[ad_1]
Created on 06.10.2019 22:29.
Last updated on 23.02.2020 17:12.
Author: Alekseev Igor.
Time passes, which puts everything in its place wonderfully, namely: once again the developers of dynamic drivers return to paper diffusers. So why don’t we learn from our mistakes and are always ready to make new ones after someone? This thought tormented me for a long time, until I decided to cast aside all doubts and make my own home theater acoustics using the existing old speakers.
And their choice, as befits a “plushkin”, is very standard – these are low-frequency 30GD-2 (75GDN) / 25GD-26 / 10GD-30, “constantly waiting in the wings” 4A-28 / 4A-32, a bunch of mid-frequency 5GDSH-5 / 3GD-38 / 4GD-35, an even larger pile of “universal” 1GD-18 / 2GD-40 / 3GD-42, etc. Tweeters usually did not live up to the moment of “archival storage” due to their former constantly low quality of performance and reliability, with the exception, perhaps, of several 2GD-36.
It all started with modeling the acoustics of the room. The goal was significant: to optimize the low frequency range for the listening point and, at the same time, to prevent the “bass” from leaving the room through the doorway into the corridor (so as not to interfere with the children sleeping in the next room).
I really wanted to “attach” the 4A-28 because of their excellent reproduction of the lower middle, and the rest of the set of heads showed an average sensitivity of 90–93 dB / W * m, which determined the choice of the low-frequency design of BandPass (higher efficiency), Fortunately, the location of the “virtual subwoofer” completely coincided with the locations of the front speakers when the set “acoustic” conditions were met.
As it turned out, the dimensions of the low-frequency link on the 30GD-2 turned out to be quite acceptable at an operating frequency range of -3dB 42-160 Hz (18Hz at -10dB). It was decided to make the mid-frequency part in an “open box” design. Do it like that! In fact, well, I do not like the work of midrange in closed boxes. All that was left was to choose a suitable tweeter in terms of sensitivity and sound, which would not spoil the sound of the “paper”. The choice fell on a product from a series of car horns – a silk dome without magnetic fluid.
As a result, we got a project with a limitation of the LF digestible power of 50 W (the stroke of the 30GD-2 diffuser does not exceed 8mm). I want to say right away that it is impossible to do “on the knee” and without measuring equipment and modeling of 4-band acoustics. This was my 9th deliberate, measured and “hard-won” project, and I perfectly understand the limitations of certain speakers on their use.
The next condition, which I really wanted to fulfill, is the implementation of a wide (± 30 deg without distortion of the frequency response) radiation pattern in the horizontal plane, as well as precise alignment of the phases of the radiation of the speakers and, most importantly, the identity of the phase conjugation of the radiation of the heads in the range of 2 octaves in the frequency range section.
And if the first condition was realized by choosing the frequencies of the band section, adequate to the diameters of the diffusers of the heads and installing side speakers to reproduce the upper middle, then the second condition required an inclined front panel. After a lot of research, the shape of the case has acquired quite a pleasant look. How it is done inside is shown in the next photo.
In the lower part – 4th order bandpass, middle wedge-shaped compartment for filters, above – mid-frequency design open box
The thickness of the outer walls of the case, the rear wall and partitions of the low-frequency section – 35 mm, is recruited from 3 layers: birch, MDF, pine. There are no body vibrations !!!
The back wall of the midrange box is perforated, and the calculated area of the holes and the thickness of the material (plywood 12mm, reinforced with beech slats) is a phase inverter for the installed 4A-28. This made it possible to reduce the amplitude of displacement of the diffuser by 150 Hz and to apply a 1st order filter from below.
Speaker 4A-28 is fixed banally outside the case. And although its operating range is limited to 200-2700Hz, I did not remove the center horn because of the possible increase in the resonant frequency. Central upper midrange – 5GDSH-5. Side upper midrange – rare 1GD-18, electrically connected to each other in parallel and in series with the central one.
The lateral speakers are glued into milled mounting holes on silicone; the photo shows the gluing process.
The tweeter is placed in a birch “bullet” -type housing, which made it possible to accurately phase match with the upper midrange in depth. The tweeter plays out the 7kHz – 25kHz band (2 +1 tweeters were selected from a batch of 12 pieces, so they wouldn’t say – there is a marriage everywhere!).
Filters of the 1st order (with the exception of the low-pass filter – the 2nd order for effectively damping the resonance of the external port of the 1250Hz phase inverter by 30dB). The acoustic filter order is standard 2nd, band matching at the level of -6 dB. Phase mismatches of bands in the 2-octave overlap region are no more than ± 15 degrees (acoustic phase characteristics of adjacent bands are as close as possible).
The central channel is on 2x4GD-35 and a silk squeaker, similar to the front speakers. The design feature is a bass reflex with an internal resonator a la AS “Symphony”. The front panel is tilted “towards the listener”. Despite the decent quality factor of the 4GD-35 heads, the acoustic resonance is damped quite well and completely fits with the frequency response of the room at the calculated location of the speaker.
The rear speakers are based on 4GD-35 and 2GD-36. 1st order electric filter – easier and better to come up with something difficult. Also, a bass reflex design, internal resonators are used, but already low-Q and 2 frequencies. The maximum vertical directional pattern is directed towards the listener (the rear speakers are ceiling mounted). Inside the speaker, the absorber is minimally applied.
Placements of acoustics are selected in such a way that the shortcomings of the frequency response of the AC’ok are largely compensated by the “acoustic features” of the room.
I, of course, understand that “crossing two camels” is a delicate matter, but in this case all the “humps and depressions” are precisely calculated. And who said that it is easier to live with one disabled two-humped camel (room)?
External finish – color scheme and alkyd varnish. Protective grilles are made of slats, covered with “radio fabric” and fixed to the front panels with special caps.
Final appearance of Agnett’s kit:
To make life easier for neighbors from below, the front speakers are installed on spikes on reinforced concrete pedestals (25 kg), glued with epoxy into birch boxes.
I will not focus your attention on the common truths about high-quality and properly oriented connection cables and details of crossover filters. Everything is matched. A good German paper broadband and my own hearing were used as an “indicator”.
All paper speakers used in this project have been modified (guerlain, impregnation of diffusers with “tricky compounds”, muffling of baskets, and 30GD-2 and 4A-28 were disassembled and, after eliminating factory assembly errors, reassembled). Yes, I forgot that the front speakers are connected to …
[ad_2]