Last week was shortened due to the 4th of July holiday. On Monday, July 1st 2013, Denis placed three additional Moos in the 9th row of the rear wall and one Moo on the 9th row of the side wall. He placed the Moos such that an antenna (half ground and half signal) would be poking into an empty cinder block (note there is very little clearance). We are hoping that this orientation may allow us to better communicate with the reader during the curing process by mitigating the effects of attenuation. In our laboratory experiments, we found that if the signal antenna is poking out of a fully submerged Moo inside water, we can still get a reading.
Inside the lab, Mike and I performed some additional experiments with the Moos to better understand how they would respond in two new environments: subzero temperature inside a mini-fridge and submerged inside wet sand.
Fridge Experiment (7/1/13)
- We placed a Moo inside of the fridge in a relatively static position along with the reader.
- Logged data and recorded observations (see google doc Moo/Experiments).
- TODO: plot data using R (Denis)
- Observations: Noticed a sudden drop in tag rate several times during experiment. Not sure what was causing this – maybe a subtle shift or moisture from the fridge. When opening the door and closing it again or repositioning the Moo inside fridge, we were able to get rate back.
Wet Sand Experiment (7/1/13)
- more info here: https://docs.google.com/a/umich.edu/document/d/1rkw7hPEzXR2AKH2-ituy-uPJaOAaj0YyDMGN48KPTBo/edit
- Paradox: We were able to get a solid tag rate (~4 per sec) from the Moo inside of a Moo that is submerged inside 2ns sand that was oversaturated with water. This appears inconsistent with our previous results because no tag rate in both pure water and wet (or slightly moist) concrete, which contains sand as an aggregate.
- TODO: Plot data from experiment.
Water Experiment (7/3/13)
- placed Moo inside very small puddle of water with antenna underneath. No tag rate as long as water is barely touching the signal (right) antenna.
- Moo Tx_counter resets after pulling Moo out of water (even after very short amount of time). This implies that the Moo is not able to receive power from reader nor communicate with water because volatile memory is reset.
- Ground antenna can be completely submerged inside water without affecting the tag rate.
Possible Ideas to improve communication
- Measure length of dipole antenna (as a multiple of lamba). How would performance be affected if we increased (maybe doubled) the length of the Moo antenna. Re-solder a longer antenna on to the Moo and measure performance affects. Who is the expert on this? Post question to the Wisp forum.
Reading this Week (with Notes):
Radiation and Antennas (chpt. 9) of Emag textbook: basic principles of antenna design (includes dipole antennas)
- Most effective performance (max power density) when object is placed at 90 degrees off of dipole axis (i.e. perpendicular position).
- Current flowing through dipole antenna has symmetrical distribution wrt center of dipole and current is zero at its ends.
- [Impedance Match] to maximize power transfer to the load, the load impedance must be chosen such that ZL = complex conjugate of Z in (i.e. RL = Rrad and Xl = -Xin).
- Friis Transmission Formula: Power transfer ratio P_received/P_transmit = Gt*Gr*(lambda/4*pi*radius)^2. (i.e. Inversely proportional to radial distance squared)
Next Steps (??)
- begin development of Moo 2.0 PCB Board modifications (have finished and sent before August)
- begin making poster for ESWeek. Gather Notes & place info inside google doc. What editor should I use? LaTeX? Some Ideas on general format and structure?
- Test other ideas to improve tag rate before side wall and front wall deployment begins (roughly 2 weeks from now).