It just so happened that this set of speakers I have used as a computer. They were purchased on the principle of the best that was enough at the time. In 2008 it was 600 wooden rubles. At the moment the cost is 1200-1300 wood.
It's hard to argue that padonki and just the sound source at a price they play completely, but with time incredibly disgusting sounding asked to change something. To describe the sound of this class acoustics difficult, but in a nutshell, it was completely absent lower bass, some significant sound pressure started from 60Hz, abruptly grew to 80Hz (and where nightmare was buzzing with all body), the absence of at least some of clearselected in reproduction, not that parts, at least some. When watching movies I had to crank the bass to a minimum, otherwise the whole sound, including dialogue was reduced to a continuous resonant bubnezh. The scene won't even speak, she just was not all sounds were clearly stuck to the speakers. All the sound through some kind of cotton veil. In General, a typical pattern for active speakers for the 1200r. However, for some reason there was a clear understanding that these speakers can still more. On the Internet there was not found a single example of this tweak acoustics, but there were pink snot fans MP3 with built-in speaker mobile phone, about the great sound of this model. Also found characteristics from Microlab, where shamelessly bralos about magnetic shielding and frequency response from 30Hz...
So at some point in time all this stuff was collected in a bag and related to work. The situation was complicated by the fact that the work we are dealing with digital technology, so of measuring equipment - mediocre oscilloscope and Chinese multimeters. Largely have to rely on your own ears and an acoustic frequency generator on the computer, good ears trained and relatively experienced. All alterations were made from scrap materials, sometimes not quite those that would like to use, but maybe it's for the best, tweak should be commensurate with the work and the original cost to the donor, otherwise it becomes meaningless.
In the first place was dismantled active speaker and emptied it into a transformer and amplifier and studied. Speculation immediately confirmed. TDA2030 inclusion in the model with maximum savings on the details. But for a start choose columns - the case of MDF though, but due to the small thickness of the walls rang like a glass when tapped. Glue the spacers I did not, went for a simpler and equally effective for such enclosures by - Vibroplast M2 over the whole inner surface.
After rolling of the hull became deaf as concrete, added a bit of holofiber, usually a lot more stuff, but I was afraid in a compact body to strangle the speaker and create unhealthy conditions for FI, so you can lose punch, if you get carried away gasket, which is not wanted. FI touch not, the wind noise they, oddly enough, had not issued, mindlessly extend not want, and persons of nowhere, a wall close and do a completely new and curved without calculations and measurements possible and not correct I think. The entire filter in these columns consists of a single capacitor on the Squeaker 2.2 µf, that its good if he wasn't a polar electrolitos of unknown origin. Replaced by K73-17 2,2 µf 63V to Play with more complex filters without measuring equipment - would be foolish. The wires were already soldered at the factory and the quality is quite consistent with the quality of everything else, so they remained unchanged. Passive column on this were all, with the former active conversation continued.
Initially the plan was to finalize the case, change the electrolytes for nonelectrolytes and change on TDA2030 TDA2050, but after housing improvements, it became clear that the TRANS and the amplifier in the column guaranteed to do nothing, appeared bass, and with them a background in 50Hz, actively induced by the transformer on the WOOFER. Previously, this was not heard. Yes and the difference in sound empty and Packed column could be heard clearly, so it was decided to make an amplifier with power supply in a separate enclosure, but Tempblock to leave in the building column, first did not want to mess with it, uprooting, and secondly there was nowhere to move it, and the space it takes up little and generally passive, respectively connectors line inputs, it was decided to also leave in place, not to carry thither mezhblochnye, but from the connectors of dubious quality internal wiring I strongly disposed by the method of wire cutters and a soldering iron.
Connector to the amplifier together with a cord was used from a dead sukria Pioneer :) it directly from the column from template signal supplied to the input of the amplifier and from the amplifier is fed, the led on template, and signal and power LEDs in a shielded pairs, inside this cable a lot, decided to play it safe. Connector classy, with a latch. Pasted into place the power switch, all the fastener holes filled with hot-melt adhesive, outside plastered black arokiam, and the inside of vibroplasty. Capacity template used Mylar from birth, so I didn't touch. This former active column was collected and all conversations with her, too, ended.
The amp started to present surprises almost immediately. By itself, the fee divorced well: earth star, the diodes in the bridge is shunted by film producers, there is space under the tiles are soldered onto power amplifiers before, there really is nothing not soldered. After the diode bridge, there are 2 of electrolyte MF and 2 on 100mkf, though they shifted specifically to the amplifier's right channel and the left goes jumper humorous section. Jumper strengthened, 100mkf electrolytes vipal and soldered on the pads large electrolytes on the reverse side of the card is lying, at the same time sotirova their ceramic 0.1 µf and poured it all with glue gun. The amplifiers have changed with TDA2030 on TDA2050, why not doing it at the factory for me is still a big mystery (in the process of the tweak was born the opinion that TDA2050 more demanding to the trim, so with TDA2030 you can save not only on the chip, but also on the harness, and the sound don't care).
As input isolation stood electrolytes on 1uf. Search in the bins gave birth to a dead sound % Kenwood were extracted 2 plenochnykh on nF and 2 on nF, brown, more data about them =) Soldered in parallel, one in place of the electrolyte, with the second back side of the Board. Filled with glue. The result of the test enable surprised - strong background of 50 Hz in the right channel, instantaneous heating of the amplifier and the transformer and hard clipping even at low volume. Went to look for short-circuit or ground. Not found. Began to change the capacitors in the TV sometimes, the culprit was found quickly.
The site parallel dynamics should stand antisana RC chain of nonelectrolyte 0.1 µf and 1 Ohm resistor for TDA2030 and 2.2 Ohm for TDA2050. The Chinese had the same jumper instead of the resistor and with TDA2030 it worked fine, but with 2050 stopped and shot the capacity. I reasonably have decided that the length of the wires less than half a meter this circuit is not needed and vipal capacity. Background we first went, but when they return the amplifier to become unstable, and gave the generation of random frequencies. Soldered in their place k73-17 0.22 µf 45V, not helped, capacity is not dead, but there was no stability. As a result all the same soldered 1.2 Ohm (that I found) instead of jumpers and all vanished, generating not heard nor ears nor the oscilloscope, the heatsink is barely warm, the TRANS is cold. Why generated stably only the right channel I never knew, the replacement of the amplifier to another new result is not given, probably something to do with the wiring of the left channel.
In the feedback circuit is electrolit to earth on MF, so required by the site. On its good too to change at least unipolare, but these are not available, so I replaced it with 47mkf electrolit, subjectively, the bass was slightly better, literally on the verge of audibility. Left to compete with high frequency noise included in columns. The site advises as close as possible to the chip to put ceramics on nutrition, 0.1 μf. I decided to understand the site literally (after the trouble with the resistor and soldered the capacitors directly to paws. The timbre of the noise changed...
Began to try one by one all in a row, falling under the arm capacitors. Changed timbre, amplitude, frequency noise. In the end, full of peace was achieved with capacitors marked M15 - how much, I don't know, I guess 0.15 UF. Now the amplifier is completely stable, not heated over 40 degrees and to distinguish without the input signal is switched from off you have only the led, even if you bring your ear close to the Squeaker with a fully turned to maximum volume. In this work with the amplifying part was over, all capacitors and protruding elements were mercilessly filled with glue gun.
Now the
It is known that by Microlab life emotic on transformers with all the consequences - low maximum power, clipping, penetration channels at each other for food, the sound of trance, etc. native trance was written volt (depolari) 0.8 A, but these characteristics caused severe doubt. Rewinding is not, iron frankly weak.
From the depths of the closet was extracted CCI 250-220-51. Depolari 11B and stated the current is 0.9 And it was getting more interesting, the iron was clearly more and the build quality significantly. It was a pity to lose precious volts on the diode bridge, so the light is extracted almost all the caches, but Schottky diodes were not found; therefore were stupidly selected with the lowest resistance of the junction and soldered instead of relatives. Assembled it all the happiness in the housing of a motorcycle battery charger, barely all fit in there. Transformer stuck on the zippers, excellent vibration and electrorate, the amp is bolted, unnecessary holes sealed with ORACAL body to the earth to plant did not, cosmetics inducing too. I will go, and time was running out. Later home assessed the case as "so disgusting, that already like", at the same time provokes to listen Sex Pistols ;)
Results
Completely disappeared buzz and in this range woofers, speakers begin quite heard to push already at 30Hz, which is really surprising. Generally articulation and dynamics the bass is very pleased, though subjectively it was less, but the point here the lack of resonance of the casing, so that now the bass control at maximum, one 4 inch driver podpinyvat. Clipping greatly moved, detailing grew monstrously like a cotton shroud with speakers pulled, the stage rises as children to school, rare echoes stick to your speakers, the sound is unobtrusive and pleasant. Generally jumped column and adjacent the top. At the moment I am sure that will easily outperform regular shirpotrebovskoy system to 5000r confident, they would have tweeters fabric and not plastic, but buying one is too cool, so if only for the freebie will come ;) HF at 14-00, more plastic grates on the ear, less gluconato already. The bonus was granted the ability to turn off the speakers on the night and put the cell phone at least an amplifier, before the call or SMS in the zone of 2 meters from the active speaker called guaranteed obscene reaction of all households. Especially at night.
In General alteration satisfied at 100% result recoup initial investments and investments and work on fine-tuning, the result exceeded my expectations.
Microlab can do good designers for subsequent finishing, and proved once again.
Author: Edward Wiclif; Publication: www.cxem.net