Active subwoofer based on 100GDN-3


Created on 20.07.2009 04:00.

Updated 04.17.2020 08:54.

Author: Nikulin Oleg.

The design of the bass link proposed by me, for obvious reasons, does not readily dock with the front acoustics, the spectrum of reproducible frequencies of which does not fall below 200 Hz … but we will not talk about that.

First of all, I would like to mention all the multiple reviews about the “quality” of most Russian speakers, with which I completely agree. And in spite of this, nevertheless, I decided on such a “feat” and spent a certain amount of time and material resources. As a result, it turned out – not in vain …

Here, in fact, is the head itself. This is the workhorse of Russian speakers of the “Corvette” type, from where it was twisted … Due to the non-novelty of the dynamics, I had to paint it. Yes, straight from a spray can of automotive acrylic paint. Attaching photo.

To calculate the body, I used the well-known program JBL Speaker Shop.

By the way, for those who do not have characteristics for this work of Soviet engineers, I attach this:

  • Fs – 25Hz
  • Qts – 0.35
  • Vs – 200 +/- 50 L
  • SPL – 0.1 dB
  • Rnom – 75W
  • Rshum – 100W
  • Rdolgovr – 150W
  • Rlimit – 300W
  • Z – 8 Ohm

After not long inferences and poking the computer keys, the following case parameters were drawn:

Vb – 80 L; Fb – 27 Hz

Following the parameters of the case, I calculated the phase inverter. And it was calculated according to the principle of “Matarazzo” (the designer of the design is Jean-Piero Matarazzo). Matarazzo, in fact, proposed a new one, which takes into account the effect of the wall of the speaker cabinet on the calculation of the length of the phase inverter.

Here the frequency is in hertz, the volume is in liters, and the length and diameter of the tunnel are in millimeters, as we are used to.

This scientist proposed a solution regarding the body of the phase inverter itself, namely, to reduce the length of the phase inverter tunnel.

Tunnel “Hourglass” and slotted version of “Hourglass”, respectively

This geometry makes it possible to shorten the tunnel in comparison with the original, constant cross-section, at least one and a half times, or even more.

Dimensions of an hourglass-shaped tunnel equivalent to a cylindrical one with a diameter of 80 mm and a length of L0

L0 Lmax d D L1 L2 h Wmin Wmax
160 100 58 81 60 twenty 50 52 103
200 125 58 81 75 25 50 52 103
260 175 58 82 105 35 50 52 104
330 200 55 82 120 40 50 48 104
400 250 55 83 150 50 50 48 105
500 300 54 83 180 60 50 45 105
630 400 54 84 240 80 50 45 106
750 450 54 84 270 90 50 45 106

The same for the original tunnel with a diameter of 100 mm

L0 Lmax d D L1 L2 h Wmin Wmax
270 175 71 100 105 35 60 69 130
330 200 71 100 120 40 60 69 130
420 250 71 100 150 50 60 69 130
530 300 69 102 180 60 60 66 133
650 400 69 102 240 80 60 66 133
800 500 68 103 300 100 60 63 135
1000 600 68 103 360 120 60 63 135
1180 750 68 103 450 150 60 63 135

What the dimensions in tables 3 and 4 mean will become clear from fig. D and d are the diameter of the cylindrical section and the largest diameter of the tapered section, respectively, L1 and L2 – the length of the sections. Lmax – the full length of the hourglass-shaped tunnel, just for comparison, how much shorter it was possible to make, but in general, it is L1 + 2L2

Technologically, an hourglass of round cross-section is not always easy and convenient to make. Therefore, here it can also be made in the form of a profiled slot (see Fig.) To replace a tunnel with a diameter of 80 mm, Matarazzo recommends choosing the height of the slot equal to 50 mm, and to replace a 100-mm cylindrical tunnel – equal to 60 mm.

Then the width of the section of constant cross-section Wmin and the maximum width at the entrance and exit of the tunnel Wmax will be the same as in the table (section lengths L1 and L2 – as in the case of a circular cross-section, nothing changes here). If necessary, the height of the slotted tunnel h can be changed by simultaneously adjusting both Wmin, Wmax so that the values ​​of the cross-sectional area (hWmin, hWmax) remained unchanged.

Here’s what I got. The parameters are as follows:

  • Lmax – 190mm
  • L1 – 114mm
  • L2 – 38mm
  • H – 60mm
  • Wmin – 69mm
  • Wmax – 130mm

A bass reflex was made of 5 mm plexiglass, then painted black.

So. I’ll make a comment here on the bottom of the front and rear walls – the front, as you can see, is double, glued and twisted. The holes for the self-tapping screws were filled with auto putty. (In the photo, they have not yet been putty) And the screws themselves are smeared with glue in order to exclude the possibility of twisting them from vibration (in my practice, there have been cases). The ends of the wall are rounded.

Chipboard material, like the entire body. All joints were glued with “liquid nails”. The rear wall has a niche for placing an amplifier, a frequency cutoff control filter and a phase shifter there. the subwoofer was planned to be made active.

At the back there is a hole for the bass reflex port.

After careful processing, the body was pasted over with self-adhesive tape for the future interior of my room. (The speaker system on the right, also of my manufacture, which I will also paste over in blue).

The legs for the sub were made of steel, tapered (but this is really not me anymore). The internal volume of the case is pasted over with noise insulation and a sound absorber is installed (it is better not to save on these components).

As for the amplifier, it was traditionally made on the famous one. And no matter what they say about her, here, they say, shit, etc. for a home-made subwoofer amplifier, I have not yet found better in the ratio of ease of strapping / quality / cost.

Here is a bridging diagram of two microcircuits, with a total nominal output power of 150 watts. This is quite enough for the buildup of the 100GDN-3.

The diagram shows light blue capacitors – these are through-feed, you can put electrolytes instead of them (but I recommend to put non-polar ones). I etched the board with ferric chloride, translated the drawing from a printout of a laser printer.

The power supply is 2-pole, shunt capacitors and dischargers are soldered in parallel to the supply capacitors. Diode mots with a Schottky barrier, 10A, 1000V each. The earth connects at one physical point. The wires connecting the power supply to the amplifier are selected with a cross-section that allows a current to pass up to 20A.

Schematic diagram of the amplifier

Schematic diagram of the power supply

Block of the cutoff frequency regulator and phase shifter

An adder and a first-order passive filter with a cutoff frequency of about 150 Hz are installed at the input. A 2nd order Butterworth filter is assembled on the VT1 transistor. The cutoff frequency varies from approximately 50 to 200 Hz. The phase regulator is assembled at the op-amp. The lower limit of the range is 15 Hz. The input signal must not exceed 1 volt, otherwise distortion may occur. The signal at the input is attenuated 4-5 times (12-14 dB), but, however …