Wavewife Swans G2 Speaker, SWANS G2 Speaker
Keywords: Huiwei SWANS G2 speaker production
Huiwei SWANS G2 speaker production
Author: Mao Dong
In 2003, DIY had a pair of Hewei "SS6.5R + SS1II" 6-inch bookshelf box, which was adjusted for more than a year. Overall feelings listening to people will also make together, the treble has always had a layer of 胧, and the low frequency cannot meet the requirements. The author likes to listen to the big dynamics such as Disco, the high energy of professional audio, and the high-profit aftertaste, it is a strong low frequency space that is vibrated. After thinking about it. Decided to use another pair of powerful speakers to meet the enthusiasm of DISCO!
By looking up the figure of nearly 10 years of large diameter, two-point frequency book box, combined with the current speaker unit, weighing power, frequency response, tone, and other factors that can be purchased in the current market. Finally locked the Wideworth Swan Series G2 Speaker. The speaker low frequency uses Huiwei "SS8IIN", high frequency adopts Hiwei belt high sound "RT2C-A".
The next step is to design and manufacture the parameters of "SS8IIN" and "R He C-A".
"SSSIIN" is a large 8-inch bass unit with a 222 mm, a lap is larger than the ordinary 8-inch unit. The caliber low frequency radiation is better! Impedance 8Ω, power rated 60W (maximum 120W). The resonant frequency is 29Hz. Sensitivity 90dB.
"RT2C-A" is the first tiered treble technology in Hewei. Also selected in Berlin, Germany, the core unit, the lower limit of 10,000 Hz. The claim can be paired with any low frequency unit, the sensitivity is as high as 93dB. From the perspective of the frequency ram, there is a large attenuation of 10kHz, and it should be considered for a certain compensation treatment when debugging. The impedance is also 8Ω, which is consistent with the low frequency unit, and the divider can not take much consideration on the impedance coordination. Power rated 30W (maximum 60W), which is more than 15W of "SS1IL", and the aluminum panel used in the high-frequency unit is thick.
The box is inverted. The low frequency unit manual recommended that the net volume of the cabinet is 40L, take into account the size of the volume, from the tank density debug angle, and the human positioning, the last reduced to 27L: Previous "SS6.5R + SS1II"
Because of volume and unit arrangement. "Backpoint" adopted. The box is used with 27 mm Chinese fiberboard.
Panel dug aperture, surface mailing. Inverting the two cases of mirroring, similar to Du Xi 2.1. A single cabinet weight is 12kg, and the rigidity of the box is verified in the later audition, and the thickness of the thickness does have a thickness. The vibration of the sides of the box side plate is also slightly 6-inch boxes than "SS6.5R + SS1II". The inverted pipe is selected from the original specifications of Huiwei. Finished product of a length of 80mm in length, because of a comprehensive consideration of the volume, cell diameter, low frequency lower limit, and the specification is most suitable for demand.
The diameter of the inverted pipe is a pair of contradictions. There is no fixed value, and their combined values are the integrated value required for the last box, which will jointly determine the resonant frequency of the entire speaker system.
The box is completed. Connect the power amplifier. Listen to the sound, first feel the clear clear clearness of the tapered high frequency is really impeccable, very penetrating.
At the same time, it feels that its high frequency is different from the SS1II of the IL, the tape-clear high frequency is displayed under an energy-sensing escort. There are different Q1 trebles than Huiwei. The style of Q1 is characterized by a mixture. Belt treble wind.
The grid is more tough, more energy sensation and penetration, DISCO just need such penetration, but the treble directivity is too obvious, and only the head is in a high-stroke horn centerline will have the correct sensitivity. A slight shift immediately feels that the sound is far from yourself. Secondly, the listening of the low frequency unit is very satisfied. The low frequency lower limit can be very deep. Its low frequency dynamics, thickness, roundness, fullness, and density make me like it. However, the frequency response of the high frequency unit and the low frequency unit is in a state of respectively. This is obviously the problem of the divider, and needs to be solved from the divider.
Analyze the reasons. Because of the M1.2 kit divider, its original low frequency unit should be a 6-inch speaker of F6. Find the parameters of F6. Sensitivity is 88dB, SS8IIN is 90dB, then there is 2DB of 2DB. The divider has no manual. I don't know where the division is in accordance with the frequency drawing of the frequency division (see Figure 1), remove R1, temporarily use a nude speaker line to short the way, audition. High frequency sensitivity has improved, but it is not enough.
Remove R5 again. This is the experience in DIY "SS6.5R + SSIII" 6-inch box, and this R5 resistor in parallel with the high frequency unit. Very impact on high frequency. Even a resistance on the series channel is not string.
If the error is installed with R5, it will give high-frequency fluid to high-frequency fluid. This resistance is usually a corresponding impedance compensation with R1, and when the high frequency unit and the low frequency unit impedance itself is very common, it is not necessary to do these painted snakes to consider, so taken. At this time, the ulcery disappeared. The sound immediately transmitted, but the ear feeling sensitivity still has not risen to the correct loudness. via caculation. The "R1" 1.5Ω value on the series channel can approximate the amount of 1.8 dB of high frequency attenuation, while "R5" on the parallel channel
The 15Ω value is an impedance compensation of approximately 3.5 dB on a series channel. Then, the R2, R3 and C4 on the series channels are connected in parallel, and it is for consideration for the wait? See high-frequency cell frequency rare curve diagram, with a significant decay of 10kHz, C4 should be directly connected by a certain frequency band from the influence of R2 and R3 attenuation. Calculate the "C4" 0.47μF company to allow the above 23kHz to pass through. The 23kHz frequency is clearly not coheed with the frequency response period required to compensate the unit. What is the use of 23kHz? C4 I reconsider. Combined with high frequency unit 10kHz or more attenuation is obvious. Then calculate the C4 to pass through the above 10kHz, taking into account the slope. The straight-through turning point is not directly selected at 10kHz at 14kHz. Because 10 Hz-14kHz is a falling edge. Regardless of the steep or slow or slow, 3KHz-10kHz is in 90dB, and will eventually be attenuated. Because attenuation standards will be based on the 90dB of the low frequency unit. It is necessary to decrease to make the entire frequency to balance. at the same time. Into the high frequency feels slow. The 91DB of the 14kHz is selected as the attenuation slope is not too early. Instead, you can make the listening feel more balance. The final calculated value of C4 is 1μF.
R2 and R3 are 10Ω, respectively.
It is 5Ω in parallel, calculating the 5Ω resistance value on the series channel will be a lot of sensitivity of 8 dB, severe excess, I don't know how M1.2 will be attenuated so, no matter what to say. Try to attenuate 3dB according to your own ideas. The three resistors were soldered to the divider, and finally the 2Ω hearing is most correct.
At this time, I feel that the overall sensitivity has been peaceful, and the sound of the sound is overwhelmed. However, the intermediate communication person is very "ear".
That is, it seems to be too prominent. The whole sound sounds split. Mainly, it is concentrated in the Sound Section of the CF.
Find reason again. Turning out the description of the high and low frequency unit, and the frequency sound graph is again examined. Suddenly discovered that the manufacturer recommended division points in the high frequency unit parameter box becomes "greater than 3000 Hz", hurry to ask the dealer, the answer is: The high frequency unit has changed twice in 2000 with 2002, and the previously seen should be old information. According to this, "RT2C-A" and "SS8iin" intermediate frequency part disconnects about 400 Hz vacation. Of course, it is not worth it. No sound will be like this.
From the entire frequency band of "SSSIIN", 100Hz-2600 Hz is an extremely flat and excellent curve. This is the reason why this low frequency unit was originally seen. Now it seems. H can be able to treat the peaks at 3.2kHz, then at least the upper limit can be increased to 3.2 kHz. By making a dashed extended reference analysis by the "sssiin", the slope of the peak is selected at 4,5 kHz. This is more conservative than 1.5 times the frequency of 3.2kHz. The circuit structure of the "R6 series C5, C6" in the DVD portion in parallel, seems to be prepared for smoothing this peak. It's just a big difference. By calculating, C5, C6 is used in parallel with 2.2 μF and 1.8 μF, while R6 uses 10 Ω. It's now in the peak. The upper limit estimate of the low frequency unit has reached 3.5 kHz. Again again, unexpected discovery "depressed" part "fill". But the "ear" portion still exists. Analysis of peak success. Extended the upper limit segment of the low frequency unit. Successful connection with high frequency units.
"Spend" has not existed this, so there is only "ear" on the sense of listening. Why is "ear"? It is too high in a frequency to cause the hearing that is too irritated, and it is not enough to listen to the frequency band, repeatedly listening to different audition signals, and finally the analysis should be that the coupling capacitance of the serial channel of the crossover high frequency part is too large. , C1, C2, and C3 are more than 15.3μF, according to the high frequency unit instructions, the divided point is 3000 Hz ~ 10 calculations. Welding this with a 3.6μF method. Listen again. The problem is greatly improved. Try to change the divide point at 3.2kHz, "ear" problem is basically solved, the speaker sounds more texture, the entire frequency is more balanced! Next, adjust the sound absorbing cotton and inverted pipe. Start all other than all other than the inside of the board. But low frequencies come to "dry", no feeling: then reduce the amount of sound absorbing cotton. Re-cut the sound of sound absorbing cotton, do only cover the backplane with panel.
Finally, I will listen, balance, and the depth of low frequency dive is very good. The delicate feeling is relatively lacking when playing people.
The divider circuit is again analyzed, and the eyes fall to the series elements of R4 and L3. If single from L3, L3 is a frequency division of Tong Tong, together with C1, C2, and C3, together with a second-order divided circuit structure. But the added R4 is added.
Let the simple 13 adjustment is more rich, and the value of 13 is unknown in amateur. The value of increasing R4 will reduce the number of bypass quantities of a frequency segment corresponding to L3, which is the same, and the value of R4 is revealed. First disconnect R4 and L3, that is, the divider becomes a first-order frequency, and then combines the memory of the ear to the varying portion of the sound. It has probably locking this R4 determined which frequency segmentation is located. Try to increase the R4 to 3N, and I feel surprised once.
Really this R4 is the root of the problem. Figure 2 is a final debug result circuit diagram. Figure 3, Figure 4 is a desirable curve.
Our other product:
