For this build, I have a couple of woofers I will test. One is the GRS 6.5in poly cone 8ohms (85dB). The other is a Pyle 6.5in paper cone 8 ohms (92dB). I’ll have to see which is the best performance as I am looking for bass extension and sensitivitnand low distortion. As 8x 6.5in woofers have a similar area to almost an 18in woofer, I am hoping distortion can be kept low since the drivers won’t have to move as much. For midrange I have choice of Faital Pro fiberglass cone 5FE100 8ohms (90dB) or 6FE100 8ohms (91dB) pro audio midbass drivers. Both very nice sounding as midranges. For the tweeter I will use a GRS 3in planar in a waveguide bezel. It will be the dipole version naturally. I am planning on passive crossover but will be open to active with DSP.
When I did this years ago with 6x woofers and a single PRV 5MR450NDY full range, the natural roll off from the woofers and full range made a perfect 340Hz natural crossover point.
There is no DSP or EQ applied here, only level matching of the full range to meet the woofer output. I’m hoping a similar situation is repeated with 8x woofers and the 6FE100 ot 5FE100. Hopefully with some more overlap so that I can shape it more with a passive crossover.
When I did this years ago with 6x woofers and a single PRV 5MR450NDY full range, the natural roll off from the woofers and full range made a perfect 340Hz natural crossover point.
There is no DSP or EQ applied here, only level matching of the full range to meet the woofer output. I’m hoping a similar situation is repeated with 8x woofers and the 6FE100 ot 5FE100. Hopefully with some more overlap so that I can shape it more with a passive crossover.
Last edited:
...and the name is fixed: EPICThank you, Istvan. This speaker is compact but finishing it is EPIC as there is a lot of surface area front and back and unfolded. You realize how much surface area is used to increase the OB bass when having to finish both sides of the baffle and both wings.
Is it a (planned) commercial speaker or is there a slight chance for diy?
A variant will be commercial so some differences from this one but for DIY I can give basic dimensions I am using here for this build.
Baffle width is 12in, overall height is 45in. 5 degree lean back. Base is 14in deep. Woofer box is for quantity 8x 6.5in woofers made by stacking 5 sheets of 3/4in MDF. So the slot is 3 sheets wide or 2.25in wide x 5.70in tall (same as hole cutout diameter). Woofer CTC spacing on woofer box is 7.0in. There are three internal sheets that form the chamber. The two outside sheets provide a circular mounting hole.
You don’t need to make a separate removable board for the mid and tweeter but it’s handy for trying different things. The wings are 7.75in deep and held by 3/4in square ledger strips. Mid (6.5in) to tweeter CTC distance is 6.25in. Mid to woofer array mid point CTC distance is 18.25in. The base is 14in deep. Tweeter to the 5.25in mid CTC distance is 5.75in. These dimensions should let you recreate more or less what I have made here. I actually don’t have any drawn plans as much of this was made based on rough sketches and on the spot adjustments to make things fit and look good. The important ones are the baffle width and woofer chamber internal dimensions and the wing depth.
Baffle width is 12in, overall height is 45in. 5 degree lean back. Base is 14in deep. Woofer box is for quantity 8x 6.5in woofers made by stacking 5 sheets of 3/4in MDF. So the slot is 3 sheets wide or 2.25in wide x 5.70in tall (same as hole cutout diameter). Woofer CTC spacing on woofer box is 7.0in. There are three internal sheets that form the chamber. The two outside sheets provide a circular mounting hole.
You don’t need to make a separate removable board for the mid and tweeter but it’s handy for trying different things. The wings are 7.75in deep and held by 3/4in square ledger strips. Mid (6.5in) to tweeter CTC distance is 6.25in. Mid to woofer array mid point CTC distance is 18.25in. The base is 14in deep. Tweeter to the 5.25in mid CTC distance is 5.75in. These dimensions should let you recreate more or less what I have made here. I actually don’t have any drawn plans as much of this was made based on rough sketches and on the spot adjustments to make things fit and look good. The important ones are the baffle width and woofer chamber internal dimensions and the wing depth.
Last edited:
Cool build.
How did you calculate/plan the chase volume (besides cone to magnet clearance) ? Any rules of thumb ?
Are you going to mirror the 6.5" bass driver orientations between the speakers (e.g. all facing left in 1 speaker and all facing right in the other speaker), alternate them within an individual cabinet and then mirror or something else ?
Thanks for sharing your projects.
The SLOB woofer chamber geometry was originally simulated in Akabak. Things to keep in mind as it is a tradeoff. Keep chamber depth small to increase bandwidth. For this reason, an 8in woofer will not have as high upper reach. 6.5in dia is about as large as one can go to get 350Hz reach on top. Keep the chamber volume small to maximize this bass upper extension and reduce lag time of woofer cone to inject air out. The slot exit dimensions should be at least 25% to 33% of the cone area to prevent velocities from being too high and causing chuffing and compression. And not too big or you revert to nothing more than an open face driver with no slot loading. Slot loading gives a bit of SPL efficiency gain - about +2.5dB according to article by Nelson Pass. The loading also reduces membrane Xmax to allow higher SPL with less distortion. The size and geometry if not optimized, may result in a midbass peak.
I think approach it as a 2 way by integrating the mid and tweeter first. But keeping in mind the ultimate max level of the woofer at 100Hz in order to match the levels of the mod/tweet to the woofer. This is new for me too and I’ll see how I muddle through this one. The crossover between the woofer and the mid might be a simple 1st order coil on the woofer and large cap on the mid/tweet unit already with its own crossover. If my estimates are correct, the woofer should be pretty sensitive despite being an open baffle.
I have used espresso brown stain before in BB plywood and it never turned out as dark as I would like. I like this color much more in that it approximates antique mahogany better. The first coat of BLO has almost dried but still slightly oily to touch. Looks very good. Natural cloudy daylight through window used for illumination here vs evening sunset light yesterday which is warmer.
This was the XKi speaker I made stained in espresso brown and finished with BLO:
This was the XKi speaker I made stained in espresso brown and finished with BLO:
I am testing out two potential candidates for the woofers. They are the GRS 6PR-8 6.5in polycone 8ohms (85dB at 2.83v) and the Pyle PDWM6 6.5in paper cone foam surround 8ohms (92dB at 2.83v). The latter should have plenty of sensitivity but Fs is not as low (as is typical higher sensitivity for pro audio woofers). Both use identical stamped steel baskets with 4 screws so I can swap them out after I determine which one I will go with.
I will be wiring them as 4 sets in parallel pairs for 4ohm each set, then series/parallel those 4ohm pairs for a nominal 4ohms. Paralling an 8ohm should give +6dB at same 2.83v, then series paralleling those shoulld give another +6dB sensitivity boost. Open baffle will lose about 6dB of sensitivity. So overall should be +6dB or 91dB at 2.83v sensitivity at around 100Hz.
I will be wiring them as 4 sets in parallel pairs for 4ohm each set, then series/parallel those 4ohm pairs for a nominal 4ohms. Paralling an 8ohm should give +6dB at same 2.83v, then series paralleling those shoulld give another +6dB sensitivity boost. Open baffle will lose about 6dB of sensitivity. So overall should be +6dB or 91dB at 2.83v sensitivity at around 100Hz.
Installed the woofers today. 4 sets of parallel and 2x series then parallel those for 4ohms nominal. Alternating drivers in and out. All wired up and tested polarity with 9v battery. All drivers push in and squeeze air out of slot on +ve polarity.
The weight of 8x woofers is really noticeable. This is a heavy speaker now. Listing it up onto the bench to do the wiring wasn’t easy. About 52lbs now I think, before tops or wings or XO boards.
The weight of 8x woofers is really noticeable. This is a heavy speaker now. Listing it up onto the bench to do the wiring wasn’t easy. About 52lbs now I think, before tops or wings or XO boards.
I tried a 1.5v battery and could not see the deflection well enough. 9v is equivalent of 1 way stroke for 10w power into 4ohms. Not too bad. I was less concerned about the mislabeled polarity and more whether or not I wired it correctly. My estimates say that 8x 8ohm drivers wired as such should give +12dB at 2.83v? 4x parallel / 2xseries / 1x parallel wiring.
I just had time for a quick measurement. Mic is 0.5m and at mid range axis. 2.0Vrms for equivalent sensitivity at 2.83Vrms and 1m.
With the GRS woofers, the overall sensitivity in the bass at 50Hz is 87dB at 2.83V. This is not bad for HiFi open baffle speaker. The -3dB point is 43Hz and -6dB is 37Hz. The upper bandwidth is reaching 550Hz which is great as that gives me some possible FAST type crossovers. This is all natural fullrange measurement with no EQ or DSP. I think this can be made into a passive crossover quote easily. Notice that the distortion is quite low. Some of the distortion peaks are from other things behind the speaker in the room rattling during the sweep.
I will also measure the Pyle 6.5in 92dB woofers next. They should be much more sensitive but lower reach may not be as good.
With the GRS woofers, the overall sensitivity in the bass at 50Hz is 87dB at 2.83V. This is not bad for HiFi open baffle speaker. The -3dB point is 43Hz and -6dB is 37Hz. The upper bandwidth is reaching 550Hz which is great as that gives me some possible FAST type crossovers. This is all natural fullrange measurement with no EQ or DSP. I think this can be made into a passive crossover quote easily. Notice that the distortion is quite low. Some of the distortion peaks are from other things behind the speaker in the room rattling during the sweep.
I will also measure the Pyle 6.5in 92dB woofers next. They should be much more sensitive but lower reach may not be as good.
I redid the measurements carefully and found a software checkbox that was clicked and attenuated the signal of the mic. The sensitivity is closer to 89.5dB at 2.83v.
Raw responses of all 3 drivers now measured. It looks like I can put the asymmetric crossovers at 500Hz and 3.7kHz:
There is a strange thing with the absolute SPL displayed on the distortion plot. At lower frequencies, it seems to not agree with the individual plots. That’s why I will go with the more conservative value of the individual SPL plot.
All data is with wings on. With wings off, the baffle width of 12in would predict the bass to start falling off by 600Hz I think. Once I get my crossover simulated, I’ll take the wings off. I don’t intend to operate in that mode so it’s not high on my priority list vs getting a crossover developed.
Raw responses of all 3 drivers now measured. It looks like I can put the asymmetric crossovers at 500Hz and 3.7kHz:
There is a strange thing with the absolute SPL displayed on the distortion plot. At lower frequencies, it seems to not agree with the individual plots. That’s why I will go with the more conservative value of the individual SPL plot.
All data is with wings on. With wings off, the baffle width of 12in would predict the bass to start falling off by 600Hz I think. Once I get my crossover simulated, I’ll take the wings off. I don’t intend to operate in that mode so it’s not high on my priority list vs getting a crossover developed.
Last edited: