It's been a challenging week for the project, but results are currently looking good. Since the last blog update, all components have been ordered and have arrived (with the exception of one or two stray parts we're still waiting for). Since then, circuits have been fully designed in both schematic and physical board layout form to facilitate testing of the components and associated circuit designs. A sensor board has been physically printed and populated with components, shown below, including the temperature sensor, humidity sensor, light sensor, hi-frequency microphone and amplification circuits. I've been testing digital communications with the various sensors, ran into some minor difficulties initially but have now overcome them and have communications working nicely. I'm about to run some calibration tests to check that the data they're collecting is correct and accurate.
Meanwhile, I've begun testing the microphones and associated amplifier circuits, checking them for performance and frequency response - vital if they're to pick up and - perhaps more importantly - distinguish between the various, relatively quiet and high-frequency songs of the bush-crickets in the 'Conocephalus' Genus. Currently, the results are looking excellent. The image below shows the preliminary results - the first waveform is a recording of the Conocephalus Discolor taken from the CD accompanying the book "Grasshoppers and Crickets of Western Europe". The waveform was played from high-quality speakers at a relatively low volume in a quiet room next to the microphone implemented on the "sensor prototype board". The resulting recorded waveform is shown on the bottom - it demonstrates an extremely high fidelity copy of the original waveform (note: the two waveforms shown are not at the exact same point in the recording, so they are subtly different), and demonstrates the mic's ability to "hear" the bush-cricket. Of course, there's more work to do in writing an algorithm to actually identify the song, but this is an excellent start.
A small amount more software work needs to be done on data-collection from the sensors, and the next objective will be to fabricate a full working main board to accommodate the GPS, SD card etc. Since the board has been designed, this will be a matter of fabricating the board and populating it with components. Hopefully the turnaround for this will be relatively rapid, though the fabrication labs and staff here are currently busy with another large project, which could be a small hindrance - we shall see.
In summary, work is going well and the results from all aspects so far are pleasing - however, it's also hard to deny that we're behind schedule from where we'd ideally be at this stage in August. I'm working hard, and keeping very long working hours to keep the project as on-time as I can. Problems are being tackled as fast as possible as and when they appear, so it's a case of watch this space, and I'll keep posting here with the latest!
Add new comment