Seeking Advice on a Custom LXmini Speaker Project!

Hey fellow audio enthusiasts,

I hope this message finds you all well. I wanted to share an exciting project I'm planning for my office desk and seek your valuable advice and expertise. I've been inspired by the renowned LXmini speakers, and I've decided to build a variation of them tailored to my specific needs.

Here's a breakdown of my project plan:

Enclosure: I intend to 3D print the enclosure for this speaker project. This approach should give me flexibility in terms of design and customization while keeping costs reasonable. I'm looking forward to experimenting with different materials and configurations.

Woofer Driver: For the low-frequency driver, I've chosen a 6.5-inch unit from Mark Levinson. I believe it will provide a solid foundation for the speaker system, offering good bass response and overall performance.

Full-Range Driver: To handle the mid-range and high frequencies, I'm considering two options: the GRS PT6825-8 or the GRS PT2522-4. Both drivers seem promising, but I'm open to suggestions or experiences you may have had with either of them.

Amplifier Board: To power this speaker system, I've selected the TSA8804 DSP-enabled amplifier board. Its integrated digital signal processing capabilities should allow me to handle the crossover function effectively and fine-tune the sound to my liking. I'm excited about the possibilities this board offers.

Now, I come to you, the knowledgeable and passionate audio community, for your guidance and advice. Here are a few areas where I would greatly appreciate your input:

1. Enclosure Design: As I embark on 3D printing the enclosure, what factors should I consider to ensure optimal sound reproduction? Are there any specific design principles or techniques that I should keep in mind? Any recommendations on enclosure volumes or materials that have worked well for you in similar projects?

2. Full-Range Driver Selection: Between the GRS PT6825-8 and GRS PT2522-4, which driver do you think would be the better fit for this project? Are there any other alternative drivers you could suggest? I'm particularly interested in achieving a balanced and accurate sound reproduction across the frequency spectrum.

3. Amplifier and DSP Configuration: Since the TSA8804 amplifier board features built-in DSP, what are some key considerations I should keep in mind while programming the crossover function? Are there any recommended crossover frequencies or other DSP settings that I should experiment with to optimize the speaker's performance?

4. General Tips and Recommendations: Lastly, I would love to hear any additional advice, tips, or recommendations you may have based on your experience in speaker design and DIY audio projects. It could be anything from wiring considerations to potential tweaks that could enhance the overall sound quality.

I'm truly excited about this project and can't wait to dive into the world of DIY speaker building. Your collective wisdom and expertise would mean the world to me and help me make informed decisions throughout this journey.

Thank you in advance for your valuable insights and suggestions. I look forward to hearing from all of you and sharing updates as this project progresses!

Best regards,

Hi Baker, you appear to be embarking on a project similar to mine, now that I've changed direction.


1. IMHO, the enclosure is a LOT to 3d print. Not sure I see the benefit, as it will have to be substantially braced to ensure low vibration. I don't like the use of the sewer pipe either. Am planning to make 5-sided tubes or towers cut from MDF, most likely. Each board identical in shape and size, with sides that are angled inwards by 54 degrees. A couple of regular pentagon pieces with holes in the center will be bracing as well as the forms around which the boards can be strapped when glued together -- along with end caps. I am way more comfortable working with wood rather than PVC piping, so...

2. I considered magetic planars too, but note that Linkwitz calls for a combination of diffraction & absorption for the back wave, which explains his structure for the cone tweeter. (Kind of OB + aperiodic). Not sure what full dipole will do to directivity pattern at the 500 Hz LR2 crossover, which leads a lot of overlap. To achieve the same directivity for the magnetic planars, a lightweight 4 sided box with perforations in the sides, and stuffing with damping might work. Also 500Hz is pretty low for the planar 8. I guess it could work with an LR4 xover, or go higher, like maybe 1 kHz. This will change the power response, naturally.

3. I have one of those 4x100 DSP amps coming. No one around here seems to have really used it or reported on how it worked when they did use it. A backup is minidsp 2x4 hd + 4 ch amp. I have some of EricaC's 2x150W amps if this route is taken.
Hmm, where to begin..

Forget about 3D printing except for possibly the mounting of the fullrange driver to the bottom enclosure, or preferably NOT connected to the bottom tube at all and instead having a free-standing stand for the fullrange driver. In other words make a separate stand for the fullrange driver that has an extended "L" bracket printed that connects to the driver and the rest of its stand.

Bottom enclosure: 8" sonotube and if you want to you can frame-around it to achieve a rectangular shape and fill-in with sand between the rectangular box and the round sonotube.


Baffle for top of sonotube with 6.5" driver attached to it should be something more ridgid than you can 3d print (like HDF/stacked masonite board with preferably some very thin visco-damping layer between the stacked boards (see Augerpro's "Monster" thread for the preferred visco adhesive caulk)).

GRS PT6825-8, and you'll want a steep crossover around 800 Hz, do not use the smaller one, it won't go low enough in freq. at any reasonable output level as a dipole without significant distortion UNLESS it is crossed-over a bit above 2 kHz (and you can't do that with the 6.5" driver in a radial configuration where there is to much pressure loss off-axis at 90 degrees above 1 kHz).

Behind the GRS PT6825-8 all you will need is a variable density layer of foam padding about the same width and height of the GRS PT6825-8; starting about a half-inch behind that driver and extending back at least 3 inches: perhaps an inch (thick) of 25 ppi foam closest to the driver and 2 inches (thick) of 35 ppi foam behind that (for a total of 3 inches thick minimum with the two layers glued together with contact adhesive).

Note: to achieve a more point-source character at higher freq.s consider 30 ppi foam in FRONT of the driver (and actually on the driver) as RAAL does for their ribbons.

Eq. as required with your DSP.
Last edited:
Last edited: