Week 01 (10 – 16 May)

Getting our idea took a while – Initially, we were keen on making a robotic dog that could serve as a companion for the lonely elderly and respond to various stimuli. However, we decided against it as we realised that the dog would be quite similar to those sold in the market. We explored alternative ideas, including a flying phone cover for drone photography, a pair of smart glasses for the blind, and even an automatic fried rice cooker!

In the end, our group of music-lovers decided to integrate our passion for music into our project by making a laser harp (Fig. 1.1). Excitement bubbled inside us like 4 cans of fizzy soda and we pitched our idea fervently to Dr Ho and Mr Tony. To our dismay, we learned that the exact project had already been done by another group of students from a previous batch. Plus, it was too simple to be considered an MnT project according to Mr Tony…

Disheartened, we returned to racking our brains. The same idea clicked in all of our heads – instead of playing musical instruments by ourselves, why not make a machine that could serenade us by playing an instrument?  Great minds think alike, as the saying goes! Ultimately, we decided that a lyre-playing machine is the most feasible, innovative and original! Dr Ho was pleased with our idea 😃 Hurrah!

 

Week 02 (17 – 23 May)

Our group had an online meeting to finalise our idea and to decide on our roles. As no one volunteered to become the leader, Bernice came up with the brilliant idea of using an online random name generator for leader selection. She hit the button, held her breath and closed her eyes to say a little prayer. However, that night was just not her night – she still suay suay became the group leader. Mesmerised by the beauty of a lyre, Vivian wanted to be the first in the group to get her hands on the instrument, so she claimed that she loved shopping and keeping track of finances. Sean asked tiredly, ‘safety do what ah?’ and he found the role of the safety manager to be quite appealing. Finally, Yisong said he ‘sui bian’. In the end, our fates were sealed as:

• Group Leader: Bernice
• Finance Manager: Vivian
• Safety Manager: Sean
• Communications Manager: Yisong

Later that week, all of us attended the 3-hour online workshops religiously. During the 3D printing workshop, Mr Hanyang introduced us to Fusion360 which could help us with the modelling of parts we may need for our projects. Additionally, he taught us about the slicing program and things we should note when using the 3D printers in the lab. We took down notes diligently such that we could put what we have learnt in theory into practice more easily when we got to the lab. During the workshop conducted by Mr Tony, we learned how we could make use of block diagrams to represent our ideas. Despite our best efforts, our heads spun with confusion when it came to the circuit diagrams 🥴. This was probably due to our (extremely) limited knowledge of electronics. Hence, we decided to do some self-directed learning!

As much as we were eager to go to the lab to take a look at the cool machines and be reunited with our fellow batchmates that week, an announcement was made for us to focus on design and purchase for the next two weeks due to the COVID-19 Heightened Alert.

 

Week 03 (24 – 30 May)

As a responsible Safety Manager, Sean reminded us to attend the risk assessment (RA) briefing conducted by Dr Alicia as all of us had no clue where we could find the online platform for RA submission. The session was really informative as she explained the risks that may be involved in each lab activity slowly and clearly. Somehow, her voice reminded us of a lovely module we all took last semester 🧐.

After the briefing, we met up virtually for a discussion on how exactly we were going to make the machine, where all of us tried our best to share our knowledge that might be useful for this project. Bernice took the initiative to create an MnT folder in Google Drive to facilitate collaboration and the exchange of knowledge.

As Bernice and Vivian were new to electronics, Sean and Yisong patiently explained what motors and Arduino boards are. Astounded by the fact that such a tiny Arduino board could act like a mini-computer, we were thirsty to learn more! Our group decided to purchase an Arduino kit to familiarise ourselves with how it works by watching a series of YouTube videos. We also decided to use servomotors for our project as they are not only small and light, but also budget-friendly.

With prior knowledge in playing the harp which is quite similar to the lyre, Vivian conducted a brief music lesson on how the lyre can be played – either plucked or strummed. While plucking the strings would produce a richer tone that resembles the sound of harps, strumming them would produce a crisper sound that is similar to that of acoustic guitars.

Harp:

 

Acoustic Guitar:

 

After an intense discussion, we decided to proceed with strumming using guitar picks. This was mainly because the plucking motion (pulling and releasing)  was more difficult to recreate as it would require higher torque especially when the lyre strings are very taut, and we were uncertain if the motors could generate the torque required.

To attach the motors to the lyre, we planned to mount the motors onto a rod and attach it to the lyre with a clamp that is similar to those found on a retort stand. As we were all gravely concerned about the clamp damaging the exquisite lyre, we came up with the idea of using rubber contacts between the lyre and the clamp. This would also minimise any vibrations from the motors and stabilise the whole set-up.

To better visualise what we had discussed thus far, all of us tried to draw a picture of the prototype that we had in our minds. Evidently, we were limited by our (non-existent) artistic talents🧑‍🎨 – the one that Vivian drew looked nothing like a lyre; Bernice struggled to draw in 3D; Yisong’s was monochromatic. We had a clear winner (Fig. 3.1)!

We constructed a block diagram to represent our AwESoME idea too (Fig. 3.2)!

The next day, we shared our vision and presented our sketch, block diagram, as well as a components list during our first draft presentation. Dr Ho said he looked forward to our final product, which got us really pumped in getting started on our project 🙂

 

Week 04 (31 May – 06 June)

This week, the long-awaited claims briefing was finally conducted by Ms Arelason – Vivian had made many purchases 💸 and was extremely kanchiong to get all her money back.

As we were unable to go to the lab, we met up virtually again to model the arm that would be used to hold a guitar pick on Fusion360. As a somewhat experienced Fusion360 user, Yisong explained various useful features of the application to us and he completed modelling the arm he had in mind in a flash. It looked really cool! Together, we made some improvements to the model and this is the final arm design (Fig. 4.1):

Unfortunately, little progress was made this week as we were all involved in the CNYSP FOP dry run.

Week 05 (07 – 13 June)

The COVID-19 Heightened Alert was over and we were finally allowed to go to the lab!

During the 3D printing workshop, Mr Hanyang taught us about offsets, slicing, the different types of infills, and how we should orientate our model for more efficient printing. He shared some limitations of the printers too. To show us how to operate them, he took us on a little tour around the MnT lab where the printers were. Fascinated by the sight of the printers in operation, we could hardly concentrate on what Mr Hanyang was saying oops. Later that day, we attended an interesting soldering workshop conducted by Mr Tony. We even made a digital clock😆 Just how cool is that??!

As we walked past the tables in the lab, eagle-eyed Sean spotted what looked like a servomotor, a stepper motor and a linear solenoid, all of which could be used to generate motion to strum the lyre strings. With these 3 options laid out in front of us, we became unsure of which motor we should use. After consulting Mr Tony and Justin, we realised servomotors might not be the most suitable for our project as they generate a fair bit of noise when they move, which may interfere with the music played by the lyre. More significantly, servomotors are not the best when it comes to the precision in the degree of rotation. Linear solenoids are noisy too and they are unable to generate the amount of force needed to strum the strings. The quieter stepper motors, on the other hand, are not only extremely precise, but they could also generate sufficient torque. It was just what we needed! The only downside of using these motors is that they are quite bulky and heavy. Hence, we decided to scrap the idea of using servomotors as we had discussed earlier in week 3, and use stepper motors (NEMA17) instead.

Since our lyre and Arduino kit have arrived, we thought we’d figure out the Arduino code to move one motor and try it on this stepper motor. We downloaded the Arduino IDE software to get started we connected the Arduino to the motor and a power source. Then, we typed out a code to make the motor work with Justin’s help. As the guitar picks we ordered have not arrived and the clamp for the lyre has not been made, we scurried around the lab in search of a makeshift pick and clamp which we found in no time at all. With our fingers crossed, we sent the code to our Arduino, turned on the power source and…

… IT WORKED! Hurrah! 🥳

 

However, there was just one problem… How are we supposed to fit 19 stepper motors across the lyre when just one covers about 1/5 the width of the lyre? I guess that’s a problem for another day as we were all drained after a long day 🥱

 

Week 06 (14 – 20 June)

To resolve the space constraint issue, we decided to keep only 10 strings and mount the motors on two aluminium rods across the lyre instead of one, with 5 motors on each rod. With this slight change of plans, we thought that using the clamp as discussed previously isn’t such a good idea after all. As our eyes shifted around the MnT lab in search of inspiration, Bernice spotted the 3D printer and a wonderful idea popped in her head – we could build a frame for the lyre by modelling it against the printer’s frame (Fig. 6.1)! We thought that it was a great idea, so we started scouring the lab for L-shape aluminium profiles and rods of appropriate lengths.

As Yisong constructed the frame, Bernice and Vivian removed alternate strings on the lyre such that there would be more space between the strings. To maximise the range of songs that could be played, the remaining 10 strings were then restrung with strings of appropriate thickness such that they could be tuned to play notes on the diatonic scale of C major, from C4 to E5. However, quite a bit of force was required to force out some of the obstinate strings from the holes. As they glanced helplessly across the table, they spotted Yisong and his enormous biceps💪 so they asked him for help (Fig. 6.2). After hours of restringing the lyre, it was finally done! Pheeeww! As an experienced string tuner, Vivian quickly did a few rounds of tuning as new strings go out of tune very quickly the first few times they are tuned.

Meanwhile, Sean worked tirelessly on a code that would allow 2 motors to work, taking into consideration the direction and degree of rotation. He tested out his code using 2 stepper motors found in the lab to ensure that they moved in the intended way.

What a productive week!

Week 07 (21 – 27 June)

With the completion of the frame last week and the arrival of most of our purchases, we decided that it was time to mount at least 3 motors on the frame. We came up with the idea of printing a mounting block to attach the motor bracket to the aluminium rod. With a surge of adrenaline, Yisong delved right into designing the block on Fusion360. He explained how screws could be used to hold the bracket to the block which, in turn, could be secured to the rod (Fig. 7.1). As a group, we improved on the design of the mounting block and proceeded to print them. We could hardly believe our eyes when we saw that the estimated printing time for each small block was 1.5 hours, which means it would be evening by the time we were ready to assemble the parts! Before leaving to fill our tummies, we started the print job for the arms which we had modelled in week 4 too.

As we entered the lab after lunch, Sean asked inexpressively 😐 , “3 notes can play any songs or not ah?” With her brows furrowed, Vivian replied monotonously to match Sean’s energy level 😐 , “Hmmm can lah… The melody of ‘Mary Had a Little Lamb’ involves only 3 notes C, D and E if we tweak it a little. We can try that today.” To facilitate the coding of the song, Bernice and Vivian then went on to calculate the delays (in milliseconds) between the notes using the note values and the song’s tempo. Excited to code our very first song, Sean hurriedly pulled his laptop out of his bag, plopped himself down on the couch, and started to punch the keys on his keyboard like a maniac.

The printing of the arms and the mounting blocks were finally completed and we assembled the parts and the motors with different kinds of bolts and nuts. With our hands, we tried to play the strings with picks of 3 different thicknesses – 0.46mm, 0.71mm and 1.00mm. We noted that the picks had to be placed at a more precise height relative to the strings if they were thicker. Hence, we decided that the picks with a thickness of 0.46mm were the best. After trying them out with the motors, however, we realised that the 0.46mm picks were so flimsy that they were unable to strum past the strings. The 0.71mm picks were able to strum past the thinner strings while the 1.00mm picks worked great for thicker ones.

TADAA! Our very first song!!

 

Week 08 (28 June – 04 July)

 

Week 09 (05 – 11 July)

As usual, our week started with tuning the lyre which frustratingly goes out of tune every few hours. We held a goal-setting session to ensure that we are all on the same page. This week, we decided to focus on mounting all 10 motors and play at least 2 more songs that are ‘not lame like a kiddy song’ as Sean says it 💁 . While printing more mounting blocks and arms, Yisong adjusted the frame and Sean wired the circuit boards and driver boards to the 10 motors. Meanwhile, at the piano, Bernice and Vivian worked on the arrangement of 2 more songs: ‘Hallelujah‘ and ‘Do Re Mi‘ (Fig. 9.1).

While arranging the songs, Vivian reflected upon the past few weeks and fell into a trance. Bernice realised that Vivian was staring into blank space, so she nudged her, ‘Eh, don’t slack leh.’ Vivian replied ‘Hold up. Remember how tedious it was to code for the melody of Mary Had a Little Lamb and Old Macdonald Had a Farm? They are such simple songs, yet it took us quite some time to calculate the delays between each note and write the code. Now that we have 10 motors that play not only the melody but also the accompaniment, how much worse is that gonna be? There has to be a better solution!’ Bernice nodded in agreement.

As they were bringing up this issue to Sean and Yisong in the lab, Justin overheard us and said, ‘If you could extract something like a MIDI file from Flat, you could possibly avoid hard-coding the songs!’ Our previously dimmed eyes filled with hope as we found the icon for exporting the sheet music as a MIDI file. Justin was really helpful and gave us a little crash course about this file type (Fig. 9.2). We learnt that each note has a unique number (for instance, C4 corresponds to note number 60), and the time information of each note is also included in MIDI files! We spent the rest of the day figuring out a code that converts the MIDI file codes into Arduino-readable codes. As it was getting late, we collected our 3D prints and quickly prepared the slides for the progress meeting the next day.

When we arrived at school for the progress meeting, we allocated a few slides for each member to present. During our presentation, we shared our Gantt chart, our financial status and the overall design of our machine (Fig. 7.1). We also talked about the progress that we have made thus far. Dr Ho, Dr Alicia and Mr Hanyang seemed quite pleased with our progress. Dr Ho suggested that we could get more lyres so that we could play songs in other keys besides C major without having to tune it. After giving it some thought, we decided that it was easier for us to transpose all the songs to C major instead of swapping the lyres for different songs.

After lunch that day, we mounted all 10 motors onto the frame and tried out the code for each of the 10 motors to play successively. When we turned on the power source, the current surged and a hissing sound was heard 🐍 😳. The guys remained calm – Sean commented, ‘What the… Why is it not working?’ and Yisong asked, ‘Where’s that sound coming from?’ Bernice and Vivian, on the other hand, scooted back immediately. Afraid that the motors would blow up 💥, Vivian and Bernice ran for their dear lives out of the Year Physics 1 Lab and into the MnT Lab to cry for help. Immediately, Justin dropped whatever he was working on and came to our assistance. For the next hour, we looked into every component as well as their connections. With so many wires all over the place (Fig. 9.3), it was really hard to find where the fault lies. In the end, Justin amped up the current and turned the power on. It worked! Apparently, the current was not set high enough, resulting in an initial surge of current. Justin stayed with us this whole time. We were so thankful for him!

 

Although we were exhausted after a long day, we connected all the wires back with ripples of excitement – With all 10 motors mounted on the frame, we couldn’t wait to try playing more advanced songs on the lyre. After converting the MIDI file code of ‘Hallelujah‘ into a code that is readable by Arduino, we sent the code to the Arduino, turned on the power source and watched intently. OoOof. Although the motors worked just fine, the initial jerk sent all the arms out of position when the power source was turned on, making it difficult for the strings to be strummed consistently… This was our best take 😓:

Bummer… We had to find a way to resolve the jerking issue!

 

Week 10 (12 – 18 July)

Inspired by the 3D printer, we decided to home the positions of the arms before each song by installing a limit switch beside each of the arms. After modelling a switch mount on Fusion360, we proceeded to print it. While waiting, we also decided to cushion using foam to ensure that the lyre does not go out of position. With Justin’s help, we also figured out the code for the electrical homing and it worked! Check this out!

 

We spent hours printing the mounts and installing all 10 limit switches and at this juncture, we realised that there would be a significant mess as each switch required 3 wires. How would we manage so many additional wires on top of the original clutter? We sought Mr Tony’s advice and he suggested that we could make use of mechanical homing instead of electrical homing. On the way back to our hall, we thought of a fabulous idea (or so we thought)! We could reattach the alternate strings which we removed and restring them at a much higher position. In theory, this should stop the arms from going out of position.

The next day, we tried it out and realised that the torque generated by the motor was so great that the picks strummed past the higher reattached strings too. Disheartened, we gazed desperately around the room to see what we could use for the mechanical homing of the arms. At that moment, Yisong spotted a long screw on the table which could be slotted into the hole of the switch mount to stop the arms from moving past a certain point. With a stroke of brilliance, Yisong exclaimed, ‘We could use this instead!’. We tightened the screws on the mounts and attached them to each motor bracket (Fig. 10.1).

After which, we decided to play Married Life on the Lyre with the 10 motors mounted. Although it was (very) out of tune, it worked! Notice how the positions are set at the beginning of each song. Cool right?? 😆

 

Having figured out the homing of the arms, we returned back to hall in high spirits 🙂

 

Week 11 (19 – 25 July)

Since we have already figured out the code for many motors, and Bernice and Vivian were having a really hard time arranging songs with such a limited number of notes, we wanted to explore the possibility of restringing back 2 more strings (F5 and G5). However, there are some strings attached to this change – the physical intervals between the strings would no longer be constant and there would be some space constraints for the motors. We looked into alternatives and we all agreed that the best solution was to purchase another lyre with 24-strings. This means that there will be enough space between the strings. We would simply need to adjust the lyre frame to make space for the greater width of the new lyre and adjust the positions of the motors.

Sean and Yisong suggested that we arrange a Pop Song Medley as most people listen to pop songs. With her limited knowledge of music of this genre, Vivian was struggling even in song selection, so Bernice her helped out and they worked together on the arrangement. Due to her love for pop songs, Bernice proceeded to arrange a couple more songs, including ‘Hey Jude‘ and ‘The Scientist Coldplay‘ while Vivian arranged ‘Let it Snow‘ and ‘A Million Dreams‘. Phew, arranging music became so much easier with 2 additional notes!

Since our machine was functional, we thought we could work on the appearance of our product. We decided to encase the electrical components using an acrylic box, and possibly install LED lights in the box which will light up whenever a note is played 🤩. Before calling it a day, we created a design for the front of the acrylic box:

Week 12 (26 July – 01 August)

Our new 24-string lyre had arrived. With much fervour, we unboxed it –  it was simply beautiful (Fig. 12.1). As Vivian couldn’t sleep because of a big fat cockroach 🪳  in her hall room, she arrived early that day to remove the alternate strings and restring the ones we needed with the correct thickness. When Bernice arrived at the lab, she joined her in forcing out the (super sharp) strings from the lyre. Their fingers got sliced a few times🩸. With their battle scars, they soldiered on and by the time they were done, half the day was gone.

As the dimensions of the new lyre were different, Yisong made some adjustments to the aluminium frame. He cut the aluminium rods for a better fit and adjusted the height of the motors. To accommodate 2 additional motors, he extended the width of the frame using additional aluminium rods on the sides. Meanwhile, Sean worked on making the wires neater, bundling related wires together using tapes and labelling them (Fig. 12.2). This makes it easier for us to troubleshoot in case something goes wrong in future.

After lunch, Bernice and Vivian went to collect the acrylic box that would enclose all our electrical components. As Bernice had just gotten her driver’s license, Vivian was excited for Bernice finally to drive her. Bernice, on the other hand, was afraid of wrecking the car. But yes, they made it alive (Fig 12.3) 😃 Bless!

When they returned, we arranged the electrical components in the acrylic box and reconnected the wires in the box (Figure 12.4), trying our best to reduce the clutter.

Week 13 (02 – 08 August)

This week, we focused on adjusting the new frame. Unfortunately not all the strings could be reached by the arms because the lyre strings are spaced differently now. Hence, we modelled a longer arm and sent it for printing. We decided that we would not stick with playing songs that only contain notes in the diatonic scale of C major, but instead those that contain any 12 notes. Sean updated the code to make way for the two additional notes and any sharps or flats that would be played. We also arranged more songs including La La Land’s “City of stars” and a Disney Medly!

For the millionth time, we adjusted the pick positions and the physical homing to make sure that the strings could be strummed.  Taking a step back to look at the machine, Bernice and Vivian were extremely bothered by the aesthetic of the machine – the bolts and nuts used were of different types and the frame was asymmetric (Figure 13.1). Just at that moment, Mr Tony walked past us and took a glance at our machine. He said, “Y’all very funny leh 🤡. The motor bracket that you bought can be directly mounted onto the frame, why use a 3D printed mounting block??”

We thought to ourselves, “That can’t do!”. We were determined to improve the aesthetics of our project by making it look less cumbersome and congested, even if that means taking the trouble of dismantling the frame we already did so far. To do this, we needed M4 T nuts which we did not have and are not commonly found in hardware stores. Vivian and Bernice called various hardware stores and drove all around Singapore while the guys dismantled the entire frame and cut the aluminium profiles to specific lengths. After hours on the road with an empty stomach, they finally found what they needed and cried tears of joy😭🤩  (Figure 13.2).

When they came back, the four of us worked together like a well-oiled machine. We assembled components of the frame back together. Our backs and arms all hurt at the end of the day from bending down and being in all sorts of positions (Figure 13.3). To remind themselves which way to turn the screwdriver, Bernice and Vivian chanted ‘RIGHTY TIGHTY, LEFTY LOOSEY’. Looking at our ‘new’ frame, we believe that it was entirely worth it (Figure 13.4).

 

 

 

 

 

 

 

 

Week 14 (09 – 15 August)

This week, we painstakingly fitted the lyre in the frame using sponges (again), arranged the electrical components into the acrylic box (again) and adjusted the arm positions (again). Much to our dismay, the screws which we had decided to use to home the individual pick arms could not stay in position despite our best efforts at tightening them. An idea clicked in Sean’s head – we could use another screw to home the homing screw!

 

Week 15 (16 August – 22 August)

Since each arm that at a fixed position, we coded the degree of adjustment for each arm such that the picks were close enough to the respective strings to strum them. Check out our homing!

We tried to play (part of) the Disney Medly and…

…. it was our first win in WEEKS 😭. While we were discussing the most suitable song that we could play for the live demonstration during the presentation, we saw Dr Ho walking down the flight of stairs. we ran to ask for his favourite song. Asked Dr Ho for his favourite songs and he said  ‘Always with me‘ and ‘Xiao Xing Yun’. Sean requested the arrangement of ‘Stay‘ by Justin Bieber and Vivian proceeded to arrange the songs. To improve the aesthetics of our machine, we installed LED lights and included the lights in our code so that different coloured lights would appear when different notes were played. We also improved the aesthetics of the lyre box by lining it with some shimmery gold paper! It looked AMAZING 😀 (Fig15.1).

 

 

Week 16 (23 August – 29 August)

Having arranged the songs that we planned to play with the lyre, we tried to play them using the machine. After filming the machine playing all the songs that we had arranged thus far, we proceeded to edit the video late into the night🦉 We were all so proud of our product that it ended up on all of our Instagram stories the next day! We were finally done HURRAH!

This marks the end of our MnT journey!

AU REVOIR~