Week 17: Applications & Implications¶
Goals for Applications & Implications¶
- Answer these important questions about the invention
- Create a formal BOM such that anyone could reproduce the device
- Update these materials as needed
What will it do?¶
This device aims to listen to and cancel disturbances caused by dental equipment. More specifically, it will detect sound waves and/or vibrations that are traveling through the patients teeth and use destructive interference to create active noise cancelation in the users’ ear canal. These disturbances are known to range from unpleasant to downright painful. The device will detect these vibrations near the dental tool, which will then get processed, inverted, combined with music of the patients choosing, and sent to a bone conducting transducer on the temporal bone, near the ear canal. In this case, processing refers to digitization and inversion as well as delay and volume adjustments. Central to this idea is that the speed of light and processing is faster than the speed of sound, even through bone tissue. Therefore, the inverted signal can ‘beat’ the original signal to the punch and a delay knob can bring them perfectly out of phase.
Who’s done what beforehand?¶
See resources page for a list of relevant projects, research papers, and resources that have aided this process.
What will you design?¶
I will design an all-in one mono bone conducting headset with the ability to play music and cancel incoming signals at the same time. It will have a slider switch to turn off the music and only show the cancelation technology. It will have the ability to adjust the volume and delay of the cancelation signal.
Extra emphasis will be put into designing a nice form factor that’s also comfortable to wear.
The disturbance will be connected to the headset via hardware with a modular cable connector. This will be useful for using diufferent disturbance sigal input options (see how will it be evaluated section?).
What materials and components will be used?¶
Electronics¶
| Component | Make, Model, and links | Specific Function(s) | Sourced from | Cost |
|---|---|---|---|---|
| MCU | ATtiny 1614 | Sound processing, digitization (ADC/DAC), central command | Digikey-Microchip | $1.05 |
| BC transducer | Adafruit 1675 | Output signal to temporal bone. Possibly also used to generate disturbance signal | Digikey-Adafruit | $8.95 |
| Class D amp module | LM 386 EK1236 | Amplify output signal | Amazon-Freedom-1 | $1.60 |
| Mic/pre-amp | Adafruit Max4466 Electret | Disturbance signal input. Also used to prove ANC test stand | Adafruit-Adafruit | $6.95 |
| WiFi Radio (1) | ESP 8266 Node MCU | Connect to WiFi, stream .Wav files | Amazon-esepressif | $7.99 |
| WiFi Radio (2) | Espressif ESP8266 | Smaller form factor of above | Amazon-espressif | $7.50 |
| Op-Amp (1) | LM 358N | Analog signal processing. Suited for typical op-amp circuitry. | Digikey-onsemi | $0.43 |
| Op-Amp (2) | LM 386 | Analog signal processing. Suited for audio applications. | Amazon - National Semiconductor | $1.49 |
| 10k potentiometer | Taiss 60PCS Potentiometer kit | Delay and volume control | Amazon-Taiss | $0.17 |
| SPDT Slide switch | surface mount labeled “0H” | Music on/off | unknown | unknown |
Programming/Interfaces¶
| Component | Make, Model, and links | Specific Function(s) |
|---|---|---|
| jtag2UDPI | Arduino Uno | Programming MCU boards |
| SAMD Programmer | SheekGeek SAMD11C Multi-use board | Programming MCU boards |
| FTDI Cable | Sparkfun 5V FTDI cable | Easy Serial communication, processing |
In addition to these components, various commodities such as resistors, capacitors, etc. were used. These will not be listed here for the sake of brevity.
Raw materials by process¶
| Process | Material | Source | Price/unit |
|---|---|---|---|
| PCB Milling | 4”x5” Single Sided FR-1 | Bantam Tools | $1.45/sheet |
| FDM printing | Galaxy Black PLA | Prusa | $29.99/kg |
| FDM printing | White Recycled PETG | GreenGate | $29.99/kg |
| Resin printing | Water Washable Resin | Elegoo | $21.99/500g |
| Laser Cutting | 1/8” black acrylic | estreetplastics | $26.99/sheet |
| Mold Production | Machining Wax | Bantam Tools | $10.35/pc |
| Molding - hard material | Smooth-Cast 300 | Smooth-on | $28.87/kit |
Where will they come from?¶
See sources section in table above.
How much will they cost?¶
See see price section in table above.
What parts and systems will be made?¶
Beyond the electronic components and modules mentioned above, the entire headset will be manufactured from scratch. This includes the following systems and sub systems:
- Electronics
- Main MCU board - milled out of FR-1
- Controls Board - milled out of FR-1
- Speaker Box
- Speaker cover - molded out of mutiple materials
- Speaker housing - FDM printed in PLA
- Headset adjustment - FDM printed in PLA
- Overhead strap - tbd
- Mounting points for electronics * FDM printed in PLA
- Access window - tbd
- Cushion - tbd
- Opposing head strap - tbd
- Known Disturbance wand
- Housing - FDM printed in PLA
- Speaker cover - molded out of mutiple materials
- MCU board - milled out of FR-1
- Modular wired connection to headset
- Unknown Disturbance Microphone assembly
- Housing - FDM printed in PLA
- Attachment method - tbd
- Modular wired connection to headset
What processes will be used?¶
- KiCAD Electronics design of MCU boards
- PCB milling of FR-1 on other mills
- Molding/Casting of hard/soft materials with wax machined on other mills
- Embedded programming with ATtiny 1614 microchip *Embedded programming with ESP8266 module
- FDM 3D printing of speaker housing
- Interfacing with Processing for on-screen delay function
- Electret microphone as Input Device
- BC Transducer as Output Device
- [SLA 3D printing of…]
What questions need to be answered?¶
How does active noise cancelation work?
Is it possible to combine bone conduction transmission and active noise cancelation?
Can the dental experience be made more pleasurable using this technology?
Is it possible to use a microphone input to cancel the vibration propagating through the skull?
What form factor(s) makes the most sense for future spirals and the dental application?
What worked?¶
The bone conduction transducer was surprisingly easy to implement, and worked just like a normal speaker. To read more about this, see week 10.
Creating a noise cancelation test stand (see ANC2) was the first time I successfully created a noise cancelation effect. This in itself felt like a huge win. To read more about this, see the ANC2 section above.
Processing the signal digitally worked very well. It simply required adding a ~ in front of the 10 bit variable, which in effect flipped all of the bits and flipped the entire signal. A delay was created using a buffer method. The only issue with this method is it was subject to the sample rate of the main loop. Luckily, the length of this sample rate and resulting delay was on the right order of magnitude for the distances/delay needed. To read more about this process, see week 15.
What didn’t work?¶
Initially, I would consider the speaker cover a failure. The soft piece was too soft to hold together whereas the hard molded pieces created tons of sound leakage and inhibited good transmission to the head. It wasn’t until I let the metal piece poke through that this offered desirable results. To read more about this, see section H1 above.
My first attempt at a 3D design was a failure. I realized that I jumped right into 3D before having an accurate sketch or concept in mind. As powerful as CAD systems are, they are not a good place for failing fast. After generating a 2D concept sketch, returning to 3D was much more fruitful. To read more about this, see section H3 above.
When starting out creating the main MCU board, I was attempting to debug everything through a serial interface. This lacked both the control and speed I was needing. It’s a good thing that Tom convinced me to switch over to using the Digital to Analog discovery kit. This kit completely upped my game. To read more about this transition, see week 13.
The ESP8266 WiFi radio module remains very buggy. This is partially because it did not fulfill an assignment and therefore it was hard to justify too much more time spent on it. The main bug that remains seems to be a WDT timeout. This sort of error occurs to an ESP module when you prevent the main loop from executing for too long and background functions are not allowed to flow. Time permitting, I will return to address this issue. To read more about this exploration, see week 14.
How will it be evaluated?¶
In spirals,
- The device will first be tested with a known disturbance signal that is sent into the tooth by a 2nd BC transducer. The signal will be inverted and sent through the users skull to cancel the original signal. This will be repeated with music playing over top.
- The device will then be tested with an unknown disturbance signal that is sent into the tooth by an electric toothbrush, and collected via a microphone. The signal will be inverted and sent through the users skull to cancel the original signal. This will be repeated with music playing.
- The device will then be tested with an unknown disturbance signal that is sent into the tooth by a dental drill, and collected via a microphone. The signal will be inverted and sent through the users skull to cancel the original signal. This will be repeated with music playing.
The performance will be evaluated subjectively by multiple users. In addition to the noise cancelation, users will be asked to rate:
- Form and function
- comfort level
- fit to their own head
- their aversion to the dentist office
- would they pay extra for this option? If so, how much?
User’s will be asked to participate in a survey to answer these questions.
What are the implications?¶
If this device performs as intended, this could greatly reduce discomfort felt in the dental chair. This could also reduce aversion to visiting the dentist, leading to healthier teeth in the population. I speculate more people going to the dentist more regularly would lead in an uptick in mild operations such as fillings, but a downtick in cases of neglect that lead to more extreme operations such as root canals. In either event, those operations would potentially be more pleasurable across the board.
Since the headphones are open ear, dental offices would potentially need to stop playing ambient music, less their patients want to hear two songs at once. That said, the dentist could also wear over the ear bone conducting headphones, without inhibiting their ability to communicate with the patient.
The connection between dental proffessions and mental health issues is well understood. Though, the widespread claims of higher suicide rates among dentists have been disputed. If this device can improve the happiness of the patients, this would have a secondary leading to happier dental professionals.
~ are stretch goals ~
All works shared on this site 'A Fab Academy Journey with Charlie Horvath' by Charles W. Horvath are licensed under Attribution-NonCommercial-ShareAlike 4.0 International
