Mod SB26ADC for Lower Fs?

One solution is to add a supertweeter with a 1st order passive at 7~8 kHz

lots of combing crossing that high and slow..

comb-filter-diagramX.gif


dave
 
Hmmm... I have measured it -- not nearly as smooth as claimed, but then I measured it w/o baffle in dipole mode. IIRC, the original FR was done with drive mounted on a large baffle. I'd have to mount it on the intended baffle and use a small traffic cone to absorb as much of the back radiation as possible.

I did use it recently on another LX521-style variant, and it worked well -- crossed ~4kHz.
View attachment 1249614

But you're right -- "maybe the narrowing of the wavefront above 2.5khz or so is advantage". I could just build X's FAST, set in the intended room & listen & measure there. If it doesn't work well enough, mod or move on. That's a practical approach, I guess.
I think the build and listen in a good call. The 10f was tested for power handling by Zaph and even at 400hz LR2, it can handle gobs of power……thermal power compression is the only enemy. I’d play around a bit and see where it takes you…..you can always add a tweeter later
 
Hmmm... I have measured it -- not nearly as smooth as claimed, but then I measured it w/o baffle in dipole mode. IIRC, the original FR was done with drive mounted on a large baffle. I'd have to mount it on the intended baffle and use a small traffic cone to absorb as much of the back radiation as possible.

I did use it recently on another LX521-style variant, and it worked well -- crossed ~4kHz.
View attachment 1249614

But you're right -- "maybe the narrowing of the wavefront above 2.5khz or so is advantage". I could just build X's FAST, set in the intended room & listen & measure there. If it doesn't work well enough, mod or move on. That's a practical approach, I guess.
OH, haha! I used the B&G Neo3 with the NEW123 in a 2 way. I tried the GR planar as well and some samples were good, others not very (bumpy response or high distortion). I had to buy 4-6 to get a well matched pair with expected response. It's obvious why the B&G originals were becoming expensive - much like the original Quad ESLs they probably fail over 50% and send them back for re-skinning the film.
 
You don't even need to mod it for this. Has worked just fine in the Protones https://github.com/Ponkala/audiopoli-protone/wiki
If I planned to use DSP crossover, I wouldn't sweat 1.5 khz either, but I need to go passive. End users want to use the new Wiim Amp, ordered at my recommendation. That + Qobuz + these speakers + 65" LG OLED + streaming video will be their AV entertainment system. To replace a lousy soundbar & various google speakers. Neighbor friends who ventured into my studio to see/hear my audio projects a while ago.
 
Too easily, I dropped a butterfly accidentally while dismounting the front plate. Take care.
Agreed. The 4 screws thread in only a few turns, and there's very little friction to hold the dome/coil in place when the faceplate comes off. I used a tiny drop of crazy glue to hold it more firmly -- still pulls off easily but stays put otherwise.
Better damping of the cavity might help though.
I tried putting a small piece of foam in the cavity. FS dropped from 590 Hz to 570, impedance peak from 6 to 5 ohms. Didn't measure FR -- want rebated baffle for that.

It's tricky to experiment with damping w/o opening up the back. Removing the front & dome assembly multiple times invites an accident. If the back was open, it'd be a lot easier, though I'd then need to devise a new easily opened/closed bigger cover.

Need to invite @Patrick Bateman to post pics & details of the metamaterial & how he applied it to his sb26. He said it transformed the sound of the sb26.
 
OK, so here's where @Patrick Bateman posted STL files of the metamaterial block he made for the SB26ADC -- probably usable with any SB26:

I took that over to my local FabLab & they had a close look at the file. The device looks maybe 2" tall, like a cylindrical extension of the magnet. It has 2 layers or "floors" of a circular maze-like pattern of channels with different lengths. The anticipated weight is ~85 grams. They will have one for me next week. If that one turns out ok -- and gives me the desired results -- I'll get a second one printed.

A good xmas surprise gift. 🙂 Thanks PB!
 
So the @Patrick Bateman meta-block for the sb26adc got finished yesterday. From the backside, it looks a lot like an ice hockey puck.

1000005157.jpg


From the other side...

1000005154.jpg


There are 4 rectangular holes in the cylindrical hollow, a pair lower, a pair higher, that lead into the maze-like channels within. At the bottom of the cylindrical hollow, PB has replicated the conical deflector of the original back cover.

Overall inside diameter (inside the perimeter lip) seems a touch bigger than that of the magnet. I think all of the original plastic end cap has to be removed. Based on PB's comments and my initial poking, this is likely to be a PITA.

1000005152.jpg


Pic above shows the tweeter, the @Patrick Bateman meta-puck, and the 6" @augerpro waveguide I had made a while ago. Neither of these mods would be possible without the efforts of these outstanding DIY audio members: thank you both!

I'm hoping the combo will give me a tweeter that can be crossed as low as 1500 Hz 2nd order while maintaining excellent dispersion to 10khz & extension to 20khz. Ie, a tweeter that's near-unobtainium. 🤞🤞

I will try to accurately record & report my findings.
 
Last edited:
Oh the back chamber comes off in about twenty seconds. I used a hand saw and it went through it like butter.
I used a Japanese hand saw that I bought from Home Depot, it's one of my favorite tools.

I think that a conventional hand saw is probably too blunt/inaccurate, and a hack saw might use a blade that has teeth that are too small to really cut through it quickly. Japanese hand saw is just about perfect.

Dewalt-Pull-Saws.jpg
 
Since I was in the hacking mode with the Yamaha 3" BE dome from the NS1000 and had the @Patrick Bateman meta-puck in hand, I kept going.

I have one of those Jap saws from Lee Valley that I've never used before. Used it. Cut the back away. Stuffed a bit of sponge to approximate what KEF said they do with their meta coax. Used double-side tape to stick the puck to the opened back of the SB26.

PXL_20240104_020452215.NIGHT.jpg


PXL_20240104_021922494.jpg


PXL_20240104_023501968.jpg


No time for FR or other acoustic measurements but the DATS results are significant.

Stock SB26ADC:
sb26adc-stock.jpg


With Patrick's meta-puck absorber:
sb26adc-meta.jpg


Fs dropped from 590 to 405 Hz. No significant change in amplitude.

This SHOULD allow safe, low distortion operation to a lower frequency.

I'd love to get FR tests done quickly, esp the with @augerpro 6" waveguide I have on hand... but busy couple days and a little trip away. Might not get a chance for a week.
 
  • Like
Reactions: GreatLaBroski
I wonder what that graph would look like if the chamber was stuffed more like how scan-speak does it with (what appears to be) wool felt:
Glad you're enjoying. 😉

I tried teased long lamb hair in the cylindrical chamber of a Yamaha JA0801 3" BE dome here and it made no appreciable difference vs poly fiber. But this was with the relatively huge volume of a 12" tall traffic cone for a back. I doubt thick felt would be useful -- too much reflection & absorption before the meta-chamber, imho. Might try a little ball of lamb hair later.

This meta-puck did not do anything good on the JA0801, btw. Too small, methinks, or maybe wrong resonant frequencies. Need to get @Patrick Bateman to tell the specs of the puck.
 
@GreatLaBroski -- forgot to mention, the stock back was filled with that type of felt already. My Japanese saw split it in half when I cut the back cap.

@markbakk -- I realize the seal between the meta-absorber & the tweeter might not have been complete because the double sided tape was partly on the felt. This might be responsible for those secondary resonances. Will try sealing it tighter with tape & run DATS again. For permanent installation, will scrape the remaining felt off the magnet before using adhesive.
 
  • Like
Reactions: GreatLaBroski