FORM: 026 Speaker

Hello again everyone, here’s another project I was working on for my first semester of ID grad school at UIC in Chicago.

This project was an assignment where we were to take cheap PC speakers from Amazon and transform them into a new product using the same internal components with a specific design goal in mind.

My design goal was to limit myself to using simple cube form and experiment with how form, details, and refinement could increase the perceived value of a product instead of using something more direct like expensive materials.

The idea of this speaker was take the speaker grill and abstract it so that the whole speaker body acted like a grill and let sound pass through it.

Here’s a link to the project for bigger pictures and thanks in advance for your feedback!

I wanted to give project documentation an overhaul so I reworked this, trying to push the graphic style and rendering detail (electronic components and hardware).

I’ve updated the previous post with new images, check out the new version above. Thanks!

The nature of the construction you have chosen (individual vertical slats) really highlights ANY imperfections in your parts, alignment, gaps, and tolerance. For this reason, your renderings are more successful than the physical model images for beauty shots. I would show a couple shots of the physical model, but would limit them to the more informative/construction level, not for aesthetics unless you plan on creating a new/better/more precise physical model. I would also point out that the issues you are having in your own physical model may be an indicator of issues you could see in production if you intended to assemble it in a similar manner, this could also be used as manufacturing/assembly key learning that could lead to you ultimately reconsidering how you would go about manufacturing or assembling the product. In that case, I would use the the physical model images, call out the imperfections (especially the the top back edge line) and show the corrective actions you took to eliminate the issues you found in this model.

I’m sorry I missed this when you posted it the first time. I meant to comment on it and then forgot.

Kudos for making this thing first off. Not only making a model, but making a functioning prototype.

Something you wouldn’t know if you have never worked on speakers is that deep slats are bad for sound. The waves bounce off those layers and start canceling or distorting each other, so those slats would need to be as thin as possible in the cone of interference from the driver. An acoustic engineer would want to make them about 1-2mm thick (like I said, ver thin) which owl make them look not very good, so then you would need to back them with scrim… not a level I would expect from a student. but I know you geek out o this stuff, so I thought you would like to know that.

A few examples of this in production products I’ve worked on:

Polk Camden Square. Notice how thin depth wise the lattice is, you especially want that in front of tweeters because the wavelengths are shorter. Not important in front of woofers, mid range drivers you could got a few mils deeper: https://www.polkaudio.com/products/camden-square

Polk N1. This bar didn’t have tweeters, but instead had an array of mid woofers, so we got away with more depth, probably about 5-6mm: https://www.polkaudio.com/products/n1-gaming-sound-bar

Lastly, doesn’t have slats, but a similar square form factor. Here we tough advantage of the huge size to have very directional left and right tweeter array (4 tweeters at various angles) then a giant upward facing woofer and mid range: Definitive Technology's Cube | Wireless Bluetooth Speaker - YouTube

OK, lastly, I’d like to see you align the visuals. The design is a cube with a triangle lopped off. The fins have a radius on the ends. Make everything relate to that. Where the AC cord comes in can have more of this feel, at least use the same rads on the part. The volume slider is a circle, I’d expect to turn a circle. If it was more of a triangular thing it would have visually said “slide me” more. Also, I could see some LEDs behind those slats to show volume level or mode or something?

Lots of notes, sorry to nitpick, but I worked on speakers almost exclusively for 6 years. Still a strong project as it.

@IDiot

Thanks a lot for the feedback. That’s a really good point about calling out the imperfections on the physical model. I should totally use it and the issues I had making it to feed directly into improvements for rev 2 of the model. That will also let me end with some new and improved beauty shots. Great ideas!

@yo

Thanks so much for sharing your experience in designing speakers, lessons learned for sure! I appreciate the geeking out :smiley:
When you design speakers are there any text books or other resources you find useful for speaker design? I haven’t worked with an acoustic engineer before and it’s really interesting to see how the physics affects the final design. Really good suggestions about the model, I’ll do a little research on sliders and work those and the visuals into rev 2. Huge thanks!

I imagine this construction would definitely impact the sound.

About the form, what can I say, it is an original idea, daring and high end in execution. You don’t challenge low-cost manufacturing that much though as such an expression can also be achieve using molds. Explore more and look into complex patternings and construction, for example those enabled by digital fabrication. Overall I like your product for its simplicity.

@ralphzoontjens

Thanks! Yeah I agree about challenging manufacturing. I think I was thinking of the low-cost in terms of using digital manufacturing based on the construction of the prototype which was laser cut from plywood and the inner frame which was FDM 3D printed and maybe not needing to invest in molds. I appreciate the feedback!

Nice process, I love that you actually made a functional model.

I don’t know of a book, but there’s a great audio forum called DIY audio that’s jammed full of every bit of information about home speakers.

Knowing how to house the speaker to suit its electromagnetic properties (qts is a big one, you’ll probably understand what this actually means in practical terms far more than I do :laughing: ) makes an enormous difference to the sound it produces. Ideally, you don’t want the sound waves from the front of the speaker cone to come into contact with the identical wave coming off of the rear of the cone, as they cancel each other out and the sound quality is reduced.

Have a look at different methods of installing speakers into enclosures (sealed, ported, different orders of band pass) and how the amount of air inside the chamber can alter the performance of your speaker. There’s a great free software called WinISD that helps you to design enclosure volumes and ports etc around different speakers if you wanted to have a play.

As far as the sound/form relation, this made me think of the Bang & Olufsen BeoLab18 which has a similar idea with the slats. I’ve only heard a few of their products so I can’t speak too much from a sound perspective, but I would hope the price tag would warrant great sound. They’re also beautiful.

B&O stuff in general does not sound that great. It is more of a design and brand play… they are also not doing very well as a company. I dot a demo of their flagship product, I think they are $20-30k a pair. I was surprised how muddy the sonics were.

@AndyMc
Thanks so much for the info! I can tell that cruising the diyaudio forum will teach me a lot. WinISD is exactly the kind of thing I was looking for, I’ll definitely give it a try for version 2. :slight_smile:

Hey Gerry,
A few notes:

  • The slider looks like a push button or a dial, definitely not a slider. That being said, the way it’s integrated with the housing is clever.
  • Overall presentation is clean. Nice.
  • Love to see the prototyping.
  • You say “low cost speakers are designed for manufacturing, not for the users.” But then you don’t solve a user-centered problem. If you’re adding value by refining the form, that’s fine! Don’t try to make the project sound like something it isn’t.

No problem. I can’t resist an opportunity to turn someone else into an audio nut :laughing: