In an earlier post I presented a capsule-shaped gag. This shape offers a few (minor) advantages to the ubiquitous ball gag. Even though this capsule gag is only a stepwise improvement over the ball gag, long journey often starts with a few small steps.
A side effect of the lengthened shape of the gag is that it has sufficient space for a set of batteries and a blinking LED.
The half-ball of the gag unscrews, giving access to 2 AAA-batteries and a small circuit board.
The circuit has to MEMS microphones, one at each end, and a microcontroller. The microcontroller right now samples both microphones 10 times per second and dumps the results. I put the results of a short test in a graph, which is further down.
In the test, I had background music at roughly 65 dB and the moans I made were also roughly 65 db when measured at approximately 30 cm with a simple sound level meter. As is apparent, the ambient sound is barely picked up. The moans are very prominent: while it may be 65 dB at 30 cm, the sound level is much higher close to the source (meaning, inside the mouth). My hypotheses was that one would be able to to distiguish sound "coming from the outside" from sound coming from the voice, by comparing the levels of the front and back microphones. In the graph results, we see that this indeed works: on moaning the back microphone reports a higher sound level than the front microphone.
The first peak, at around 11.5 seconds, may be an error in capturing the microphone or in data transfer. The capturing software is still simplistic, it does not filter out outliers (or even check for them); and the microcontroller runs on its internal RC resonator, which isn't the best when doing RS232 communication.
The circuit does have a low-pass RC-filter with the cut-off frequency at 500 Hz. The idea is to filter out the "hiss" sounds of breathing. These sounds were indeed not an issue in my test. However, I haven't tested without these filters, so I don't know if breathing sounds are even an issue.
Now that we can detect a voice, how do we use that? Ah, there's the rub. I have currently no plan for the next step for this experiment. There is room in the capsule gag for a wireless module, but again: what do you do with it when receiving a signal from the gag. When moving toward the "voilette" gag design (as described in the mock patent), the next step would be the air valves, but I think that project is beyond my "hobby scope" budget.
Comments
Post a Comment