• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Questionable Loudspeaker Enginerding with Aubey Industries and Weeb Labs!

airborne

Active Member
Joined
Mar 18, 2022
Messages
131
Likes
22
Again, this is simply a driver testbed assembled from scrap materials; it is not the loudspeaker design described within the OP.

The CP-09 is a relatively expensive, high distortion driver and exhibits narrower horizontal directivity than the unoptimized testbed. It is closer to 45 degrees at the -6dB point.

View attachment 193229

the way your speaker has wide directivity up to about 5 khz and then just turns into a laser beam will make it sound lifeless - lacking "air"

the strength of CP-09 is directivity is constant ...

but admittedly it can't be integrated into your design ...

relying solely on driver size for directivity control is extremely wasteful when directivity can be shaped by waveguides ...

BUT ! it's your project and you get to do it any way you like :)
 

airborne

Active Member
Joined
Mar 18, 2022
Messages
131
Likes
22
all right relax guys - i joined ASR an hour ago and just wanted to do some test posts to see if the forum is alive and if others can see and reply to my posts

test was successful - thank you !

my reason for joining is to get some criticism for my own design, that i will post later, but i wanted to test the waters first by posting some criticism of somebody else's design just to see how it goes ...

before writing a 10-page description of my own design with 20 illustrations you know ?

see you guys around !
 
Last edited:

Katji

Major Contributor
Joined
Sep 26, 2017
Messages
2,990
Likes
2,273
all right relax guys - i joined ASR an hour ago and just wanted to do some test posts to see if the forum is alive and if others can see and reply to my posts

test was successful - thank you !
Alive? heh?
...and why might it not be possible for people to see your posts and reply to them?
 
OP
Weeb Labs

Weeb Labs

Addicted to Fun and Learning
Joined
Jun 24, 2020
Messages
606
Likes
1,423
Location
Ireland
Recently, I have been working on a small desktop monitor that makes use of very inexpensive components. This monitor was designed to meet the requirements of my specific desk, room and listening position, so do bear that in mind should you find yourself interested in replicating it.

The requirements were as follows.
  • Shallow, inert cabinet which could sit flush with my monitors which matching the appearance of my ash desk.
  • Ruler flat response down to 80Hz.
  • Ideally suited to a 2.1 configuration.
  • High SPL capability with minimal distortion products above 700Hz (both harmonic and amplitude modulated)
  • Wide but not necessarily perfectly smooth directivity.
  • Very inexpensive.
With these criteria in mind, I drew simple plans for a suitable cabinet.

1664569645542.png


The total depth is a mere 100mm, which corresponds to the amount of clearance between the surface of my primary display and the rear wall.

Next, I selected the drivers and materials. The objective here was to locate drivers which would deliver the best possible measured performance at the lowest possible price while fulfilling the wide directivity requirement. Eventually, I settled upon the following.

  • 2x Dayton Audio ND13 (13mm Tweeter) - $20
  • 2x Tectonic TEBM35C10-4 (35mm Midrange) - $34
  • 2x Dayton Audio TCP-115-4 (100mm Bass Midwoofer) - $29
  • 1x MDF sheet (18mm) - $20
  • 1x Ash sheet (30mm) - $20
  • 1x Wood veneer sheet - $7
Total cost for two monitors (excluding DSP/amplification): $130

The cabinet material is MDF but I chose to make use of real 30mm ash for the front baffle. This eliminated the need for a front veneer and enabled me to created sealed rear chambers for both the midrange and tweeter without compromising cabinet volume for the bass midwoofer.

I then set about cutting and assembling the cabinets.

IMG_3259.jpg
IMG_3261.jpg
IMG_3290.jpg
1664571153533.png


With the cabinets assembled, I performed a quick test fit of the drivers and began implementing a simple DSP crossover via my JAB5s, correcting driver responses as necessary.

1664571340208.png


I wanted to cross the ND13 as low as possible, without running into excursion limits resulting from its high Fs. This will suffice for now but as we shall soon see, 4.5KHz is quite conservative with the current 24dB/oct filters.

1664571526484.png



There are still a few things to finalize (BMR mount print, 6-pin DIN, router flush trip, veneer, staining) before this project is complete but it is coming together quite nicely. On-axis is currently ruler flat to 80Hz, which aligns perfectly with my calculations for this cabinet.

unknown.png



Distortion performance is also excellent above 700Hz, approaching -60dB at 86dB SPL in the most critical portions of the band. Below 700Hz is much cleaner than this measurement indicates but there are several very loud desk and equipment resonances down there, so much of that can be disregarded.

unknown.png



I do not yet have a proper CEA2034A spin for these monitors but preliminary off-axis measurements (disregard actual frequency response) indicate very reasonable behavior up to 70 degrees off-axis.

unknown.png



That's all for now! Further updates as I complete the remaining work.
 
OP
Weeb Labs

Weeb Labs

Addicted to Fun and Learning
Joined
Jun 24, 2020
Messages
606
Likes
1,423
Location
Ireland
I now have a CNC, which will both enable me to design more sophisticated cabinets (modelled waveguides, for instance) and greatly improve precision.

748CBC0A-5502-44C0-9F41-C5524EC94A33.jpeg
 
OP
Weeb Labs

Weeb Labs

Addicted to Fun and Learning
Joined
Jun 24, 2020
Messages
606
Likes
1,423
Location
Ireland
In order to rapidly familiarise myself with the CAD/CAM process, I chose to design and machine a simple cabinet for a pair of BMR drivers. The Carveco and G-Code files will be made available on GitHub, for those interested in replicating this little project.

My initial cuts were a mess due to a random x-axis coupler failure.

1669906552184.jpeg



Once that issue was resolved, I was able to produce extremely precise cuts. A locating hole was added for the driver position, which enabled me to accurately machine both sides of the material.

1669906667389.jpeg
1669906685029.jpeg
1669906783039.jpeg
1669906835134.jpeg



After performing a test fit, it was then time to give the edges of the material some attention from a knife (ideally a router table with flush cut bit). The results were very clean and I soon had a very nice pair of cabinets to assemble.

1669907041119.jpeg
1669906989015.jpeg
1669907156442.jpeg



With both cabinets assembled, it was time to sand everything perfectly flush and apply the veneer. This entailed determining the most effective means of cutting said veneer. Ultimately, I decided that a 0.8mm endmill was much more effective than a manual knife.

1669907535932.jpeg


The end result is rather attractive, in my opinion.

1669907244467.jpeg



I do not yet have anechoic measurements or a CEA-2034A (soon) but the purpose of this project was to learn about the CAD, CAM and machining process rather than to produce a very well behaved speaker. I may port these cabinets, depending upon the measurement results.

My next project will entail machining a computed waveguide into a baffle. That should be quite an interesting one.
 

Rednaxela

Major Contributor
Joined
Mar 30, 2022
Messages
2,135
Likes
2,760
Location
NL
Amazing!

Out of interest, how did this thread’s original project turn out?
 
OP
Weeb Labs

Weeb Labs

Addicted to Fun and Learning
Joined
Jun 24, 2020
Messages
606
Likes
1,423
Location
Ireland
Amazing!

Out of interest, how did this thread’s original project turn out?
It is still under development and I expect to build one in the near future.

I chose to port the BMR cabinets and in conjunction with DSP, the result was an F3 around 50Hz and an F6 around 44Hz with very reasonable SPL capability. It bears repeating that this is a single 3.5" driver. With a slightly longer and narrower port (currently 30*190mm), I am quite confident that an F3 of 45Hz is attainable at reasonable SPL.

image.png
 

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
3,613
Likes
7,347
Location
Stow, Ohio USA
I have good news for you!

The 3rd revision of the cabinet (the final release version) that I am working on right now is definitely going to be capable of cardioid bass and maybe cardioid lower-midrange. Additionally, the upper "bookshelf" cabinet will be separate from the lower subwoofer section, so people who don't want to build a massive floor-standing speaker with 8 drivers can instead just build the upper section.

Here's a sneak peek of one of the cardioid sub prototype designs from before the cabinet was split into two separate components:

View attachment 192537

Btw, if the woofer of choice for this is still the Dayton E180HE, suggest you look elsewhere. The Directiva team considered but held off and glad we did. If your goal is low distortion, it does not fit that bill. From VoiceCoil testing...

1670253366390.png
 

Plcamp

Addicted to Fun and Learning
Joined
Jul 6, 2020
Messages
860
Likes
1,318
Location
Ottawa
Thanks for posting your work!

Looking at your criteria on first post caused me to wonder if you considered a horn solution?
 
OP
Weeb Labs

Weeb Labs

Addicted to Fun and Learning
Joined
Jun 24, 2020
Messages
606
Likes
1,423
Location
Ireland
Btw, if the woofer of choice for this is still the Dayton E180HE, suggest you look elsewhere. The Directiva team considered but held off and glad we did. If your goal is low distortion, it does not fit that bill. From VoiceCoil testing...

View attachment 247953
Distortion is rather high above 200Hz but as a 6.5" sub-bass driver, this is very well behaved.
 

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
3,613
Likes
7,347
Location
Stow, Ohio USA
Distortion is rather high above 200Hz but as a 6.5" sub-bass driver, this is very well behaved.
As a subwoofer (crossed under 100 hz), it may be fine.,,

However, that puts a lot of content load on the RS180 mid-woofer that could be better distributed in my experience. This may result in undesirable compression for the design. Otherwise, if later you find you want to crossover higher with it, then you have more distortion than other comparable drivers. Am just sharing some considerations that will hopefully help save some potential design iterations. Without knowing more about your crossover plans, it is difficult to say.

Hope it helps!:)
 

Rick Sykora

Major Contributor
Forum Donor
Joined
Jan 14, 2020
Messages
3,613
Likes
7,347
Location
Stow, Ohio USA
A bit of a retraction on my part as I later realized I was thinking RS150 rather than RS180 used here. :facepalm:

Compression likelihood is much less with the larger RS180. Looking more closely at the OP, the E180HE is being used as a subwoofer so agree it is less of a concern too. Still do not like the distortion optics, but for the price, may be hard to match.
 
Last edited:

jamescarter1982

Active Member
Joined
Dec 11, 2021
Messages
186
Likes
54
In order to rapidly familiarise myself with the CAD/CAM process, I chose to design and machine a simple cabinet for a pair of BMR drivers. The Carveco and G-Code files will be made available on GitHub, for those interested in replicating this little project.

My initial cuts were a mess due to a random x-axis coupler failure.

View attachment 247046


Once that issue was resolved, I was able to produce extremely precise cuts. A locating hole was added for the driver position, which enabled me to accurately machine both sides of the material.

View attachment 247047View attachment 247048View attachment 247049View attachment 247053


After performing a test fit, it was then time to give the edges of the material some attention from a knife (ideally a router table with flush cut bit). The results were very clean and I soon had a very nice pair of cabinets to assemble.

View attachment 247066View attachment 247065View attachment 247069


With both cabinets assembled, it was time to sand everything perfectly flush and apply the veneer. This entailed determining the most effective means of cutting said veneer. Ultimately, I decided that a 0.8mm endmill was much more effective than a manual knife.

View attachment 247080

The end result is rather attractive, in my opinion.

View attachment 247074


I do not yet have anechoic measurements or a CEA-2034A (soon) but the purpose of this project was to learn about the CAD, CAM and machining process rather than to produce a very well behaved speaker. I may port these cabinets, depending upon the measurement results.

My next project will entail machining a computed waveguide into a baffle. That should be quite an interesting one.
this looks really good ! I'd love to have the ability to be this precise and not get covered in sawdust all the time . roughly how much does one of these type of cnc cost nowadays ?
 
Top Bottom