
Aaron Michelson
@Aaybob
Jan 20 2017 20:43
Spot something incorrect or not working? Want a feature/tool to do
something?
[Please file an issue!](https://github.com/bewest/decoding-carelink/issues
Eric
@ecc1
Jan 20 2017 20:44
@SandraK82 no problem. you found the git branch? not sure why it hasn't been merged yet
Aaron Michelson
@Aaybob
Jan 20 2017 20:45
Seems like some corrupt bits of memory. do you have an image from when it was working? or have you tried that?
Sandra Keßler
@SandraK82
Jan 20 2017 20:45
i found my install was on dev therefore i just pullen current dev and prep now for new install
nope sdcard was mirrored weekly and is fine, have already tried with switching to checket sdcard
Chris Oattes
@cjo20
Jan 20 2017 20:46
@SandraK82 Are there errors in /var/log/syslog?
Sandra Keßler
@SandraK82
Jan 20 2017 20:48
Only the expectet: Jan 20 21:46:40 looppi Timed out or other comms error - Received an error response Timeout - retrying: 1 of 3 Jan 20 21:46:40 looppi Timed out or other comms error - Received an error response Timeout - retrying: 2 of 3 Jan 20 21:46:42 looppi Timed out or other comms error - Received an error response Timeout - retrying: 3 of 3
Aaron Michelson
@Aaybob
Jan 20 2017 20:48
but can you write to the SDcard an image from when RPi ran loop?
Sandra Keßler
@SandraK82
Jan 20 2017 20:49
@Aaybob already did, made weekly backups and replayed a version from right after install
Aaron Michelson
@Aaybob
Jan 20 2017 20:50
so you want to reflash linux or TI.
Sandra Keßler
@SandraK82
Jan 20 2017 20:52
linux on raspi was restored to working condition, now im just trying to do other stuff, like setting up a complete new rig and updating old setup to newest versions
something will work as it usually does
Tim Street
@tim2000s
Jan 20 2017 21:15
@Aaybob Made some progress on pancreabble today. Hopefully I'll get it working over the weekend and can then PR to the docs.
Know what I need to do to get peb-urchin-status.sh working and seem to have been able to add reports. Definitely getting there.
Matthias
@ceben80
Jan 20 2017 22:38
@SandraK82 possibly a WW pump issue, try to implement this here
https://github.com/oskarpearson/mmeowlink/wiki/Non-USA-pump-settings
I had similar issues like you as I remember regarding this thrown errors
Sandra Keßler
@SandraK82
Jan 20 2017 22:39
maybe, just Testing with a complete new setup
Sandra Keßler
@SandraK82
Jan 20 2017 22:56
oki did everything fresh and new (new pi, new card, new ti stick, new setup) and now it works again
@/all thanks for the Troubleshooting help, even it did not finish
PieterGit
@PieterGit
Jan 20 2017 23:01
@SandraK82 : i noticed this can be caused by a git error that is not fixed by git fix scripts. this can be solved with a mv .git ../.git.corrupted ; git init
@SandraK82 : other thing i noticed that sometimes the ti usb stick tends to communicate worse if it's hot, sometimes really unplugging it makes a difference. We might need to check al the alias parts of mmtune to find out which one is causing the error
Sandra Keßler
@SandraK82
Jan 20 2017 23:06
i will try again next on the old setup
just copying the sdcards
ejoe132
@ejoe132
Jan 20 2017 23:08
Just received my explorer board, just wanted to confirm, if I want my edison setup to run where I dont have wifi I need to have my dexcom plugged in correct?
Aaron Michelson
@Aaybob
Jan 20 2017 23:14
@SandraK82 :smile: copy sd cards and a flash of edison, or was it flash of ti? english "somtimes unplugging it really makes"
"sometimes just un...."
Sandra Keßler
@SandraK82
Jan 20 2017 23:18
all 4 ti now on latest master (1.0.0)
Sandra Keßler
@SandraK82
Jan 20 2017 23:51
yeah with freshly flashed ti and copied card even my old riss runs fine again
Aaron Michelson
@Aaybob
Jan 20 2017 23:52
sorry you had to go through that but you did well
Sandra Keßler
@SandraK82
Jan 20 2017 23:52
thanks @Aaybob
even more pressure for me to get the combo pump ready to use ;)
i hate the big rig, however i dont want to buy an Edi :)
Dana Lewis
@danamlewis
Jan 20 2017 23:58
@ejoe132 yes and with a second battery also plugged in to power it.