In the Monitor 4 project this woofer was used in a sealed box with a large volume to have a Qtc near to 0.6 in order to obtain the best transient response.

Here I am making an attempt to improve the performance even more by using an H-frame then a dipole configuration.

The H-frame design has advantages over the classic flat dipole because it is more compact under the same conditions and high frequency peaks are generated outside the frequency band used but in any case a high slope pass low filter low should be used.

Here there is an article of Martin J. King about a comparison of the Bass Performance of Open Baffle, U Frame, and H Frame Speakers.

To compensate the normal loss of efficiency of the dipole it is necessary to use a filter, in this case an active filter is placed before a dedicated amplifier.

This active filter increases about 12-15dB at 30Hz therefore it is necessary to use a high power amplifiers and only woofers with large excursions and high power are able to support this conditions.

This project born to be combined with the new system with the Purifi mid-woofer and the Mundorf AMT but it can be used in any system with mid-woofer able to produce 80Hz.

The first my test of this H-frame subwoofer was on a little 10" Wavecor woofer but this giva a more defined and drier sound, even the mid-frequency range seems cleaner.



BOX DESIGN

The box was built using 25mm MDF.




WOOFER

I have used an old K.C.S. 18" S-1846 woofer.






SIMULATION

This H-frame has been simulated using the Matcad worksheet by Martin King on a Windows XP virtual machine.
Here has been used the data of the Faital 18FH510 woofer.

MEASUREMENTS

Follows the frequency response of the loudspeaker without the active filter, this is like the simulation model.

You can see the typical resonance peak at 200Hz determined by the length of the case and this will be this will be reduced significantly with a high slope filter.

Follows the various frequency response using the active filter with the dipole compensation.

 

ACTIVE FILTER

I have got on Ebay an active crossover module to start like base (NE5532 Subwoofer processing circuit).

The original schematic have a big problem, the output voltage change with the frequency setting.

so I have changed some components and the schematic to obtain a good low-pass and to add a dipole compensation.

 

The main problem of this previous circuit is the 50Kohm potentiometer for frequency change which is not linear but logarithmic so it is not easy find the right position.  

 

Follows the new filter with a true variable 24dB/oct + dipole compensation

Here pcb files

n.	C1 (nF)	C2 = 2xC1	R1 and R2	C1 (F)	C2 (F)	F (Hz)
1	150	300	3570	0,000000150	0,0000003	210,0
2	150	300	4120	0,000000150	0,0000003	182,0
3	150	300	4700	0,000000150	0,0000003	159,5
4	150	300	5490	0,000000150	0,0000003	136,6
5	150	300	6200	0,000000150	0,0000003	120,9
6	150	300	7150	0,000000150	0,0000003	104,9
7	150	300	8250	0,000000150	0,0000003	90,9
8	150	300	9530	0,000000150	0,0000003	78,7

R1                 1Kohm  1/4W  1%
R2               100Kohm  1/4W  1%
R3,R11,R1,R27    3570ohm  1/4W  1%
R4,R12,R20,R28   4120ohm  1/4W  1%
R5,R13,R21,R29   4700ohm  1/4W  1%
R6,R14,R22,R30   5490ohm  1/4W  1%
R7,R15,R23,R31   6200ohm  1/4W  1%
R8,R16,R24,R32   7150ohm  1/4W  1%
R9,R17,R25,R33   8250ohm  1/4W  1%
R10,R18,R26,R34  9530ohm  1/4W  1%
R40,R41           270ohm  1/4W  1%
R38,R39          3000ohm  1/4W  1%
R36               56Kohm  1/4W  1%
R35                3Kohm  1/4W  1%
R37                 0ohm

C1,C2,C3.C4,C11,C12   0.15uF WIMA MKP
C5,C8                 0.22uF WIMA MKP
C6,C7              470uF 35V
C9,C10             100uF 25V

IC1,IC2             NE5532
U$1                 LM317
U$2                 LM337

B1                  diode bridge 1A

SW1,SW2,SW3,SW4     dip switch 8 way

It is necessary a 3.3uF on output signal to this filter. 

 

 

 

 

 

 

FINAL CONSIDERATIONS AND LISTENING SESSION

PHOTOS