How I learned to make my own speakers
Why my summer was awesome despite not going as anticipated
Before the real article starts, I would like to state that this was something I had written in September shortly after the summer was complete. However, I did not post it because I felt that it was not written well and was too hard to understand due to some jargon.
This is being posted now with edits because I’d rather have this out in the world, than not at all.
In a previous post, I had publicly announced that a larger-scale project was coming from me this summer. It didn’t come. I had expected my summer to be filled with plenty of time for me to do my personal projects, a ton of reading and to learn to play my new electric guitar. None of this happened.
Yet, this past summer was a great one.
My promises were not fulfilled, but it was for a good reason — my Model and Prototype Development class (called Model Shop for short). This has hands-down been the best class I have taken at SCAD. Although the class falls within the Industry Design faculty, it is a required class for me as a User Experience Design student. Despite the fair amount of overlap between the two industries, much of what I learned in that class is unlikely to be directly related to any work I do in the future. Regardless, Model Shop was an absolute treat. In this class, students learn the various tools that are required to make prototypes and develop future products. The variety of lesson topics range from foam model making to woodworking to vacuum-forming plastic.
The Final Project
The final project for the class was to design a speaker. It was the perfect way to take the various skills we had learned and apply them in order to create something that was functional.
Initial Brainstorming and Decision-Making
The project did not outline specific design elements and many of the variables were up to the maker. You could make a small portable waterproof speaker out of PVC pipe, or massive wooden speakers that produced thunderous bass. As a result, there were little constraints.
Constraints aren’t necessarily a bad thing. As seen time and again, constraints can breed innovation. As a result, I wanted to figure out what constraints I wanted to work within.
In order to do so, some questions had to be answered.
Who was I designing for? What type of speaker was I designing? What features did I want to include? How many drivers was I going to use? What was the budget? (Side note: while speaker-making can produce higher quality sound at an equally-priced branded speaker, it can still become an expensive endeavor.) What type of music was I going to be listening to? And from what device(s)?
These were some of the many questions that were running through my mind. Having taken in all the considerations, I watched various YouTube videos about speaker-making (shoutout to KirbyMeetsAudio) and then let my mind do some of this thinking unconsciously for a few days.
A few days later, while waiting for the paint to dry on one of my other projects, I sat down with my dotted sketchpad. I began by drawing some quick sketches and annotating those sketches with some details and ideas. I knew I wanted a speaker that was relatively powerful, and portability was not something I cared about. It just had to stay in one place. I thought perhaps I could incorporate a digital assistant in it, such as Amazon’s Alexa or the Google Assistant. The two ideas that I really enjoyed were a boombox-style Google Assistant speaker or two bookshelf speakers that could be used as a stereo home theater setup.
Focusing on these two ideas, I went to my professor to discuss what I would need to buy and how feasible each idea was. He showed me the drivers, amplifier and power sources I would need to buy for each set-up. We also talked about what the advantages and challenges of each system. Due to the unnecessary complexity, additional costs of the board required to compute a digital assistant, and desire to create something for my parents, I decided to go with the home theater set-up.
Fleshing the Design Out and Developing the Foam Prototype
The next part of the process was creating a foam prototype to ensure we made an appropriately-sized enclosure for the system while also getting an understanding of the steps required to create the real enclosure. This step was the professor’s way of making sure we did not mess up with costly materials.
At this point, I knew the general direction I wanted to be headed in. There were going to be two speakers, a left and a right speaker, allowing for stereo sound. Each speaker itself would have two-way crossover design with a woofer and tweeter.
I know you may not have understood a word of what I said above, so let’s quickly go over the terminology.
A speaker driver is essentially a motor that produces audio by moving a diaphragm back and forth; it is what most people consider the speaker itself.
A woofer is a loudspeaker driver that produces relatively low frequencies.
A tweeter is a loudspeaker driver designed to produce higher frequencies.
A crossover is “an electronics device that takes a single input signal and creates two or three output signals consisting of separated bands of high-, mid-, and low-range frequencies.” (Source) It consist of resistors, capacitors, and coils. In my case, I needed to split the input audio into two (hence a two-way crossover) so that the highs could go to the tweeter and the lows to the woofer.
In terms of materials, I wanted to have a nice wooden design that was complimented by a textured fabric. My initial design was a vertical box with tapered cylindrical legs to give it some height.
However, with some time given for my idea to settle, I decided I wanted to do a horizontal set up so the speaker could in some capacity serve as a table. The top surface would also extrude to offer some more surface area. The idea for tapered legs was kept, but instead of a cylindrical form, I went for a rectangular design.
With the design in mind, I had to figure out dimensions. The dimensions could not be whatever I wanted because there were many factors that had to be taken into consideration. When designing speakers, the enclosure volume is not to be taken lightly because it determines the sound. There is a significant amount of physics behind this, but for our class’s purpose, we did it in a rather unscientific but still informed way. I needed to hit around .52 cubic feet of internal volume — keyword here is internal — meaning my calculation needed to take into account the width of the material being used. The challenge here was finding appropriate proportions while hitting the desired volume. After a while of tinkering with the math(which could have been cut down with knowledge of 3D modelling software), I figured out the dimensions of the enclosure and began cutting up my foam pieces.
Testing Out the Speaker Drivers
Part of the reason we built the foam prototype was to have an enclosure to test the sound in. Although foam and wood have different acoustic properties, one could get a good sense of how the final product would sound in the foam prototype.
The first bit of audio I successfully tested was my woofer driver. It had a good amount of bass but sounded muddy. This was because I was still not using the tweeter and I did not have the crossover setup to test the drivers in unison. A crossover is how a speaker understands which frequencies to play when there is two or more types of drivers present. Because I had a woofer and tweeter setup, the woofer had to play the lower frequencies whereas the tweeter would hit the higher frequencies. That’s why when I tested without the crossover, the woofer driver sounded muddy. A woofer is better suited for lower frequencies and when forced to play higher frequencies, it will not produce pleasing results.
Because my speaker components were part of a kit, the crossover was already designed for me and all I had to do was put it together on the PC board. Despite this, I faced a lot of issues at this part of the project. My lack of soldering knowledge made things much harder; I made many simple beginner mistakes that drained some of my time. Luckily, my dad who works with electronics for a living was available over video call to help me troubleshoot. In his typical manner, he told me how I was making so many small mistakes and the solutions were super simple. After making the changes, my crossover worked liked a charm and I got a rough estimation of how my speaker would sound.
Material Choice
With a design to aim for and confirmation that my drivers worked, the next step was buy the building materials. The two types of wood I used were Medium-density fibreboard (MDF) and southern yellow pine. MDF, although considered a cheap and ugly-looking material, is a good material to use when making speakers. Its acoustic qualities are great, so much so that it is even used in speakers that cost thousands of dollars. Furthermore, I was going to cover the MDF with fabric anyways. The reason for using pine on the other hand was that it was a wood that would stain well. Although staining wood does not produce the same results as buying the right wood does, I needed the wood to look dark like walnut to match my hardwood flooring. However, the one constraint the professor had set on us was that we were not allowed to use hardwoods because they can be difficult and potentially safety hazards for first-time woodworkers. The fabric I chose to use was a dark blue denim because it had both the color and texture I wanted.
Time for the Real Deal
It was finally time to start making the speakers themselves.
With this being my first large-scale woodworking project, I was nervous to say the least.
I started by cutting up the MDF to make the main box of the enclosure. I used mitered cuts of 45 degrees to make the joinery easy. This process took a little time to perfect because I needed to repeat cuts of equal length quite a few times. I ended up using much more wood than I should have, but luckily I had bought more than enough MDF.
With the enclosure’s cuts done, I had to make holes for the speakers to fit into using the plunge router and a circle-making jig my professor had made specifically for this class. For my design, I wanted to flip the circle cut-outs from one speaker to another because the hole for the tweeter was much smaller than the woofer. As a result, when placed beside each other, they would look like a mirror of each other, rather than an awkward copy.
I also had to make my legs. The process for this was quite long. I had to first plane and joint the pine I had bought to make sure the surfaces were flat and perpendicular. Then, I made cuts to match the rough length and width I wanted. Because one board was simply not thick enough for my leg width, I had to glue 6–7 pieces together. After the glue dried, I needed to make the final cuts. I, first, made everything square. The next part is where it got a little bit trickier. Although we learned various tools and jigs in my Model Shop class, I did not learn how to use the taper jig because it is rarely used in Industrial Design and is rather a furniture design tool. My professor very kindly took time out to give me a demo on the jig. He even helped me set up the right angle for my cuts. After that, I just had to make various passes through the table saw using the jig.
The last major component I had to work on was the top panel of the table. I ended up planing my pine too thin for an appropriate top. As a result, I had to improvise. I cut MDF boards to put below the pine and glued the panels together. After it dried out, I used the disk sander to give the panel a radius and then hand-sanded to make the curve smooth. With that done, I used a chamfer bit on the router table to chamfer the edges on both faces of the board, with the top chamfer being more pronounced than the bottom.
The Cast Element
As part of the project, students were required to cast an element of their design. This was to ensure we learned all the tools we could use to make products in the future. The cast element I decided to do was simple yet important to my design. It was going to be the letter K to represent my last name, Khatri.
I created the initial reference model from high-density Ren foam. Don’t let the term foam trick you. The foam here acts much more like a softwood such as poplar or pine. I cut my general shapes using a band saw and scroll saw and then, sanded from 120 up to 2000 grit to ensure a nice mold (as the cast will pick up all imperfections of the original)
After that, I used Mold Star 16 to create the silicon mold. The process for this is pretty easy (I’ve skipped some parts in my explanation, but this should explain the general process). You, first, measure equal parts of Part A and B and then mix them together. As a result of the mixing, a chemical reaction begins to happen and the new liquid begins to solidify. Before the reaction gets too far into the process, you must pour the liquid over your reference model in a container. Once this is done, you put the container in a pressure pot to take out any bubbles that may have formed from pouring. After 30–45 minutes, you can take the container out. Lastly, remove the mold from the container, and your reference model from the mold itself. Now, you have your own silicon mold that can create objects replicating the reference. This mold can be used to cast various materials including metals, plastic, and concrete.
The process used for casting is very similar to mold-making process. I used Smooth-Cast® 326, which is basically fast casting resins that are able to color match. I also bought gold mica powder to make my cast look like gold. The only real difference in preparing the cast resin compared to the mold is that you mix in dye or, in my case, the mica powder into Part B before mixing part A and B together. Here you want to leave the resin in the pressure pot for 1.5–2 hours.
A Crunch to Finish
With all major componentry done, I needed to do the finishing steps. This had to be done in a serious time crunch and sadly, I had to sacrifice some quality to meet my class’s submission.
I had planned on covering the main part of the enclosure made of MDF with the dark denim I had bought. First, I had one of my friends cut the denim for me to the appropriate size using fabric scissors. There were two main ways I could have gone about attaching the fabric to the MDF, spray adhesive or staples. Because spray adhesive is quite unforgiving and achieving appropriate tension was very important, I chose to staple it. This choice did not make the process as easy as I thought it would be. I had cut the fabric to the right length but I wish I had given myself more slack to pull the fabric from the inside. This part was also a two person job because one person needed to be pulling to achieve appropriate tension whereas the other needed to be stapling.
I also had to stain and spray paint parts of my project. This part had to be done in a rush because I was limited on time. Unfortunately, this part of the process is very time sensitive. Spray paint and stain both need time to dry between coats and after the final coat. On top of that, the high levels of humidity in Savannah do not make this part of the process any easier. Even 24 hours after spray painting, there was noticeable stickiness on the surface indicating it still needed more time to dry. The wet paint slightly damaged my pieces right before submission when I stacked the two top panels on each other. On top of that, when I taped the painted bottom half to stain the top, the tape left marks on the painted surface. The fixes for both these problems aren’t necessarily hard, but it was wasted energy and time.
Showcase/Reflection
This project really pushed me. It was my first proper woodworking project ever. On top of that, I didn’t choose an easy design to make. There were some complex cuts, especially the tapered legs. With that being said, I am very proud of this project and can’t wait to make more with the skills that I learned here.
The few things I want to change are minor but important finishing touches. It’s not even the grade that matters here (which I did fine on anyways). Having spent so much money on the product, I just want a properly finished project. These finishes include polyurethane application on all surfaces, tighter and less messy attachment of the denim, better soldering and some other small quality-of-life changes.
In terms of sound, I am more than content. Since I used two drivers, a woofer and a tweeter, I got great quality on both the highs and the lows. The highs were especially what surprised me. For the class presentation, I played Sanctuary by Joji and if you’ve heard that song, you know how crazy the vocals and highs are. The speaker reproduced that audio with fidelity that truly blew me away.
Update: Finishing Touches
After summer, when my parents drove down to Savannah to visit, they took the speakers back home so that I could finish them properly when I came back home for the winter break. Luckily, my dad was at home to help me with a bunch of these finishing touches. These touches had less to do with the looks and more with functionality and longevity.
Here are a list of the finishing touches we did:
- We re-soldered some of the points that were questionably soldered for a stronger hold and better connection.
- I redid the spray paint on the bottom of the top panel due to problems caused by humidity on the original paint layer.
- We added foam on the inside speaker to help with sound quality.
- I added another coat of polyurethane on the top panel. I did one coat of polyurethane on the legs.
- We bought banana plug cables that made our system plug-and-play rather than using stripped wire ends.
- We attached the fabric in a more appropriate way using nails, instead of staples.
With all these changes, I can officially call this project complete and I am so ready to use these speakers to watch some movies and listen to some music.
A Thank You to My Parents
Before I end this post, I just want to thank my parents for all the times I haven’t been able to say it. I don’t often verbalize how grateful I am and this speaker project is a gift I truly want them to be happy with. The beauty of this gift is that it’s custom-made and can last a lifetime.
If you are reading this Mom and Dad, your trust and faith in my choices is something I appreciate more than you realize. When I chose to come to SCAD, you guys were two steps ahead of me in preparation and did not worry about the feasibility of my choice. You knew that I was capable and allowed me to realize my aspiration. My first project of this scale only made sense to be dedicated to you guys.
Thank you truly!
