Project Logs

Collapse

• LEDE Project Gateway Build

  Sotirios Vrachas • 06/12/2017 at 15:07 • 0 comments

  For gateway devices with limited memoy

  git clone https://github.com/SotiriosVrachas/lede-project-source.git lede
  cd lede
  git checkout v17.01.1
  make defconfig
  make image -j 3 PROFILE=tl-wr703n-v1 PACKAGES="kmod-usb-ohci kmod-usb-uhci kmod-bluetooth kmod-bluetooth_6lowpan bluez-libs bluez-utils"
  

  where 3 in

  -j 3
  

  is the number of CPU cores that you have available plus one and the tl-wr703n-v1 in

  PROFILE=tl-wr703n-v1
  

  is your device model.

  scp bin/targets/*/*/*sysupgrade.bin root@lededevice:/tmp/sysupgrade.bin

  whare
  

  lededevice

  is the network address of the device.

  ssh root@lededevice

  sysupgrade /tmp/sysupgrade.bin
  

• Case design

  Sotirios Vrachas • 06/12/2017 at 14:00 • 0 comments

  I want to replace the stock handle of umbrellas with my handles that included the magic.
  Experimented with clear nylon and clear petg 3d printing filaments.
  After testing both the petg was a lot more clear, transparent, and was snappable making prototyping a lot easier.

  
  
  I designed a concept case to test the lights and the defusion of tha petg. i sanded the case and then applied clear varnish. made a coin cell + led and enclosed it in the case to see how the lighting will work under different condition.

  
  
  Dew to space constrains and the fact that one led dose not seep to be enough, concluded that smd leds will have to be used.

  Took the dimensions of the original hanle and experimented with making a base for the common M14 screw umbrella's have for the handle mound
  

• Case design

  Sotirios Vrachas • 06/12/2017 at 13:58 • 0 comments

  I want to replace the stock handle of umbrellas with my handles that included the magic.
  Experimented with clear nylon and clear petg 3d printing filaments.
  After testing both the petg was a lot more clear, transparent, and was snappable making prototyping a lot easyer.

  
  
  I designed a concept case to test the lights and the defusion of tha petg. i sanded the case and then applied clear varnish. made a coin cell + led and enclsed it in the case to see how the lighting will work under diffrent condition.

  
  
  Dew to space constrains and the fact that one led dose not seep to be enougth, concluded that smd leds will have to be used.

• Firmware

  Sotirios Vrachas • 06/12/2017 at 13:54 • 0 comments

  Introduction
  After researching the available platforms and concluding on the nRF51822, the choice of the firmware had to be researched.
  Candidate RTOS with 6loWPAN support
  Nordics SDK
  RIOT
  CONTIKI

  
  
  Contiki port for nRF51822 does not handle the radio, so this one was eliminated

  
  
  Then the Nordic's sdk have a binary blob for the radio and a special non free clause in the license.

  
  
  The RIOT seems to be the only Free and Open Source rtos that seems to support the platform, their is even reference in the nordics blog for the nrf52.

  
  
  The RIOT documentation is unclear if the radio of nRF51 is suppoted or not .

  
  
  Following the RIOT instructions for nRF51822 lead to a working system without radio.

  
  
  Tried flashing riot after flashing softradio and investigated how the Nordic sdk was loaded to memory.

  
  
  Then when ahead and compiled and flashed the IoT SDK from Nordic specifically the CoAP Server Examples. those required a softradio that was not production ready and will never be according to Nordic.

  
  
  I manged to ping the device over ble with the linklocal ipv6 that matched its Mac addresses.

  The latency reported by ping was 100ms no mater the distance between of the radios.
  

View all 4 project logs