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

[freddi]

well - how good can ya do for Beta8cx ? - its not a huge vas - I think US speaker has wrong spec (?)


http://www.eminence.com/pdf/Beta_8CX.pdf

 

[xrk971]

well - how good can ya do for Beta8cx ? - its not a huge vas - I think US speaker has wrong spec (?)


http://www.eminence.com/pdf/Beta_8CX.pdf

It's not too bad - the active volume is only 14 liters and 10in x 1.0in x 15in long vent, so the cabinet is internally 15in high, x 10in wide x 7.5in deep - with 0.5in ply the external volume is 0.85 cu ft. The fb is 62Hz and f3 of 60Hz with max SPL of about 108dB at 17 volts drive.

 

[freddi]

it would be almost the same size - a bit smaller than Karlson's Eight/"K8" and about an octave better on the low end.

 

[xrk971]

it would be almost the same size - a bit smaller than Karlson's Eight/"K8" and about an octave better on the low end.

Well, that's good to know that it performs better than a scaled down K15 or a K8. The K8 probably had better power handling and higher max SPL - although, if I tuned the XKi to 100Hz, it would probably be pretty high too.

 

[BYRTT]

.....W8-1772 should work but I would caution use of any very light mass cone driver as there are substantial pressures involved here that may collapse the cone if a sufficiently high amplitude transient came through.....

Serious talk there .

Did you sweep impedance RS100P sitting in there, or is the pressure not to be seen in such plot.

 

[freddi]

most of the FR speakers I see aren't capable of folding up - they will just burn - I was surprised how loud
a Beta12cx played in an X15 cabinet copy - the big taiko transient was so powerful that I expected the cone
to break - but things were fine - it moved less than 1/4" peak to peak. Frequency response wise. one would say
there was no bass - but it caught those upper bass transients which are missed with most blh and reflex.

X - here's a pretty strong (~2% efficient) Sammi in a diy "kinda K8" - fb on K8 ran around 95Hz - K8 was a bit
larger than XRK971's proposed XKi for Beta8 cx

An externally hosted image should be here but it was not working when we last tested it.

 

[xrk971]

Serious talk there .

Did you sweep impedance RS100P sitting in there, or is the pressure not to be seen in such plot.

Yes, I sure did make a sweep with my brand new DATS woofer tester that I got for Xmas

I am just speculating about cone collapse but I do know that there is serious air volume pumping out of the narrow K-aperture.

Here is the RS100P-4 impedance measurement in the little XKi:

 

[freddi]

no discernible front chamber effect upon small signal Z sweep

 

[xrk971]

I will have to remove the front K-aperture and do a sweep to see if there is. There should be.

 

[BYRTT]

Yes, I sure did make a sweep with my brand new DATS woofer tester that I got for Xmas

I am just speculating about cone collapse but I do know that there is serious air volume pumping out of the narrow K-aperture.

Here is the RS100P-4 impedance measurement in the little XKi:

Thanks plot XKi/RS100P, at resonances area stroke looks strait and normal but from 320Hz and all up stroke has very small saw-toot form, could this be the pressure sign or inherent the driver also at free air. Example my A10.2 also have tooth but only 6 from 200-10kHz not those micro tooth all the way.

X, if preferred seems that printing the plot from DATS software can look cleaner than a program capture. Software for DATS miss the smart "Capture" bottom as REW has for creating picture files. A free alternative way is print from DATS to MS XPS file document printer, open this and capture only the plot with MS SnippingTool.exe which can ouput png or jpg files. Attached is a print example of a known Vifa.

 

[xrk971]

Ok, here is the impedance sweep with and without the K-aperture. The aperture pulls the tuning to the left a bit and slightly changes the lower freq peak value.

 

[xrk971]

Effect of EQ boost on HF's to improve "Air"

I tried increasing the boost in the upper frequencies to see if the sense of "air" or openess could be improved. I applied a +10dB high shelf starting at 12kHz with a Q=0.8. You can see the effect in the plots (dark green is the base, red is boost in 12kHz and up, blue is additional boost at 4kHz +2dB Q=0.9). The following clips represent the green (baseline case) and full EQ boost (dark blue). I also swapped out the paper driver with the aluminum cone and noticed that the aluminum cone has about 2dB more bass and the HD plot is cleaner. But all this comes at the expense of a ringing impulse response. See the IR's for the paper and alumimum cone below:

Here are the sound clips with and without EQ boost:

Without:
http://www.diyaudio.com/forums/attachments/full-range/461448d1422174486-xki-xs-ab-initio-karlson-6th-order-bandpass-xki-rs100p-4-rp-no-eq-boost.zip

With boost:
http://www.diyaudio.com/forums/attachments/full-range/461449d1422174486-xki-xs-ab-initio-karlson-6th-order-bandpass-xki-rs100p-4-rp-eq-boost.zip

Paper IR (with EQ boost):

Al IR:

Harmonic Distortion (paper):

Harmonic Distortion (aluminum):

And here is what the aluminum cone sounds like, I think there is sibillance: http://www.diyaudio.com/forums/attachments/full-range/461450d1422175098-xki-xs-ab-initio-karlson-6th-order-bandpass-xki-rs100p-4-rp-alum-eq-boost.zip

 

[freddi]

yeah - sibilant - can you apply maybe 3 thin and small strips of cellophane tape to the cone? the bass is remarkably good

 

[xrk971]

Neat idea Freddi - cone treatments - never thought of that. Pretty reversible too. I will give it a try and report back. Maybe this may be the saving trick.

Yeah, the bass is remarkable - cleaner and lower HD than the big 5in Tang Band in a sealed Hypercube.

I suspect the HD of the Al cone is better than paper when I swapped drivers I noticed two driver screws were undersized and a bit loose. I swapped for bigger ones and tightened. Must be basket or baffle vibration.

While I am at it I will apply CLD to the K aperture.

 

[xrk971]

It's not too bad - the active volume is only 14 liters and 10in x 1.0in x 15in long vent, so the cabinet is internally 15in high, x 10in wide x 7.5in deep - with 0.5in ply the external volume is 0.85 cu ft. The fb is 62Hz and f3 of 60Hz with max SPL of about 108dB at 17 volts drive.

I am going to have to revise this design based on practical consideration of how big the magnet and rear side mounted compression driver is - it won't fit in this small of a box. As a result the volume will be larger with more bass extension but probably less power handling. I have a Beta 8CX and PRV 280Ti ordered and on the way. Hopefully will be able to test this out.

 

[Greg B]

Neat idea Freddi - cone treatments - never thought of that. Pretty reversible too. I will give it a try and report back. Maybe this may be the saving trick.

The scotch tape trick works well. I prefer the red "transparent" stuff. It is actually cellophane.

BTW, for those that don't have or know akabak, here is the online calculator I mentioned that will do series BP6, series BP8, etc. "advanced speaker box calculator"

 

[xrk971]

Thanks for the link GregB. I tried the last variant for rear and front chamber with two vents from my phone and when I click calculate it thinks then returns nothing. Have you gotten it to work yet?

 

[xrk971]

Added 3 little strips of smooth scotch tape in a radial pattern on the aluminum cone, plus reduced the "air" EQ boost and it seems to have cleaned up the ringing considerably.

Here is new response and harmonic distortion:

Impulse response:

Minimum phase:

Sound clips to come but I don't detect the sibilant sounds anymore.

 

[freddi]

nice distortion figures - - re: calculator - it takes the preloaded data such as Beta8cx and does something when the "calculate" button is pushed -its difficult for me to see - I'm not sure how to work it in the simplest mode muchless other. If no port diameter is entered, does that shut off the respective port? - if you can run DOS programs and I still have them (?) there were a private versions of Bob White and Bob Bullock's BOXMODEL called DCBOXMODEL (dual chamber reflex) and BPBPBOXMODEL the latter, which IIRC could model BP6a and BP6b. It had a couple of bugs where data had to sometimes be entered twice to get the correct plots but was cool. I think it also may have handled some passive EQ cases besides active and the front chamber could be deleted to do direct radiator sealed and vented. There was a commercial program (expensive} - maybe "SpeakEasy" which did series and parallel bandpass boxes. What are the car audio guys using for BP?

 

[BYRTT]

.....Here are the sound clips with and without EQ boost.....
.....And here is what the aluminum cone sounds like, I think there is sibilance.....

Prefer the sound from the paper one without EQ verse EQed and the alu one.

Wonder if the EQ correction instead of being pure minimum or linear phase but corrected half by minimum phase and half by linear phase could do a trick in a natural way in the area where the driver is not a piston.

.....I have a Beta 8CX and PRV 280Ti ordered and on the way. Hopefully will be able to test this out.

...........

Added 3 little strips of smooth scotch tape in a radial pattern on the aluminum cone, plus reduced the "air" EQ boost and it seems to have cleaned up the ringing considerably.....

Great.