Differences

This shows you the differences between two versions of the page.

--- wind_sensor:meeting_minutes_mar_24_2017 [2017/03/25 03:24]
mwu [Ultrasonic Wind Sensor]
+++ wind_sensor:meeting_minutes_mar_24_2017 [2021/09/19 21:59] (current)
@@ Line 3: / Line 3: @@
 =====Acoustic Wind Sensor=====
   * Updates:
-    *
+    * Replaced microphones on the pcb and conducted tests again
+    * Re-began testing 4-microphone set-up
+      * Unexpectedly, when the microphones were faced perpendicular to the direction of the wind, they still output the same wind speed
+        * This is most likely due to the fact that the microphones are omni-directional
+      * Results for a single microphone set-up are also still inconsistent on the pcb.
   * To Do:
-    *
+    * Solder microphones onto a perforated board to see if the traces on the pcb were the problem.
+    * Build wind channels for the microphones
+      * Will probably entail recalculating coefficients, since the funnels will probably affect the output of the microphones
 =====Ultrasonic Wind Sensor=====
   * Updates:
@@ Line 15: / Line 21: @@
     * Developed an algorithm that involves using the Teensy to generate the 40kHz square wave instead of using the 555 timer. This is mainly due to avoid the possibility of a current bottleneck with trying to drive the transducer using the 555 timer.
       * The algorithm involves the Teeny sending 10 25us pulses and then using thresholding and ADC to record the pulses. Since the tof of a pulse is longer than the time it takes the Teensy to send the pulses, we don't have to worry about potentially trying to record while still sending. Average a bunch of samples and we should have a result.
+        * Quickly wrote a program to generate a 40kHz square wave using the Teensy's TimerOne library. Observing it on the oscilloscope, the square wave looked pretty clean. We then tried to drive the emitter using the Teensy's square wave and observing the results. The results are below:
+          * For some reason, when moving the transducers farther apart, sometimes we are getting an increase in amplitude, rather than a decrease. It seems like we're getting a good amplitude around 6-8cm so we should test that distance some more and redesign our algorithm if need be.
+          * The breadboards aren't completely level and this threw our amplitudes off. We have to find another breadboard that's level with the one we have now.
     * Did a bit of research into TDCs (time-to-digital converters). These chips are designed specifically for measuring the time between two events.
       * One TDC chip we're looking at is the TDC1000 by Texas Instruments (http://www.ti.com/lit/ds/symlink/tdc1000-q1.pdf). It features the ability to drive two emitters and receivers, as well as the ability to generate the signal for the emitter and count until the receiver hears the signal.