Author: oball

As you’ve now established a set of sub-systems necessary to implement your project, what new tools are you looking into learning so you are able to accomplish your planned tasks?

To accomplish the hardware tasks planned for this project, I have had to learn how to use various software tools, along with a handful of physical tools.

The main software tool that I have been learning to design the PCBs for the project has been KiCAD. In the past, I have worked with Altium and EasyEDA but I wanted to branch out and learn more tools. EasyEDA served me well for a handful of designs but I wanted slightly more professional software, and one with a larger community. Although I also have experience with Altium, I personally do not enjoy it for personal projects, and I wanted to familiarize myself with the workflow associated with software other than Altium.  Learning KiCAD primarily consisted of familiarizing myself with the location of various options and menus, such as customizing the DRC rules to match those of JLCPCB. I also had the opportunity to learn how to import symbols from Digikey and create/import footprints for parts. In the past I have either been able to find parts in the built-in catalog or in my company’s parts library, so it was good to learn the process of creating my own parts. I also had to learn how to create a precise board outline and mounting holes, which I have not had experience with in the past. In previous projects, I have just created a board and then created a mount to fit it in CAD, but for this project, my boards must mount onto existing hardware such as a Raspberry Pi. I have also been learning new skills in Arduino to test my hardware. For example, I have learned how to use the NeoPixel library to run the LEDs on the PCBs as well as learned how to detect interrupts with the Teensy and operate the D-flip-flops on the PCBs in order to accelerate my testing.

In terms of physical tools, I have been learning to work with PCB stencils and surface mount soldering. I have dealt with some surface mount soldering of 0402-0805 resistors and capacitors, but not in large quantities or anything more complex. Additionally, in the past I have always hand-soldered these parts. However, since we need around 14-15 of the fretboard PCBs, I am learning how to use PCB stencils, solder paste, and hot plates in order to expedite the soldering process. I have also been learning about a variety of sensors in our quest to find solutions to some of our sensing problems. This consists of learning how capacitive touch sensors work, as well as how the inductive and piezoelectric pickups used in many guitars function.

For the future, I am mainly interested in learning how MIDI files work a bit more so that I am more familiar with the work being done by software and firmware and can better integrate the hardware with these systems.

Weekly Update

This week I primarily worked on the assembly and testing of the fretboard PCBs and the design of the Pi-Hat.

The assembly of the fretboard PCBs went well. This was done using the SMD stencil and a hot plate. In total 5 were assembled for Tushaar to use for testing. The pad for the current limiting resistor was replaced with a surface mount diode. The resistor will be added between the fret pad and the fret, so the safety of the system is not compromised by this. The functionality of the LEDs was tested, along with the D-flip-flops used for driving each fret high.

Top: One of the fretboard PCBs on a hot plate
Bottom: Initial tests of the LEDs on the fretboard PCB

There are a couple of changes that need to be made to the PCB, most of which were implemented this week. The first change is adding a pad for the SMD diode in series with the current limiting resistor. The next change was to switch out the through-hole pads with pads on just the surface of the PCB. If solder flows though the through-hole pad it can result in an uneven bottom of the board, making mounting it more difficult. The next change was to make the PCB narrower by around 1.5mm. The initial measurements of the PCB were taken from a photo, and as a result, the PCB is slightly too wide to fit next to the 14th fret. The final change that needs to be made is to adjust the spacing of the LEDs. Currently, the spacing is reasonable for the middle of the guitar, but by making a few variants of the board, we can better match the spacing of the strings. The updated board layout is shown below

D1-D4 on the left are the LEDs. The top pads are the connections for the LEDs, and the bottom right ones are for the fret-driving flip-flop, which is on the right side of the board.

This week I also worked on the Pi-Hat PCB. The main change to the plan for the PCB is to offload the strum detection to a perfboard. The circuit needs a lot of tuning and this can be better done on a perfboard. This reduces the risk of the strum detection not functioning.
The PCB consists of the following components:

1. A 5V barrel jack connector with filter capacitors. The 5V directly powers the Pi and LEDs and the Teensy is powered through a Schottky diode. This prevents backpowering the micro USB port on the Teensy, allowing the Teensy to be plugged into a computer while the system is also powered. It also prevents the USB port from attempting to drive the LEDs, which would draw too much current.
2. The Teensy 4.1. A handful of the I/Os are sent directly to pads for future needs. 4 pads have external pull-down resistors and will be connected to the strings of the guitar. 10 of the I/O are routed directly to Pi GPIO pins allowing for communication. There are currently no plans to use all 10 of these, but this enables expansion in the future.
3. A header to mate directly to the header on the Pi. In addition to the GPIOs routed to the Teensy, 4 I/O pins are broken out to pads for future use
4. An active buzzer driven off a Teensy PWM pin. We can PWM the power to the buzzer to control the volume of the buzzer. A resistor pad is also included for current limiting the buzzer, although the current plan is to place a jumper on this pad.
5. A 3.3V to 5V logic level converter to convert the 3.3V signal from the Teensy to a 5V signal for the LEDs. This may or may not be needed according to testing, but according to the LED datasheet the signal should be at least 3.6v

The current layout of the board is shown above. The Pi header is on the right side and the Teensy is the component to the left of the headers. The top left is the power barrel jack. Along the left side is I/O and at the bottom of the board is the buzzer.

For the most part my progress is on track. The fretboard PCBs still need their LED spacing revised, but this can be done fairly quickly once the desired spacing is determined. The Pi-Hat PCB design is ahead of schedule. These boards will likely be ready to be sent out in the same order, reducing the shipping costs incurred. The PCB assembly turned out to be faster than I expected, so this will not take as long as indicated on the schedule.

The DigiKey cart has also been compiled for the PCB orders, so that can be sent out this week.

The final thing that I worked on was the strum detection circuitry. It is mostly working on a breadboard so I hope to perform some final testing of that before soldering it up.

My main deliverables for next week will be ordering the PCBs and sending out the DigiKey order. I also hope to solder the strum detection circuit on perfboard.

What did you personally accomplish this week on the project? Give files or
photos that demonstrate your progress. Prove to the reader that you put sufficient
effort into the project over the course of the week (12+ hours).
“ Is your progress on schedule or behind? If you are behind, what actions will be
taken to catch up to the project schedule?
“ What deliverables do you hope to complete in the next week?

This week I primarily worked on the initial design of the main PCB for the guitar that will connect the Teensy, Pi, and the the various sensors/outputs. The current plan for the PCB is to provide power via a 5V barrel jack. This will be split and sent to the Pi, Teensy, and LEDs. 3.3V will be supplied by the Teensy’s onboard linear regulator, and the ~200mA limit is more than enough for our 3.3V needs. The Pi to Teensy communication will be done simply through a handful of traces connecting pins on the Teensy to Pi GPIO pins. These will be used for both UART communication and for interrupts sent from the Pi.  The current plan is for 1 UART channel and around 10 general-purpose lines running between the Teensy and the Pi. This should be sufficient for the number of interrupts we are currently planning for, and allows for expansion in the future or the addition of some other form of communication between the Pi and Teensy. Both the Teensy and Pi will have a handful of their GPIOs brought to solderable pads to enable additional sensors to connect in the future. These will be accompanied by sets of GND and 3.3/5V pads on the board to power external sensors/devices.

The main other components on the board will be a 3.3V-5V level shifter for the LEDs, an active buzzer to serve as an easy-to-control metronome, and the strum detection circuitry. Since the guitar shipping from Amazon took longer than expected, the fine-tuning and testing of the circuit has been delayed a bit. Due to how sensitive the circuit is, I will likely try to keep as many of the components identical to what I used on the breadboard tests. I could switch to SMD versions of the components, but this introduces risks that could hinder the performance of the circuit. I also plan to include a handful of potentiometers throughout the circuit to fine-tune the performance on the PCB.

The overall shape of the board has been one challenge

I believe that I currently have the outline constructed and the mounting holes in the correct locations. As Tushaar and Ashwin establish communication between the Teensy and Pi, I have been picking out what GPIOs to route where on this PCB. The top right section will be used for connecting to the LEDs and will house the level shifter for them. All components in this area will be low profile so the USB port on the Teensy is not impacted. The empty space on the left will be used for the active buzzer and the strum detection circuitry. The edges of the board will be populated with various pads to enable connections to things such as the fret PCB D-Flip-Flops. Any major circuitry that needs to be added on will only need power, ground, and a handful of I/O pins, which will be available on the board. The circuitry can be put onto external perfboard/PCBs and wired up.

I also worked on testing a handful of methods of detecting the specific string that is strummed. I tested a capacitive sensor but I think the delay of that will be too high, and it will not operate correctly with something else driving the string. We did have one other idea; connect a GPIO to a metal guitar pick and every ~10ms drive it high (while the frets are all driving low) and check if the pick is in contact with any of the strings. Combined with the strum sensing, this would enable you to tell when the user strums and what string. This would have the same safety features as the fret sensors, so there is no hazard to the user.

I also did a bit of testing with active buzzers. While they are not typically designed to have their volume controlled by PWM, it is possible to control the volume of an active buzzer quite well as long as the PWM frequency is sufficiently higher than audible frequencies. This means that we would be able to have the user be able to customize the metronome volume if we choose to use an active buzzer.

Due to shipping delays, I will not be able to assemble the PCBs until next week. I hope to get them on Monday and assemble at least 1-2 by Tuesday night before our team meeting. This will allow Tushaar to begin to test their functionality. I believe if we can get this done before break we will be in good shape.  I will also test the strum detection circuitry on the actual guitar so that I can finalize the values and put them on the PCB. I hope to finish up the rough layout of the PCB this week and fine-tune it over break so it can be ordered soon.

My main deliverables for next week will be a couple assembled fret-sensing PCBs, more finalized values and layout for the strum detection, the strum detection tested on the guitar, and a rough layout for the Pi-Hat PCB

Weekly Question:

This week, the knowledge that I learned in the following classes aided my design process:

18320, 18623, and TAing 18100: These classes helped me understand the design considerations and techniques when building analog circuits. For 18320 and 18623 has helped me develop skills in hand calculating circuit performance, and 18100 TAing has given me experience in actually building circuits and making them robust. This was relevant primarily for designing the strum detection circuitry, which handles analog signal processing.

18240, 18320: These classes helped me understand the timing requirements and considerations when creating circuits with flip-flops, which we used in the design of the fretboard PCBs. This allowed me to understand the data given on datasheets for flip-flops, enabling me to be confident in the choice of flip-flop I made.

I don’t have any formal coursework in PCB design. However, for past personal projects and internships I have done PCB design. My experience in the past was with Altium and EasyEDA, but for this project I decided to use KiCAD. This was mainly done since KiCAD has become a very popular tool recently and I wanted to switch away from EasyEDA. I was mainly able to use my experience from previous design software to figure out how to use KiCAD, but I also found posts on KiCAD forums useful when learning application-specific keybinds and shortcuts.

This week, I finished the design for the fretboard PCBs and put in an order for the boards and the parts. The main changes to the design were the layout of the I/O pins, adjusting the spacing of the LEDs to match the guitar we are ordering, and adding a ground plane. The pinout was changed to make the PCBs easier to chain together without overlapping wires.

In addition to updating the PCB and ordering the parts, I also continued working on the detection of strumming in hardware. I have a design that currently functions on a breadboard but needs fine tuning. This will likely wait until next week when we have the physical guitar, as the tuning of values should be done for that system. The general processing done to the audio signal is shown below.

I have also begun the initial design process for the PCB that will house the microcontroller, strum detection circuit, buzzer for the metronome, logic level conversion for the LEDs, and power distribution. My current thoughts are to make a hat for the RPi. This will have the Teensy, a barrel jack for power, an active buzzer, and a handful of headers/pads to connect to the LEDs, fret sensors, and microphone.

I am currently on schedule. For next week I hope to test the strum detection circuit on the actual guitar. This will allow me to fine-tune the component values used for the circuit and ensure that the system will work. We have a fall-back plan of using the piezoelectric pickup that is built into the guitar. We can tap into this signal (or the amplified version) for likely a much higher signal-to-noise ratio signal compared to the microphone. I would also like to begin planning the general layout of the PCB, such as creating the general outline for the hat and the pin headers to attach the PCB to the Pi. If the fretboard PCBs come in, I also aim to begin assembly of the boards so testing can begin soon.

This week I primarily focused on performing an initial analysis of various aspects of the design. This consisted of determining how well a microphone will be able to pick up low frequency guitar notes, developing preliminary circuits for detecting strums, and measuring the latency required to write to WS2812 addressable LEDs.

For performing preliminary tests on picking up low frequency signals, I connected various electret microphones to an oscilloscope and observed their output waveforms when measuring a low frequency guitar note played from my computer. I found that using microphone breakouts with 50-100 times gain was more effective compared to creating an equivalent amplifier stage external to the microphone. The immediate amplification of the signal on a breakout board helped to eliminate most of the 60Hz noise that was introduced when routing the signal to an external amplifier stage. As such, if we choose to pursue a microphone-based approach, using a board such as Adafruits Electret Microphone Amplifier – MAX4466 with Adjustable Gain would be a good option. I did observe that low frequency notes were often drowned out by external noise and movement, which will significantly increase the complexity of a microphone-based approach for detecting strumming.

One observation I quickly made was that detecting a ~40Hz note in around 40ms using an FFT would be nearly impossible, since less than 2 periods of the signal would be collected. Additionally, measuring high frequency signals would require that the microprocessor is constantly sampling audio, which would take its attention away from driving the LEDs and talking to the Pi. As a result, Tushaar and I agreed that some amount of processing must be done on the analog side first. I experimented with various potential solutions and have had success with a 1st order low pass fed into an envelop detector, which is then fed into a differentiator circuit. The output of this system is a voltage spike that corresponds to sudden increases in the amplitude of sound waves. This would then be read by an interrupt on the microcontroller. This circuit has shown promise, but it will still detect sounds such as claps and output a signal. I hope that by putting the microphone in the resonant chamber of the guitar, or by tapping into some guitar’s built-in string pickup system, we can increase the signal-to-noise ratio of the system.

I also ran some basic tests on a strip of WS2812 addressable LEDs. From my testing I determined that writing to the ~56 LEDs that we will have should take around 2ms, which puts an upper limit of 500Hz on how fast we can sample audio signals.

This week I also worked on the design of the fretboard PCBs and picked out parts for them. They are mostly complete pending some slight sizing changes in order to best match the LED spacing to bass guitar string spacing. I hope to order the preliminary round of boards sometime this week. The design is made to be easily daisy-chainable. I decided to run two sets of power lines, one for the LEDs and one for the flip-flops driving the frets. I did this to avoid any ground-loop issues being caused by the voltage drop along the LED power wires. This does increase the number of wires that will be on the fretboard however, so it may be worth looking into consolidating for the final PCB run if this proves to be too many wires for comfort. The schematic and PCB were made in KiCad. I have not used KiCad in the past but I decided to make the switch to it due to the larger user community compared to my previous EDA tool.