XKi - X's ab initio Karlson 6th Order Bandpass

Having spent several weeks reading this and several closely related threads on this K-Aperture design tome, it's time to ask several questions regarding the XKI designs in relation to a project we are about to undertake.

Over the past 40 years I have always had the Audio system done "My Way" and recently I have relented to the WAF to reduce the floor space and visual presence of the "Audio Altar", a phrase christened by my wife some 4 decades ago when I came home with my first pair of Maggies(II's) and placed them aside the 50" Mitsubishi front projection TV. I do admit, it did take on the appearance of the Iconostasis of a Russian Orthodox Church .

We have an existing older Home Theatre system and some significant remaining parts of a well used and thought of Stereo System that we would like to try to combine to reduce the visual and physical presence in a rather small room(13' x 15').

The HT portion will have new(upgraded) Full ranges drivers that will replace existing units plus additional units. We have auditioned several and have decided on Scan-Speak 10F4424G's for all 9 positions. The 2 existing small Subs will remain with the likelyhood of 1 additional noisemaker for newer "action" films for the grandchild and his "Games".

The existing "Music" drivers we would like to use are a pair of Heil AMT's. I believe xrk971 has had experience with these and in fact built a K-aperature Z-frame Dipole system for them in another thread. I believe he completed the build but I'm not sure he ever stated he was completely satisfied with the marriage of the two systems. That was my issue with the Heil's as I tried multiple 12" & 10" units in the original cabinets and never achieved a satisfactory hand-off between the two. Even if he was satisfied the box size of that 2 x 8" Sub would not be welcome in the new layout as there is barely enough room for the existing HT Subs. In any case, the Original Heil cabinets have long since rotted away and I am left with the AMT's and no place to put them.

What we would like to try would be to use the XKi design with the 10F4424G's for the front 3 channels of the new HT system to utilize it's expanded lower frequency response, maintain a constant "voice" thru the 80hz-~2KHz range and likely achieve better dispersion in the room. Hopefully the XKi box size can be refitted into the available space(there will be woodworking required anyway.) the rest will use they're existing enclosures and locations.

The Musical Listening System however will attempt to use the Heil's, transitioning to L & R Channels of the HT speakers as a Mid-woofer(X-ovr @ ~2k) and then passed down to the existing Subs @ ~80K. We will at that point need to add a second set of speaker feeds to the XKi's and transition the new wiring back to a new set of separate amps, X-ovr's, dsp and Dac strictly for use in the Musical listening mode, otherwise they will revert to Full-Range in HT Mode using it's own different wiring, amps, processing etc.

All this to reclaim my AMT's, save some space and atone for my decades of overzealous pursuit of something that's not likely to be achieved in my remaining years or if it is, I 'll not be able to hear it anyway as my HF hearing has really fallen off.

In reviewing the specs on the 10F's I get:
T-S Parameters
Resonance frequency [fs] 90 Hz
Mechanical Q factor [Qms] 3.2
Electrical Q factor [Qes] 0.32
Total Q factor [Qts] 0.29
Force factor [Bl] 3.9 Tm
Mechanical resistance [Rms] 0.49 kg/s
Moving mass [Mms] 2.8 g
Suspension compliance [Cms] 1.1 mm/N
Effective diaph. diameter [D] 68 mm
Effective piston area [Sd] 36 cm²
Equivalent volume [Vas] 2.0 l
Sensitivity (2.83V/1m) 89.8 dB
Ratio Bl/√Re 2.2 N/√W
Ratio fs/Qts 309 Hz
Nominal impedance [Zn] 4 J
Long-term max power 30 W
Linear excursion ± 2.6 mm
Max mech. excursion ± 7 mm

Can these be made to work?
 
Faitalpro 4Fe32

Hi again!
Well I managed to remove the vertical part of the duct and re-lined the cabinets with the wonderful carpet underlay. The boxes now appear to be behaving as a BR with useful output down to 55Hz. Sadly they still exhibit a boxiness which A) is a shame, and B) I can't stand:( Particularly noticible on percussion, piano and Roy Orbison.
If there is any further experimentation to be had the mids and upper mids are too much as if they are in need of some serious bsc, and suffering from some distortion coming through the cones, which is what I was trying to eliminate when I overdid the stuffing.
 
Founder of XSA-Labs
Joined 2012
Paid Member
Hi again!
Well I managed to remove the vertical part of the duct and re-lined the cabinets with the wonderful carpet underlay. The boxes now appear to be behaving as a BR with useful output down to 55Hz. Sadly they still exhibit a boxiness which A) is a shame, and B) I can't stand:( Particularly noticible on percussion, piano and Roy Orbison.
If there is any further experimentation to be had the mids and upper mids are too much as if they are in need of some serious bsc, and suffering from some distortion coming through the cones, which is what I was trying to eliminate when I overdid the stuffing.

Do you have any measurements? Can you show some photos of what you did? Have you tried lining the inside of the K aperture with felt? Also maybe a puff ball of stuffing in the upper front chamber. These boxes always have lots of bass and in no need of BSC. Are you sure you don't have a leak on the internal vertical divider? A pinhole leak will kill bass. With foam/felt behind area adjacent to driver and bottom of box where 180 deg turn is, line with felt.

Sorry you are having some issues here - it's unusual.

I just noticed you are using 4AFE32, you want the 4FE35 or 3FE35 higher Qts version - the lpwer Qts 32 wil have less bass.
 
Founder of XSA-Labs
Joined 2012
Paid Member
The Musical Listening System however will attempt to use the Heil's, transitioning to L & R Channels of the HT speakers as a Mid-woofer(X-ovr @ ~2k) and then passed down to the existing Subs @ ~80K. We will at that point need to add a second set of speaker feeds to the XKi's and transition the new wiring back to a new set of separate amps, X-ovr's, dsp and Dac strictly for use in the Musical listening mode, otherwise they will revert to Full-Range in HT Mode using it's own different wiring, amps, processing etc.

Sorry, I am having a hard time understanding what you want to do?

Heil AMT's as tweeters

SS 10F's as mids

and a sub below 80Hz?

where do you want to use the XKi topology? You are missing mid bass with a larger driver like a 5in or 8in. for 80Hz to 400Hz.

I would suggest this:

Heil AMT's above 6kHz

SS10F 500Hz to 6kHz (to cover the telephone band for excellent coherence)

Dual 6.5in or 8in mid bass for 80Hz to 500Hz (like SS 8in or Dayton RS180P or RS225)

Your choice of sub below 80Hz.

I would put the XKi aperture on the mid bass for the dual 6.5in or dual 8in for 80Hz to 500Hz.

The 10F should either be on a waveguide to boost the sensitivity to match the Heil (94dB) or you need to pad the Heil AMT.

Two 89dB woofers in parallel in XKi should get 95dB.
 
Sorry, I am having a hard time understanding what you want to do?

where do you want to use the XKi topology? You are missing mid bass with a larger driver like a 5in or 8in. for 80Hz to 400Hz.

I would suggest this:

SS10F 500Hz to 6kHz (to cover the telephone band for excellent coherence)

Dual 6.5in or 8in mid bass for 80Hz to 500Hz (like SS 8in or Dayton RS180P or RS225)

I would put the XKi aperture on the mid bass for the dual 6.5in or dual 8in for 80Hz to 500Hz.

The 10F should either be on a waveguide to boost the sensitivity to match the Heil (94dB) or you need to pad the Heil AMT.

Two 89dB woofers in parallel in XKi should get 95dB.

Yes, I see where you are going with this...........

I was hoping to use the XKi aperature on the 10F and see where that would take that 3" driver down to. Can I ask you to check to see what this driver or 2 of them would do in a XKi aperture, do you think it can improve below 500Hz? I was going to pad that Heil down, it is a small listening area and we don't use high volumes.

Did you move your Heil X-ovr up to the 6K range? using the 2 RS225's? I never tried using 8" before there always seemed to be a smearing or loss of articulation in the 1-2K range with the 10's & 12's.

I suppose I could replace the existing subs in their locations and in their place use the dual 8" RS225 as you did in the Z baffle. Then Build new sub cabs and place them in different locations around the room.
 
Founder of XSA-Labs
Joined 2012
Paid Member
I would not advise trying to get any bass out of the 10F - they just are not designed for that - their strength (and make no mistake, there is not another driver in th $100 price range that can match it) is the superb clear and fluid mid range/telephone band (500Hz to 6khz). If you put the 10F in an XKi, there will be too much cone movement and it will distort and you get intermodulation distortion of your nice mids. Plus, the K aperture will color the nice flat frequency response of the 10F. Finally, a pair of 3.5in drivers just won't make much bass - with exception if it is in a TL or BLH.

The Heil's are almost full range drivers as they can go down to 700Hz. I only suggested crossing at 6kHz since you wanted to use the 10F.

But if you really wanted an XKi for the 10F/8424 to reach 80Hz here you go

4.35 liter box with 33:67 volume ratio split front:rear; 4.5in internal width with 4.5in wide x 0.375in high x 6.5in long vent. Use a -12dB/oct Butterworth high pass filter at 80Hz to prevent over-excursion.

You will get something that reaches 74Hz F3 with 0.9mm excursion. With xmaz of 2.5mm reached at 8volts and 95.7dB max SPL.
 
Last edited:
broken.png
drewan
broken.png

diyAudio Member

Join Date: Mar 2015
Location: UK


The Faitalpro 4FE35 gets mentioned a few times, but I don't see if it is recommended, or the box sixel unless it is the first box size posted. Asthe 4FE35 is the only really affordable driver to play with in the UK I'd love to try this one
broken.png
29th February 2016, 01:39 PM #445
broken.png
xrk971
broken.png

diyAudio Member

http://www.diyaudio.com/forums/members/xrk971.html

Join Date: Oct 2012
Location: Virginia



As I said I have bass :) but also oodles of boxiness. Many thanks for all the time and energy you put into these projects, too bad this one didn't turn up trumps for me :(. so I'll try another:)



Can you get 4FE32? The Qts parameters are a much better match for 6th order bandpass.





As I said I have bass :) but also oodles of boxiness. Many thanks for all the time and energy you put into these projects, too bad this one didn't turn up trumps for me :(. so I'll try another:)
 
Last edited:
I would not advise trying to get any bass out of the 10F - they just are not designed for that - their strength (and make no mistake, there is not another driver in th $100 price range that can match it) is the superb clear and fluid mid range/telephone band (500Hz to 6khz).

Thank You xrk971, you've made your case. We choose the 10F4424's because of their strengths in the "voicing" range and don't wish to detract from it, we'll consider that issue resolved so I can at least begin the HT reconfig.

....but what todo with the AMT's..... it's really more of a floor space issue . A nice tall, thin set of say 6-6.5" stacked woofers in a KaZba? or V-frame?(Sl's521sub turned on it's side.)
 
that's not much different than a single 12" Karlsonator - I know its work to simulate but might be interesting to see what a single and double would do. Would f3 for the 125 liter double case be ~60Hz? Would 2.83v sensitivity for the double woofer box be above 100dB? There's not a lot of xmax but think its pretty linear within that figure. I get about an 80Hz corner with one pym1298 in a regular K12 tuned in the low 60s. A K-tube on top with compression driver will easily match one pym1298/kappa12a

I had a go at the single.
37 cm wide
50L rear chamber
12.5L front chamber
duct 25.4 cm long by 5 cm high (taking into account 2.6 cm end effect at rear and acoustic mass at front, although will likely still tune a bit lower than the sim)

fljxhl.png


Hi again!
Well I managed to remove the vertical part of the duct and re-lined the cabinets with the wonderful carpet underlay. The boxes now appear to be behaving as a BR with useful output down to 55Hz. Sadly they still exhibit a boxiness which A) is a shame, and B) I can't stand:( Particularly noticible on percussion, piano and Roy Orbison.
If there is any further experimentation to be had the mids and upper mids are too much as if they are in need of some serious bsc, and suffering from some distortion coming through the cones, which is what I was trying to eliminate when I overdid the stuffing.

I wonder if some melamine foam directly behind the drivers would reduce the highs coming through the cones? From reading, wool has a relatively flat attenuation curve compared with small diameter fibres such as glass that attenuate more strongly in the mid-range. Maybe a small amount of glassfibre would help without killing the bass?

I have some time this weekend to tinker with the damping. A lot of the boxiness in mine seems to come from the front chamber and some wool felt on the sides has helped.

Got them sanded and varnished.

2qdo4tk.jpg
 
If a single pym1298 can really come close to that prediction then it could be an excellent speaker running a K-tube on top - I've not seen K12 go much below 80 with high sensitivity twelves.

I'm hoping the big clunker coax below has XKi potential - its not as hot as my 1114 edgewound aluminum coil Martin/Eminence 12cx nor a B&C 12pe32, but sensitivity is decent, thermal power handling with a 4" coil and 4.8mm overhang should allow pretty good input power, and moving mass isn't terrible considering the 4 inch coil.

I've a number of 12cx including a light duty Beyma - Eminence seem pretty close to spec or better - 1114 with no break-in has qts ~0.27 - my Beyma is off with higher qts than spec but its woofer extends to past 5KHz on-axis


Def_Driver TSG big 12cx |Notes: TSG 12-1844 Eminence 12CX 109oz magnet - Qts 0.27 -4.8mm overhang 4" coil - 2x5" HF horn

SD = 532.36 Cm2
Fs = 44.41 Hz
Mms = 57.4 g
Qms = 9.95
Qes = .27
Re = 5.52 ohms
BL = 17.96 Tm
Le = 1.68 mH
Vas = 86.6L
 
Founder of XSA-Labs
Joined 2012
Paid Member
Def_Driver TSG big 12cx |Notes: TSG 12-1844 Eminence 12CX 109oz magnet - Qts 0.27 -4.8mm overhang 4" coil - 2x5" HF horn

SD = 532.36 Cm2
Fs = 44.41 Hz
Mms = 57.4 g
Qms = 9.95
Qes = .27
Re = 5.52 ohms
BL = 17.96 Tm
Le = 1.68 mH
Vas = 86.6L

Freddi,
Thanks for putting TS parameters in Akabak ready format - makes things much easier for me.

I think this can work in a rather large XKi: 75 liter total volume, 33:67 volume split front:rear, 14in wide x 1.25in high x 11in long vent. You can get 95dB sensitivity and F3 of 42Hz.

541899d1459837433-xki-xs-ab-initio-karlson-6th-order-bandpass-xki-tsg12cx.png
 

Attachments

  • XKi-TSG12cx.PNG
    XKi-TSG12cx.PNG
    21.4 KB · Views: 906
tough decision - which XKi to pick? Beta10cx isn't particularly sensitive but can move some air and a reasonable value

(I think the double PA130 XKi may be the sonic and value pick of the litter but I've already P-Audio 10, and other Cx)

simulations:
Beta10cx in GregB's full size Karlsonator (12 inch model)
E27Qoyo.jpg


Beta10cx in a 60 liter XKi at 18.5 volts

461370d1422138107-xki-xs-ab-initio-karlson-6th-order-bandpass-xki-b10cx-freq-xmax.png


Beta10cx's woofer in-room relative spl vs Visaton BG20 with each in a slit vent K12
vDrgFR6.gif


P-Audio SN10C in a 34 liter XKi alignment
O2r16R9.jpg

SredniVashtar 's sim (thanks) pym1298 in a 62.5 liter XKi
i28PM1X.png
 
xrk971 and SredniVashtar

how do you properly set the gap from the vent's exit to the back of the wings/aperture plate so its not too tight/loose?

60 liter XKi for W8-1772 and Beta 10cx
yKHZPMq.jpg


how about the gap for this 34 liter XKi for SN10C which has a 1 inch tal x 12 inch wide vent?
GqKP0ep.png


and the single pym1298 XKi ? 37 cm wide
50L rear chamber
12.5L front chamber
duct 25.4 cm long by 5 cm high (taking into account 2.6 cm end effect at rear and acoustic mass at front, although will likely still tune a bit lower than the sim)
 
Last edited:
Founder of XSA-Labs
Joined 2012
Paid Member
Bolexsbm has made a 4 x 3in driver Karlsonator I think. I have sim'd a quantity 9 x 1ohm Bose 901 drivers. It gets pretty impressive sensitivity to do say 9 x 91dB 3FE22's in series parallel. It should work well. The K aperture will clean up any comb filtering issues you may have. Someone should try it and do a measurement.

Good candidates are moderate Qts drivers so box won't be too big.