ET Build: Ripole sub & SS 10F/8424

[zmyrna]

ripole design reduce fs and efficiency of drivers
SLS12 already has low fs and efficiency
most diy'ers use 15 or 18" proaudio drivers in their successful ripole builts
i would do an H or M frame if I were to use SLS12 dipole
and ripole past 80hz is definitely a NO NO because of cavity resonances
you have the SLS12; build and see yourself

 

[Calvin]

Hi,

Fluid, sorry ,but no offense was intended .... just a bit of teasing
For displacement volume You calculate the lnear stroke times effective membrane area.
The Seas driver You mention has a linear stroke of 28mm p-p.
So indeed the displacement volume is greater than that of the larger SLS12.
Regarding the SPLmax chart You need to add the -6dB drop of dipoles towards sinking frequencies .... hence the dipole figures wpuld be lower.
But we didn't talk about SPLmax but of the amplitude response.
That is greatly dependant on the drivers Qtb.
The higher the Qtb the less bass lift is required to linearize the response.
The higher Qts SLS12 since requires less bass lift than the low Qts L26RO4Y.
Under certain conditions the natural response may suffice the taste of the user or requires just a couple of dBs.
I prefer to let the driver do what it 'wants to do', instead of brute power forcing it.
Besides the cost of the SLS12 is considerably lower than the Seas's.

jau
Calvin

 

[Mayuri]

So judging from the graph and dipole roll-off I should still be able to get over 100dB at 30Hz from 4x SLS 12? Seems more than reasonable, as I live in an apartment building where very high listening levels would surely bother other residents.

Judging from the simulations, it seems cavity resonance with Ripole starts around 200Hz? But I will build, see and measure myself

 

[fluid]

So judging from the graph and dipole roll-off I should still be able to get over 100dB at 30Hz from 4x SLS 12? Seems more than reasonable, as I live in an apartment building where very high listening levels would surely bother other residents.

Use SL's spl_max1 spreadsheet if you want to get a better idea

A dipole has only 1/4 as much output at 30Hz compared to a monopole of the same driver.

This is a screenshot of the spreadsheet with input for 4 SLS12 drivers (the path length difference is a guess so that could have some influence) more like 87dB but still probably high enough for your purposes

 

[Mayuri]

I see. I am completely satisfied with over 90dB at 35Hz, I can cut the very low frequencies out with MiniDSP to protect the drivers.

Another thing: How would I get around breaking the internal spherical shape of the bowls to minimize the internal modes? The internal dimensions are around 18cm width on all except the driver side, so it isn't a perfect circle.

 

[Calvin]

Hi,

How would I get around breaking the internal spherical shape of the bowls...

.... by not using a sphere
That´s why I used the slightly elliptical shaped flower pot .... I thought it´d be a better shape than a sphere ..... acoustically as well as optically.

jauu
Calvin

 

[fluid]

Another thing: How would I get around breaking the internal spherical shape of the bowls to minimize the internal modes? The internal dimensions are around 18cm width on all except the driver side, so it isn't a perfect circle.

If you line it with wool felt and stuff it with fibreglass that will do a good job of reducing contributions from the cabinet.

I did some experiments with damping materials and acoustic foam works in a similar way to felt but is more bulky and less easy to fit into a sphere. Polyester damping is nowhere near as effective as fibreglass particularly at lower frequencies.

 

[bushmeister]

Another thing: How would I get around breaking the internal spherical shape of the bowls to minimize the internal modes? The internal dimensions are around 18cm width on all except the driver side, so it isn't a perfect circle.

It is quite simple to completely eliminate the spherical predisposition to a single large modal node - first it is never as bad as it first seems as the driver is internally set back into the sphere, and therefore the magnet assembly etc breaks up some of the problems just by being in there.

As Fluid has suggested - wool felt lining followed by fibreglass wool really absorbs a huge amount if done correctly.

Lastly, a few well chosen internal braces can easily subdivide the internal sphere into two different sized hemispheres, or even more. I find a brace running back from the driver - with magnet bracing, slightly offset to one side (so it doesn't block any driver venting and creates two asymmetrical hemispheres) removes any issues - often before any damping is needed.

I have many spherical builds under my belt and this is an over-hyped issue which can be dealt with fairly easy with common sense techniques.

 

[Mayuri]

Good to know. I have some recycled felt and fiberglass, so I think I'll use those for the damping. The internal structure will have some obstacles inside, such as different braces etc, so it will probably be fine.

The construction is progressing nicely, I have holiday at the moment so got all day to tinker and build. In the picture there are legs for the amplifier and subwoofers along with IKEA cutting boards for speaker stands with the bowls themlselves. I'll use a sandwhich structure for the stand with a 10mm steel plate between two cutting boards. The plate weights 10kg.

 

[Mayuri]

The hemispheres with bracing and some recycled felt for damping.

 

[Mayuri]

Gluing time!

On the other hand, my fancy Sorbothane isolation didn't work out for a surprising reason. The magnet of the 10F is so strong that every time I test mounted it, the driver either pulled the mounting washers off-axis or otherwise made it very difficult to correctly align the long bolts. In the end I decided that I can't reliably mount the driver with this setup, so I plugged the mounting holes with wood plugs and will use normal screws for mounting. A setback, but hopefully I will still manage to get listenable audio out of them

 

[Face]

Cute little death stars.

Sent from my SM-G935V using Tapatalk

 

[Mayuri]

That reminds me, got to listen to some John Williams music. That man is a soundtrack genius.

Todays progress picture are the stands getting there. 5 clamps is not nearly enough, but that's all I own for now so got to stick with that. I finished one stand yesterday and am very happy with it, very solid and sturdy. You can really feel the weight of it

 

[Mayuri]

First cutting board all glued up and dried. Next step is sanding and lacquering. I'll use the same 10mm rubber feet at the bottom of the board as I did with the Dayton Mini Monitors, they are handy little buggers.

Currently project is going like this:
-Spheres and the leg they stand on have one coat of lacquer. They require light sanding and at least 1 more layer of lacquer before I am satisfied with them.
-Both bases will be dry this evening, and I'll sand & lacquer them so they can dry overnight.

-Amplifier parts are all here. I ordered a 10pcs of the TDA8932 modules to be sure I get the TDA8932BT version of the chip which seems to be the best. I was in luck, I got 6 of those, the rest were the other variants. With the amplifier I'll have AC filter I pulled from old subwoofer module providing a bit cleaner AC for the Meanwell 24V 10A DC module and 2 extra power sockets that will house the MiniDSP and NAD 912 amplifiers.

-The boards for amplifier and woofers are cut and ready. The shop I ordered them from botched some of the measurements on the woofer bases a bit, so they need sanding before I start assembly

 

[Mayuri]

My drivers for this project (+ extra SLS 12 if I bust one)! The ScanSpeaks look tiny.

 

[Mayuri]

I'm going on a holiday for 3 weeks, so project will continue in the last week of July. Some teaser pics of the amplifiers and speaker before I go. Thanks for the interest and support

 

[zas]

Hi,

it happens i'm using 4xSLS 12 (830669; recent ones with fs ~31hz), and a SS 10F/8414G10 in my own setup inspired by Hestia V project and my previous open baffle project.

Each SLS is in a H frame (~33x33x33cm, speaker at ~12cm from front), i added a bit of acoustic wool in the rear cavity (so it isn't exactly a dipole), 2 "boxes" per side, horizontally placed. With proper equalization i get the response slightly under 30hz at 0db, but that's something around 16db to get sufficient bass at listening position in my room, and it needs a high Q notch at fs. Speakers on each side are connected in parallel to one amplifier channel, so that's ~4 ohms.

The 10F is used as a high med, from ~1khz, while low med is an Eminence B102. The 10F in dipole cannot do the low med, and after various attempts, i ended with a 8'' speaker as being optimal.

The 10F is an impressive speaker, it works almost like a (good) tweeter, going to 20khz, with an astonishgly good horizontal dispersion, and very low distortion.
Though i added an AMT HR3326A as super tweeter that give a bit more "air" in the highs (high pass, 15khz@6db/oct-35khz).

Basically the whole thing is very wide band, low distortion, i use the minisharc to equalize phase and amplitude, it feeds 2 stereo dacs (curryman), and those feed 2 FirstOne stereo amps (160w/8ohms, 250w/4ohms), bass = active filtering, med/high = hybrid active high pass around 200hz, very simple passive between B102/10F/HR3326A. That's almost 2 years i'm tuning it, and it measures very well, as it is a pleasure to listen (whatever i listen, from chamber music to heavy doom metal).

I love this 10F, which is not expensive, very low distorsion, and really full range from 200hz.

I didn't try the 10F in a box, and i have doubt you can easily match 2 big SLS 12 with one single 10F, but your project is interesting.

 

[xrk971]

Hi,

it happens i'm using 4xSLS 12 (830669; recent ones with fs ~31hz), and a SS 10F/8414G10 in my own setup inspired by Hestia V project and my previous open baffle project.

Each SLS is in a H frame (~33x33x33cm, speaker at ~12cm from front), i added a bit of acoustic wool in the rear cavity (so it isn't exactly a dipole), 2 "boxes" per side, horizontally placed. With proper equalization i get the response slightly under 30hz at 0db, but that's something around 16db to get sufficient bass at listening position in my room, and it needs a high Q notch at fs. Speakers on each side are connected in parallel to one amplifier channel, so that's ~4 ohms.

The 10F is used as a high med, from ~1khz, while low med is an Eminence B102. The 10F in dipole cannot do the low med, and after various attempts, i ended with a 8'' speaker as being optimal.

The 10F is an impressive speaker, it works almost like a (good) tweeter, going to 20khz, with an astonishgly good horizontal dispersion, and very low distortion.
Though i added an AMT HR3326A as super tweeter that give a bit more "air" in the highs (high pass, 15khz@6db/oct-35khz).

Basically the whole thing is very wide band, low distortion, i use the minisharc to equalize phase and amplitude, it feeds 2 stereo dacs (curryman), and those feed 2 FirstOne stereo amps (160w/8ohms, 250w/4ohms), bass = active filtering, med/high = hybrid active high pass around 200hz, very simple passive between B102/10F/HR3326A. That's almost 2 years i'm tuning it, and it measures very well, as it is a pleasure to listen (whatever i listen, from chamber music to heavy doom metal).

I love this 10F, which is not expensive, very low distorsion, and really full range from 200hz.

I didn't try the 10F in a box, and i have doubt you can easily match 2 big SLS 12 with one single 10F, but your project is interesting.

Nice system Zas. Do you have photos? I am sure it sounds phenomenal. Are you using FIR and convolution methods or straight IIR filters? Have you tried transient perfect filters?

 

[Mayuri]

Hi,

it happens i'm using 4xSLS 12 (830669; recent ones with fs ~31hz), and a SS 10F/8414G10 in my own setup inspired by Hestia V project and my previous open baffle project.

Each SLS is in a H frame (~33x33x33cm, speaker at ~12cm from front), i added a bit of acoustic wool in the rear cavity (so it isn't exactly a dipole), 2 "boxes" per side, horizontally placed. With proper equalization i get the response slightly under 30hz at 0db, but that's something around 16db to get sufficient bass at listening position in my room, and it needs a high Q notch at fs. Speakers on each side are connected in parallel to one amplifier channel, so that's ~4 ohms.

The 10F is used as a high med, from ~1khz, while low med is an Eminence B102. The 10F in dipole cannot do the low med, and after various attempts, i ended with a 8'' speaker as being optimal.

The 10F is an impressive speaker, it works almost like a (good) tweeter, going to 20khz, with an astonishgly good horizontal dispersion, and very low distortion.
Though i added an AMT HR3326A as super tweeter that give a bit more "air" in the highs (high pass, 15khz@6db/oct-35khz).

Basically the whole thing is very wide band, low distortion, i use the minisharc to equalize phase and amplitude, it feeds 2 stereo dacs (curryman), and those feed 2 FirstOne stereo amps (160w/8ohms, 250w/4ohms), bass = active filtering, med/high = hybrid active high pass around 200hz, very simple passive between B102/10F/HR3326A. That's almost 2 years i'm tuning it, and it measures very well, as it is a pleasure to listen (whatever i listen, from chamber music to heavy doom metal).

I love this 10F, which is not expensive, very low distorsion, and really full range from 200hz.

I didn't try the 10F in a box, and i have doubt you can easily match 2 big SLS 12 with one single 10F, but your project is interesting.

Sounds very good, I'd also like to see how it looks

I thought I was done with building for now before the holiday, but the building bug bit again and I finished the woofer amplifier yesterday. Because hey, why not stay up until 3 am and solder

Here is my TDA8932-based amplifier made for driving 4x woofers from two channels. So basically it amplifies a stereo signal to 4 channels. The amplifiers are driven by a Meanwell 24V 10A power supply, and are separated from each other by a diode in each power lane. Extra features are a oversized cooling made from recycled computer CPU coolers and a handy slot in the amplifier enclosure to fit all your wires in (wife really appreciates this). Not shown in the picture, but there are 3 100mm holes in the slot protected by computer finger guards for ventilation. There is also an extension cord functionality so you can drive all the connected devices (NAD 912 amplifier for full range drivers and MiniDSP). The power goes through an AC filter (coil and cap) salvaged from an old subwoofer module. Everything is under the one mains switch, so you can shut down power to everything else from this amplifier.

The enclosure is constructed from 21mm and 4mm BB plywood and lacquered several times. The feet are from IKEA

 

[fluid]

Mayuri nice looking cabinet, if you haven't already you might want to include some ventilation under the heatsinks to aid in convection, those heatsinks might get pretty hot.