• 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!

Introducing Directiva - An ASR open source platform speaker project

muad

Senior Member
Joined
Jun 8, 2019
Messages
415
Likes
471
As promised, here is an r2 update...

First of all, let me say that most of the work was done by @TimVG and @ctrl. Other team members helped with driver selection and my main role was to rebalance requirements as needed. I anticipate they may add some commentary to this post, but both are very busy with other projects rn. So the first target was to design a small monitor with better directivity than r1 and at a lower cost. Am pleased to say, we meet this goal with one noteworthy exception. We had hoped that we could get enough bass extension to use with a subwoofer, but fell back to needing to pair with the planned bass module. We plan to revisit a derivative monitor design with lower bass extension later.

I had mentioned that simulations indicated a need to narrow the baffle to get better directivity, so we went with a simple shape that is like a small Salk Veracity. To push constant directivity even lower in frequency, we decided to use the approach used by the Dutch & Dutch 8C. @TimVG had some experience with the approach and @ctrl did several sim iterations to refine. Once this was done, Tim had gotten a CNC and decided he wanted to deploy it to build the prototype. His design takes the original shape and rounds off the top. So, it is reminiscent of a mantel clock as show here...

index.php


The tweeter as expected is the Seas DXT and the woofer is a SB15NBAC30-8 midwoofer. The SB was chosen as a balance of size and price performance. Here is a look at the normalized horizontal directivity:

as
index.php

As you can see, the prototype exhibits fairly constant directivity well below 1 kHz.

And here is a preview of the frequency response with an initial crossover:

index.php


While some aspects may still change, will know better once I build the less rounded cabinet and we test it. Like my r1 measurement system upgrade, Tim is in the process of upgrading his. I will start as soon as I wrap up the passive crossover for r1 (hopefully this week).

Cheers!

Rick and the Directiva r2 team.
The pictures aren't showing up for me on mobile.
 

NYfan2

Active Member
Forum Donor
Joined
Jan 10, 2021
Messages
209
Likes
446
Location
Netherlands
Same here, no pictures (PC with Firefox)
 

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,423
Likes
7,940
Location
Brussels, Belgium
How about now?

So as Rick already mentioned. This is a prototype - so is the crossover which I implemented in maybe 10 or 15 minutes. Please view this as the top module (2-way) of a larger system (3-way).


C1xhKE1.jpg
AtTIw8O.png

Zh9fbZD.jpg
Phenomenal work! Is there a reason why Rick was under the impression that the cabinet might change?
 

TimVG

Major Contributor
Joined
Sep 16, 2019
Messages
1,181
Likes
2,573
Phenomenal work! Is there a reason why Rick was under the impression that the cabinet might change?

Rick's version will be somewhat simpler to build, yet should be similar in performance. I made this version in layered mdf, but it does require templates and a copy router bit (and for ease of workflow, a small router table).

20211228_121404.jpg
 

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,423
Likes
7,940
Location
Brussels, Belgium
Rick's version will be somewhat simpler to build, yet should be similar in performance. I made this version in layered mdf, but it does require templates and a copy router bit (and for ease of workflow, a small router table).

View attachment 177920
I was wondering if you could build cabinets for fellow members living in Belgium :p. Not for free of course.

I can do Ground plane measurements on the weekend too as an extra opinion / data point.
 

TimVG

Major Contributor
Joined
Sep 16, 2019
Messages
1,181
Likes
2,573
I was wondering if you could build cabinets for fellow members living in Belgium :p. Not for free of course.

I can do Ground plane measurements on the weekend too as an extra opinion / data point.

Oh I have space for groud plane measurements too - the weather is just a bit off at the moment though :)
I can make templates, do driver cutouts - perhaps full cabinets over summer when I'm not too busy - but I'm not a professional woodworker, so keep that in mind :)
 

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,423
Likes
7,940
Location
Brussels, Belgium
Oh I have space for groud plane measurements too - the weather is just a bit off at the moment though :)
I can make templates, do driver cutouts - perhaps full cabinets over summer when I'm not too busy - but I'm not a professional woodworker, so keep that in mind :)
Cool!

What is the roll-off behavior at the moment btw? From the screenshot looks like the F10 is around 100Hz.
 

TimVG

Major Contributor
Joined
Sep 16, 2019
Messages
1,181
Likes
2,573
Cool!

What is the roll-off behavior at the moment btw? From the screenshot looks like the F10 is around 100Hz.

That's about correct. The idea is to implement the top module somewhere between 200 and 300Hz - and let the bass module take over from there. Look at it as the 2-way part of a 3-way system.

This also means that under normal circumstances, no further EQ would be needed on the top module when placing it in a room.

Edit: as Rick mentioned, a ported variation is something that we're looking at - for those that don't desire the extended directivity and wish to use it as a sub/sat system.
 
Joined
Feb 23, 2021
Messages
36
Likes
179
How about now?

So as Rick already mentioned. This is a prototype - so is the crossover which I implemented in maybe 10 or 15 minutes. Please view this as the top module (2-way) of a larger system (3-way).


C1xhKE1.jpg
AtTIw8O.png

Hi,
Does the above directivity plot really show vertical directivity? (or is it horizontal)

Thanks
 

TimVG

Major Contributor
Joined
Sep 16, 2019
Messages
1,181
Likes
2,573
Hi,
Does the above directivity plot really show vertical directivity? (or is it horizontal)

Thanks

This is the normalised horizontal plot. We have full plots but since the crossover used here was a very quick experimental one, there is nothing yet to share in this regard. Since this is not a concentric design, expect some compromise in terms of vertical response - as is it with all non-concentric designs, there is no getting around physics.
 
Joined
Feb 23, 2021
Messages
36
Likes
179
This is the normalised horizontal plot. We have full plots but since the crossover used here was a very quick experimental one, there is nothing yet to share in this regard. Since this is not a concentric design, expect some compromise in terms of vertical response - as is it with all non-concentric designs, there is no getting around physics.
Thank you.
I am really excited about this project. :)
It is really heartening to see these kind of developments and collaborations pushing boundaries in the DIY space.
Waiting eagerly for next set of results.
 
Last edited:

Brsvde

Member
Joined
Nov 1, 2021
Messages
8
Likes
8
Amazing work! I think @ppataki would approve of the Raptor-coating. I am definitely looking forward to the progress and am very impressed so far!
 

Savi

Member
Joined
Nov 15, 2020
Messages
69
Likes
51
Nice work ! Do you use a resistive port like d&d 8c to achieve a cardiod patern ?
 

TimVG

Major Contributor
Joined
Sep 16, 2019
Messages
1,181
Likes
2,573
Nice work ! Do you use a resistive port like d&d 8c to achieve a cardiod patern ?

Yes, I had some understanding of this from a previous project. After looking into patents and AES papers, modifying and existing (cheap) speaker, ctrl was able to work on a simulation model, and measured results correlate very well.
 

gab

Member
Joined
Mar 20, 2018
Messages
45
Likes
40
Rick's version will be somewhat simpler to build, yet should be similar in performance. I made this version in layered mdf, but it does require templates and a copy router bit (and for ease of workflow, a small router table).

View attachment 177920
Tim - what thickness MDF layers are you showing here? Looks like maybe ~0.75" or 18mm with the front panel doubled up so maybe overall cabinet depth ~ 7.5" give or take.
 
Last edited:

TimVG

Major Contributor
Joined
Sep 16, 2019
Messages
1,181
Likes
2,573
Rick - what thickness MDF layers are you showing here? Looks like maybe ~0.75" or 18mm with the front panel doubled up so maybe overall cabinet depth ~ 7.5" give or take.

12mm layered with a 25mm baffle, although internal thickness throughout is indeed 18mm except for the back panel which is 12mm, although it is dampened with bitumen.
 

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,423
Likes
7,940
Location
Brussels, Belgium
12mm layered with a 25mm baffle, although internal thickness throughout is indeed 18mm except for the back panel which is 12mm, although it is dampened with bitumen.
It would be really great if you could share the templates. I'm seriously considering joining as a 'testing' perspective.
 
Top Bottom