How do your speaker designs usually begin?

[Psycoacoustics]

Agreed that princess phone is useless, however I can use that jack if you are ready to get rid of it.

Thanks buddy, I needed a good laugh today!

I’ve only built three pairs. The first one was a copy of Direct Acoustics Silent Speaker. I read a review, thought they were cool, got tons of help from the people here and made an ok speaker.

The second pair were the Caritas. I liked the charity aspect, and wanted to do a hardwood baffle because most everyone here said it was a bad idea. They still sound and look great! Must be because of my super secret hardwood baffle attachment method! LOL!

The third pair were inspired by winning a pair of tweeters at InDIYana. Wolf helped me a ton, and they sound so good I have trouble justifying building more.

I really envy all you guys like Tom that have such vision! I can build, but for design, I need a muse!

Have Fun! Mark

[jhollander]

Agreed that princess phone is useless, however I can use that jack if you are ready to get rid of it.

[djg]

One interesting (to me) concept that was not built was a Linkwitz LX521 design. I designed a functional 1/3 scale LX521. I thought it was cool enough that I emailed SL to ask if it was OK (I heard he considered the shape to be IP). He said it was fine to build, but questioned why would I make something that had obvious short comings. I guess his definition of cool was not the same as mine.

Totally cool, totally useless.

[guitar maestro]

My last one started with a fat wallet, and right now it's very anorexic.

[bjaurelio]

Except for a sealed subwoofer, which really doesn't count as a "design" I've only designed my outdoor speakers. I should be posting an update soon once they're painted. I wanted to try controlled directivity by matching an 8" woofer to the SEOS waveguide. When I determined that an outdoor speaker would be the best place for me to put the speakers, that added constraints of low price and poly woofer. After looking at driver options to meet those constraints, I really only had a single woofer option. I selected a relatively similar priced Dayton tweeter since it was well reviewed on PE for a $20 tweeter and was the right size to fit the SEOS.

I have ideas for future projects, but they all revolve around different ideas I want to try (transmission line, coaxial for in ceiling atmos speaker, high end drivers when I'm more confident in my skills, etc)

[civit]

I think I have four motivating patterns for speaker design:

1A What is the best sound I can get in my living room within reasonable size constraints? These designs tend to have less focus on aesthetics and simplicity - I want a speaker that looks like I'm a hobbyist here. These designs are becoming less important to me. For these, I generally think in terms of what cool drivers I want to use, since they are ultimately what limit performance here. These are on hold until I see what wavecor offerings PE will have. I'm adding two SDX10 sealed subs here, so future designs don't need to go down past 50hz or so.

1B Designing based on a technology or feature I'd like to experiment with or hear. Transmission lines, open back or dipole midrange, driver arrays, low order crossover slopes.

2. Designing based on a specific use case - normally for other people, but sometimes myself. I needed some small, full range desktop monitors. My friends have a new place and could use something which fits with the decor. My brother likes skinny towers. Something designed for corner placement. Actual 'bookshelf' speakers. These are probably the most interesting designs, and are a good way to use up spare drivers (my spare drivers aren't anything special, but certainly exceed the quality of those used in sub $2k speakers.)

3. Designing speakers that make reference to the past in some way - a big 3 way with modern but low cost drivers - a modern reinterpretation of an interesting design, such as the Dunlavy SCM series, or the DCM timewindows, or something similar. Right now I'm designing a version of the timewindows using the ND120 driver and a small neo tweeter. There is a lot to be learned from designs from the 70s and 80s.

For each case I look for suitable drivers and simulate off axis in VituixCAD. My bottleneck is making cabinets, but I'm hoping to make some connections to a local makerspace to help me out there.

My spare drivers are the following:
4x rs225-8
SS discovery 15M midranges
BG Neo 8s midranges
Peerless ne123 midranges
4x SB26TCN small format tweeters
SB26STAC tweeters

[jhollander]

I use Fort Wayne and MWAF as motivation to build something new. I look for new speaker design concepts for ideas I want to try, like a pending omni speaker. I also like to pair my tweeters and mid-woofers for future builds.

I roughly sketch a concept, then check which drivers would work. I usually know which drivers will play well together from my previous pairing work, so I’m determining the actual box volume. Once I have a concept, drivers, and box size I’ll draw the speaker in AutoCAD and check the proportions.

Depending on the design, I’ll decide if I need to add a wood or finishing feature to the design. If there are color considerations, I usually pull in my wife to consult on the colors.

One interesting (to me) concept that was not built was a Linkwitz LX521 design. I designed a functional 1/3 scale LX521. I thought it was cool enough that I emailed SL to ask if it was OK (I heard he considered the shape to be IP). He said it was fine to build, but questioned why would I make something that had obvious short comings. I guess his definition of cool was not the same as mine.

[ohaple]

I always design in this order:

1. Identify the need or use case. What room will it be in? What will the audience be? What sort of content will be played?
2. Identify the budget.
3. Once those two are done, I have a good idea of the size of boxes that will work.
4. Once I know the size of the boxes, the price point, and the use case, I start researching drivers and select drivers that will for in the box, and within the budget.
5. Once I have chosen the drivers, I start on simulating and properly designing the boxes using Sketchup. One key limitation here is if I intend to use CNC or laser cutting, I need to design in such a way that the parts fit our cut bed.
6. Then I design the crossovers and determine whether active or passive crossovers are best for the project.
7. Buy the parts and build them.

I can see how those who have made many sets would start with an idea and then chase the idea, but to me speaker building is still a pragmatic thing to do. I don't want to design speakers without a use case in my life or in the life of someone I will give them to. I did that on my first design, trying to chase down a TL translam cabinet, but quickly realized it would not suit my use case and abandoned the design.

[Sydney]

Agree with the use of a CAD/drawing software ( like Sketchup ); use to use paper and took mechanical drawing in the 60s. Software adds a lot of power ( measurement calculation for one ) and in the case of Sketchup an extensive 3D library to utilize.

[Thump]

For my current and only set, which are in pieces waiting for me to finish the vapor exhaust system in the garage, the driving factor from the start was:

1. Need replacements.
2. Want them to be extraordinary.
3. Commercial equivalents to what I had in mind were in the $5k - $10k range.
4. Decided if I'm going to do something like that, then I want to literally try to do it myself as part of the investment.
5. Found the best options PE had in raw drivers - Esoteric line all in parallel x3 towers minus tweeter (RT2 instead).
6. Found the best options PE had in raw XO components.
7. Jumped into the DIY world to figure out how to go from nothing to a set that'll last us the rest of our lives.
8. ... to be continued.

I wound up visually modeling our living room, TV and other equipment I have in there. The media wall was from actual picture I took standing in the back centered which became the wall inside SU.

Went from there creating all kinds of crazy **** to see how it would look. In the end, simple, elegant towers + raised CC won the day and off on the journey I went.

Once these are phase 3 which is 100% done (which now simply means paint) the subs are next. SU will be invaluable at being able to mock up stuff based on what actually exists in there, their intended locations and most importantly if I can come up with something I'm happy with where the subs have two towers on top of them.

That last part has been harder than I anticipated - coming up with something I actually like, that I can technically achieve with the tools present. It may turn out to be simple wins again, but never know.

Playing around with what-if's is definitely part of the fun, and frustration too.

tomzarbo Hopefully us noobs were welcome to "post" in here as well. I thought it would be good to see the differences. There are likely many initially coming at it from a pure clinical necessity POV. A new set is needed and instead of yet another boring commercial product, the notion of DIY breached the barrier and created a completely new unexpected possibility.

The passion you actual seasoned experts have, and share with those of us who know next to nothing is a vital piece to helping gain new people and inject that passion into them, creating endless generations. Catching DIY fever and never stopping is extremely compelling. Experts generous with their knowledge and welcoming and friendly to new faces is pretty important to help make that happen.


For my journey, one thing that showed up immediately was the desire to achieve fabrication that precisely match intended measurements, and the frustrations that happen when things don't work out that way. It infected my entire start to finish mindset when I began the center channel without knowing it. By the end of it, the fatigue was extreme having faced that at every single piece of the process.

By the time I was finally able to start the towers, reflecting back on what had happened during CC creation made me realize I lost a lot of fun factor because of the constant focus trying to be perfect. This did allow me to loosen up and try to focus more on the fun and "act of doing" vs "medical equipment precision outcomes vs measurements".

[Brian Steele]

My builds usually happen because I want to test something out (hence almost all of them have names that start with "POC" for "Proof of Concept". The last subwoofer I built for example was to confirm that a MLTL design workbook that I put together produced accurate parameters for a Hornresp sim.

One of the few exceptions was the Blastoramas, because oh hell no was eldest daughter going to be allowed to use her dad's good Mission speakers for her planned house party. They've been used in a few house (and beach) parties since, usually accompanied by the POC3 TH subwoofer.

[a4eaudio]

1) In your pencil sketch, I wouldn't have made it past the first picture. I would have seen the picture of Ron Rico rum, thought about what cocktails could be made from rum, and would have made no further progress on DIY speakers.

2) Have you messed around with Sketchup? I usually get an idea from looking round at other speakers (DIY or commercial) and there is something that gets my attention. I have learned Sketchup at a basic level, but its nice to mock something up proportionately correct to see if the idea in my head actually has any merit. (Disclaimer: then I proceed to measure and/or cut wood wrong and screw it up somehow anyways. Like they say, measure twenty times and cut once. (I think they say measure twice, but that is not enough for me)).

I will say that with your tendencies to go with curved cabinets that Sketchup is harder, but also the reward may be greater. I don't like rectangular cabinets so messing around with curves and rough dimensions allows me to see whether my first inclinations look good or bad.

You can also figure out the volume of some odd shapes, so if you know the box volume you need but have weird curves you can get close.

[Kornbread]

Well, since money is always a concern ... my projects start with whatever closeout, or special, or 'price point' driver looks promising. Seriously.

As far as actually drawing what I'd like it to look like; I got absolutely no, zip, nadda, artistic ability.

[Wolf]

I usually start with drivers I plan on integrating, and the form follows function; sometimes snowballing into something else along the way to either learn something new, test my skill level, or just make it plainly more difficult to build.

Later,
Wolf

[augerpro]

For a few years it started with wanting to use certain "bling" drivers. Nowadays I start with "how do I want the sound to radiate?" Is it high directivity? Wide dispersion? In what pass band? Omni, dipole, monopole? Once I define that, it informs the size of drivers, the crossover topology, baffle layout, etc. When that is all specced, then I try to find drivers that can fulfill those properties. Only then will use distortion, max SPL, sensitivity, to weed through those drivers.