« May 2006 | Main | July 2006 »

MTS310 hardware problem

June 23, 2006 by Duncan |

Have sent off 19 MTS300 and MTS310 back to Crossbow for rewiring (2-3 week turn around). Have installed Moteworks v2 rel B. So for time being focusing on application of MTS400 (3 off) and the A/D sensor boards....

nVIDIA on fire

June 23, 2006 by Duncan |

Alvise found this when looking for reviews on the new laptop he is getting.... nothing like seeing other peoples experiences. I like the quote from the guy who posted this *It is only a matter of time until such an incident breaks out on a plane*...


Ad hoc routing

June 20, 2006 by Duncan |

Animated gif showing some time stamps of ad hoc routing over the duration of a 2 day trial - note changes in routing patterns - probably due to changes in battery strength and physical changes in space (people, obstructions etc).

screenshot_24hr_cycle_topo_sml.gif

Getting things done

June 13, 2006 by Duncan |

Met with Dan Hill today talking about interactivity in the built environment and away days - an unusual mix... the conversation reminded me to look up: Getting started with “Getting Things Done” | 43 Folders

20 node 24 hours

June 13, 2006 by Duncan | | Comments (1)

screenshot_24hr_cycle.JPG This is a screenshot of a 24 hour cycle of 20 motes measuring temp, light, battery, mic, accel and mag. Motes are micaz, sensor boards are MTS300 and 310's. Gateway is an MIB510 connected via serial to USB converter. Data being captured into Postgres db on sony laptop. Code on gateway is XMeshBase as shipped with moteworks 2.0. Code on motes is xmts310 from xmesh folder, again shipped with moteworks 2.0 - the only change was my addition of the following to hard code into the application a timer interval of 30 seconds:
(in static void initialise after timer_rate has been set, inserted at line 169)

timer_rate = 1024 * 30;

Curious observation - why does the temp sensor readings oscillate so much during the day but then do not between 9.30pm and 3.00am? The fluorescent lights are on timers so they would have gone out at around 9pm but would not have come back on continuously at 3.00am (might have come on intermittently as security walked through the building...) Is this sensitivity on the analogue sensor? Am i getting intereference. I seemed to get the same 4 degree oscillation at home where there are no fluorescent lights. Could it be the 60Hz power?

20 nodes fluctuations

June 12, 2006 by Duncan | | Comments (1)

screenshot2.JPG The 20 node network gave some interesting trends but weird results. Not sure why there is so much fluctuation in temperature and seems a little too periodic.... lodged a request with the support desk...

wind at southend

June 09, 2006 by Duncan |

WindGURU at Southend
Tides at Sheerness
WindGURU at Queen Mary

20 nodes

June 09, 2006 by Duncan |

20motes.jpg Finally got all 20 nodes up and running. 1 gateway and 19 nodes with either MTS300 or MTS310 sensor boards. Had it running overnight - moteview crashed a couple of times will need to try and figure out why. Need to go and review data to see how valid it looks. nice feeling to see all 20 motes networking in an adhoc manner.... :-)

blinkybug.com - gallery

June 05, 2006 by Duncan |

Found these from posting to Make site - what an excellent use of *spare* bits and pieces.... the blinky bugs

5 nodes

June 02, 2006 by Duncan |

1 gateway and 4 sensor motes running. Having trouble deciphering code to get all sensors working. Have tried to use the XMesh samples as my base. Used the tutorial file (containing the temp only code) and expand it to include the other stuff. All the examples seem to be slightly different implementations though so figuring out what is and what is not required has been tough. Had some fun at one point trying to figure out that the x in AccelX had to be capitalised.....

Ended up reverting back to the MTS300 sensor board sample files and using those with XMesh - there is also some other stuff going on in there (performance assessment) - but have hard coded the timer rate since i could not figure out how to define APP_RATE (i think you do it via the header file ***.h but could not get it working - logged a call to the crossbow help desk)

code finally up and running. have chosen a 30 sec interval for timer firing. interested in the implications for battery life - btw will need to look at low power stuff (is it possible on micaz?) otherwise batteries only last a few days....

next step - 20 nodes and get sampling over a couple of days.....

photos

www.flickr.com
This is a Flickr badge showing public photos from pseudonomad tagged with blog. Make your own badge here.

latest del.icio.us

my del.icio.us

Powered by Movable Type 3.2