Noob Speaker design question... Starting from scratch.

[617]

USB measurement mic is not suited to loudspeaker design measurement. There are workarounds but they are extremely cumbersome in my opinion.

 

[fpitas]

Also here:

But you want to design your own loudspeaker

My take on DIY, of which I've done a lot, although mostly in the 80's and 90's when it was easier to do as well as the best designers out there: It is worthwhile as a challenge and learning experience. It is a good solution when you want something that is not adequately addressed by the market, such as a speaker designed for a specific placement, a particular size to fit a difficult spot, or an unusual but valid design principle such as CBT, floor-to-ceiling line array, or dipole. Or floor-to-ceiling dipole line array. Also for very large speakers where much of the expense in retail is moving the thing. Normal speakers are more cost-effectively bought than built, particularly given the "known good" aspects of speakers tested here or by Erin.

My take is that, if you're not real picky about how it sounds and measures, DIY can be fine. A great sense of accomplishment. But getting good at it takes years, and acquiring knowledge most people never knew existed.

 

[Rick Sykora]

You can get a calibrated USB measurement mic for about $100 USD and REW (software) is FREE.

And... measurement isn't going to help that much unless you are going to experiment with different drivers or modifying the crossover, etc. It could help with tweaking the port but you'd need to make the measurements outdoors to eliminate room effects, and the design software should give you a pretty good idea of the bass response.

I would also generally recommend building your first speaker with more affordable components, but it's your money and you know what you can comfortably afford... And I kind-of like the idea of a coax speaker because the woofer & tweeter are matched and made to work together.

If you are just building an established design, a USB mic would help you to validate the overall speaker. However, if your goal is to build a speaker that compares favorably to a KEF speaker (as the OP indicates), you need to measure drivers installed in their intended cabinet. WinISD is good but very basic. VituixCAD is much more comprehensive software. From its manual on microphone measurements…

Note! Single channel measurement systems such as USB microphones (with latency variations by default) are not
recommended for speaker engineering due to timing and phase variations and normalizations. REW should NOT be
used without electrical loopback as timing reference or cal and timing reference for acoustical measurements to
avoid timing manipulation by the program.

For that matter, unless you want to build an impedance box, a DATS is mandatory. So, add another $120 or so to my $400 estimate.

 

[Middle Earth]

The CSS kits are designs and well reviewed
Yes, make it up and it is a crap-shoot

 

[fpitas]

The CSS kits are designs and well reviewed
Yes, make it up and it is a crap-shoot

There are some very good kits. The pitfall there is driver obsolescence.

 

[Thomas savage]

Hi,

Had some spare time over the holidays and I kinda started looking at my speakers.
Was surfing da web and landed on a hi fi site and this speaker caught my eye.

Fyne Audio F1-8 Bookshelf Loudspeaker. Ok, its 10.5K (USD) and a bit out of my price range.
But I thought about building something similar. Like a poor man's version.

I wanted to build it around Radian's 2CRF6430 12”coaxial with planar ribbon HF driver.
(The data sheet even has a crossover recommended.)

So assuming that the driver and cross over are selected... I was curious about the speaker cabinet.
The Fyne has a downward facing port that hits a cone, and I was thinking that it would be possible to print the entire cabinet using a 3D printer.
There's a lot of videos on the topic of 3D printing and I think its something I can experiment with.
Its also an excuse to get a 3D printer... or I could just bug my nephew and buy him the PLA.

My question(s)...
1) Is this insane? The drivers are ~$450 each and the crossover is ~$105 each.
2) In terms of frequency range... good enough for a bookshelf. 40Hz-30KHz.
3) Passive vs DSP. Since its a coaxial, w a recommended known crossover... why would you consider a DSP?
4) What test equipment should I buy?
5) I like the design, but would it make sense to start w a standard box?


Note: This is something I thought about doing... I figure its going to be at least a grand on the drivers/crossover / wiring, then the cost of the cabinet.
(Printing because I gave away my woodworking tools and I'd need to get a new router.)
Also, if I had to buy speakers, I'd consider the KEFs Meta R3 which is roughly 2 the price. The idea of building is part hobbyist if I have the time.
The KEF LS50Meta are roughly the same price as this build... could do a cabinet like theirs as an option. (Although the Fynes look cool)

Thx.

It is insane but fun and understanding will at a cost come from such a insanity. Its kinda insane to be interested in beyond human perception audio performance, that one could argue most of these pages are comprised of .

Hobby is fun , embrace the insane , the process will be rewarding . Assuming your not stealing the plans for the death star , so there's no harm , no dead bothans.

 

[617]

If you are just building an established design, a USB mic would help you to validate the overall speaker. However, if your goal is to build a speaker that compares favorably to a KEF speaker (as the OP indicates), you need to measure drivers installed in their intended cabinet. WinISD is good but very basic. VituixCAD is much more comprehensive software. From its manual on microphone measurements…

Note! Single channel measurement systems such as USB microphones (with latency variations by default) are not
recommended for speaker engineering due to timing and phase variations and normalizations. REW should NOT be
used without electrical loopback as timing reference or cal and timing reference for acoustical measurements to
avoid timing manipulation by the program.

For that matter, unless you want to build an impedance box, a DATS is mandatory. So, add another $120 or so to my $400 estimate.

The problem isn't the $400, it's understanding how a latency compensating loopback measurement allows you to measure relative phase.

 

[im_gumby]

Ok, call it a derivative if you prefer, most of what I posted still applies...

Certainly, part of the speaker cost is the fabrication and you can absorb that yourself. However, as a speaker is a system design, KEF (and others) apply CAD, do vibration analysis and apply materials beyond what most of us hobbyists can. This keeps them from doing expensive redesigning. From experience, I have more than a few drivers around that did not fit my design plans in the end. If you are putting in the effort, plan to buy spares too. You may blow one during design or in later use. The hobbyist is also their own service and support. I get I may sound discouraging but hope you (and others) choose to pursue the effort knowing the full set of challenges.

Exactly,
To be honest, right now its a thought exercise.
My day job takes priority and sucks up my free time. I mean, it could be months before I get to think about this again.

I was thinking coaxial speakers in that I remove some of the issues and the driver manufacturer also has a recommended crossover to pair with the driver.
This puts the bulk of the work on the cabinet / box design.

And yes, this is a hobby thing.

I mean I may end up buying different sized drivers and then building new boxes for them.

Thx everyone!