I’m Dr. Feng, the designer of Muro Box, the world’s first app-controlled music box. Let me share my maker’s story, how a firmware engineer, without any prior mechanical background knowledge, invented Muro Box from scratch.
Before telling you my story, I would like to clarify that I didn’t mean to challenge anyone. The pros and cons of tools and methods may differ from person to person. My following comments only present my experiences in searching for the solution to create the first workable Muro Box prototype.
If you google “electric music box”, you will find many results. Most of the results are toy-like music boxes with piezoelectric speakers (buzzer) or loud speakers. Its operating principle is simple: one main MCU (Micro Controller Unit) with Flash Memory (internal or external, depending on its degree of integration) to record melodies. The sound from a speaker is driven by a PWM signal. Most of the specialized MCU for toys have dedicated a PWM hardware with an internal amplifier to drive the speaker. The following image shows an example of a typical electric music box.
However, this kind of electric music box is not what I was looking for. This type of music box neglects that the unique charm of a “true music box” is its mechanical movement which we’ve kept.
In other words, what makes a music box attractive is its rolling cylinder plucking the comb to produce the crystal-clear sounds of notes.
If you continue searching for “mechanical music boxes” online, then you will find many makers’ projects. I categorized these projects into two types as follows.
Their music boxes usually include a large wooden cylinder with many holes on it. There are wooden pins inserted into the cylinder at different locations on the cylinder to create melodies when plucking the comb made with iron or something that looks like a xylophone. The following image is an example of what their projects may look like.
Their music boxes often combine many solenoids or many step motors for controlling which note to be plucked. They use a set of arms or wheels to transfer the mechanical force to its musical comb for generating music. The following image is an example of what their projects may look like.
However, this type of music box is still not what I wanted to do. This type of music box is more like public artwork for exhibits. They usually have a higher budget for its materials and parts.That is to say, this type of music box cannot be mass produced to sell to general customers. Another limitation is that its size is often too large for most users in their daily life situations.
My first proof of concept prototype is shown in the following video. This prototype includes 11 solenoids to knock an 11-note xylophone. The testing MCU is Nordic 51822, a common BLE module. BJT are used to amplify the signals from MCU to drive the solenoids. The whole prototype cost me one week to complete. I googled a midi file of “Mary Had a Little Lamb” online for free download because my wife suggested this simple children’s song for my test. Fortunately, the midi file’s size is small enough for me to program it directly into the MCU to test playing.
Although the concept is simple, when I tried to carry out this plan, I ran into endless challenges and frustrations. My previous training as a firmware engineer only covered the technical knowledge of the electric parts, not including the mechanical design. Therefore, at the beginning of the Muro Box project, I had no idea of how to design any workable mechanical structure.
I was dreaming like many engineering guys who tried to invent something great. I thought once I finished the software coding, someone will magically show up later to help me fix the rest of mechanical problems and then the final product can be made soon. This kind of illusion stayed for several months, until my brother came to my home to wake me up with harsh comments and suggestions. This was a reality check I needed. After realizing the fact, I must put aside my coding work and search for professionals to help me solve the mechanical design problem.
My first step was to ask for help from a factory concentrating on music box movement in Taiwan. This factory (Kyooh Precision Industry Co., Ltd.) is the only music box factory in Taiwan, and it already has 40 years of experience in making traditional music box movements.
The fact is: Factory managers were too busy to listen to a crazy maker’s idea because they already knew that most makers’ ideas could never be mass-produced and will never become real products. After realizing the fact again, I knew that I had to solve the mechanical design problem on my own first to win their trust before asking for their technical support in design for manufacturing (DFM).
My second solution was to find design studios specialized in prototyping to help me. I learned this solution from visiting the exhibit of inventors’ works in the world trade center in Taipei, Taiwan. I asked the experienced inventors how they made their prototypes. Then, I realized that there are design studios to help clients make their prototypes.
The statement in this quote shows that I had to confirm the design and complete the payment within 7 days, no matter the prototype works or not. After 7 days, the contract would expire and no more revision could be made. However, a real workable product can never be fully tested in 7 days. There always be validation/development cycles. The service fee is too high with unexpected risks. Even though they said that they could provide feedback after the end of the contract, this service was not written in the contract, and I doubted about how much time they could spend on my case.
When looking back at how much time I have spent designing Muro Box, if a design studio can really make a good Muro Box prototype, it is fair for them to ask even 1 million NTD, and I will pay for it. In fact, the real problem is that Muro Box will never be a good product by relying on a design studio’s service.
An inventor, or a concept initiator, may naively think that the degree of involvement in product design of a design studio is as high as his. In fact, a design studio is just an agent trying to take as many cases as possible and get those cases done quickly to earn money. If I don’t know what the product should look like, how can an agent design the product for me?
At the beginning I tried to save money by finding a freelancer of industrial design online to take my case. I spent 20,000 NTD for the following illustration in four months. It turned out to be trash.
I still remember the day. I got this 3D model illustration and then I looked for factories to produce a prototype based on it. I was fortunate to run into a nice lady working in a factory. She told me why this was a problematic design. She told me an example that CNC machines cannot cut a 90-degree angle from inside, and the 3D model was full of this kind of faults. I suddenly realized that I hired the wrong guy. How could he even don’t know this basic limitation of a CNC machine? After this frustration, I realized that it was also my fault because I should have known that this person was not the right one from Day One.
Drawing a lesson from such a bad experience, I tried to go back to find a reliable product design studio. I hoped that I could meet a more experienced mechanical engineer to become my partner. This time, I finally met the right person who saved me from repeating the same mistake. Mr. Liao is how I called him; he is the very key person that saved me from spending more money and time on the wrong directions. He told me the truth:
I would thank Mr. Liao forever because he became my technical consultant for free even though he did not take my case. I started to learn everything in mechanical design on my own with the advice from him.Thanks to Mr. Liao’s advice, I made my first workable electromagnet sample as shown in the video below. The next step was to test its ability to pluck the comb of a music box.
At that time, I already accepted the fact that I had to rely on myself to bring Muro Box into life. By following Mr. Liao’s suggestions, I started to learn Fusion 360 to draw a 3D model of Muro Box piece by piece.
After discussing with several factories and design studios, I already learned that the cost of prototyping is too high. I had to find a cheaper alternative. Then, I decided to buy a 3D printer with the best customer reviews. Although my 3D printer, LulzBot Mini, is not the best deal with the lowest price, its quality is stable. I started to test various designs more efficiently with my 3D printer.
When looking back on my product development journey, it is worthwhile to buy a 3D printer to do tests at home. Each 3D printer has its own limitations and requires different adjustments to increase its precision. Only when I have tested it many times, I can finally get its best performance. Muro Box requires the dimension error to be lower than 0.1 mm, so understanding the limitations of a 3D printer is very important.
When the design of the mechanical part was about to settle down, I started to wonder how to design the look of Muro Box. I already knew that an engineer guy’s design is not going to help me attract potential customers, so I decided to find someone to help me design its look. Then, I started to find experienced design studios. An old friend recommended a design studio whose team members were industrial designers from big companies, and this collaboration gave me some hope.
Unfortunately, the experienced design studio still could not solve all of the problems in product design because I hope Muro Box to show its mechanical movement from the transparent lid as the key attraction and the design studio was afraid to handle the mechanical part. In fact, our communication problem in designing Muro Box is a typical problem faced by many product inventors. The product inventors often had the wrong expectations about the degree of involvement of a design studio and even expecting a design studio to solve all of the technical problems. When I heard that the leader of this design studio told me that he is just an agent taking my case, I knew that I had to rely on myself to complete the design of my ideal smart music box, Muro Box.
The product’s look was determined: the Proposal C in the above image was my first choice. When it’s time to assemble parts to test its music playing performance, I made a wrong decision again. I tried to save money by using a 3D printer for making the parts, but plastic material cannot replace the real metal parts in a mechanical music box. The truth is that the material matters a lot in order to produce the crystal-clear sound in a music box. When I replaced the original metal parts with the 3D printed plastic parts, I accidentally created new problems. Until this stage, I finally realized that saving money should not be the only thing to be considered in making prototypes.
Until now, I still often think that I already made enough mistakes and learned from them, but the next lesson from a new mistake came to me right away. For example, I once had an unfortunate experience with the electromagnet. When I needed the coil to arrive soon to complete the assembling tests, I did not realize that its factory was in full capacity a month before the Chinese New Year holidays. Their lead time was delayed for two months. In order to catch up on my test schedule, I decided to order the raw material of magnet wire to make my own electromagnet. That’s why I made a video to record how I made my own electromagnet for Muro Box, and this video was shared by many makers in Taiwan.
Now, when I looked at the classical music box movement, I realized that every tiny part of it was designed and crafted carefully after endless experiments. Although I was too naive to believe that innovating a music box was not that hard when deciding to jump into this trade, this music box innovation journey let me witness the efforts in manufacturing it. At least I could still share my experiences with other makers like me.
I believe that Muro Box as the world’s first smart music box will be the biggest innovation in the 200+ years of music box industry. I hope the mechanical music box lovers can program their own music, set up their own playlists with our app, and upload unlimited melodies to our music cloud library to share their favorite songs with other music box lovers around the world. I hope when the users are holding Muro Box at hand, they will remember how much time and effort I have spent on it.
