"Type: 10W 3D voice success rate
Source of this document: http://www.geocities.jp/team_zero_three/poweramp3/index.html
Moth
"The Third" "10 W + 10W Stereo Amplifier Production
In the beginning
I made it again. The third. Even though it is not a self-for yourself, it is not a self-use, and "listening to the sound of self-made amplifier", but I asked "I want!" So he decided to make a 10 W of an amplifier. By the way, I was told that five people to "want!" . .
As usual, we have exposed to ignorance, but please skip the suspicious part with personal departure.
I will write it once, but my friends are eating Tada. That's because my experiment is a little. If you're looking for a friend? Are you saying "Do you're", but it does not say "experiment" etc., but it seems that you are looking at the atmosphere somehow (actually it will be "the experimental stand" rice field).
10W + 10W stereo power amplifier produced. Succeeded in putting a sense of luxury by painting the case black
Moth
overview
The output is about 10 W + 10W. This time, I have a major part in his digit of Osaka Nihonbashi. It should be noted that it is 15V-0V 15V-0V 2A toroidal transformer. It is put in the middle of the middle of the middle of the entrance and put it in the middle of the middle of the middle (if there is stock). I have only seen this transformer here. The price is a little higher than 4,000 yen, but clearly monaural he is 20W and 10W + 10W class semiconductor amplifier. If you are walking around the inside of the store with a large-scale part of the amplifier, you're trying to make a shop or a clerk to another customer and it's embarrassing. And since I bought two other parts for 2 days, I may have been able to learn my face to my clerk. At least 3 units will go to buy at least 3 units, thank you> Sore by clerk. Although the semiconductor amplifier kit is demonstrated per the front of the cash register, it seems that the one made for this kit is selling a single item. This time he uses 15V-0V 15V-0V 2A toroidal transformer as a power transformer for current amplification stage.
Power supply voltage for voltage amplification
Since the voltage of the current amplification stage is set to one side 15V, the power supply of the voltage amplification stage is one side 18V considering the loss of the bias voltage. 18V-0V A transformer called "LH2812" of 18V-0V 40mA is sold. Because it is a small size, he can be implemented on a 2.54 [mm] pitch board. However, because the capacity is 40mA, it is necessary to make sure that the circuit constant should be determined so that no current flows.
Schottky barrier diode
The shot key barrier diode was used for the power supply rectification diode. The Schottky barrier diode is short because the recovery time is short and the order voltage is lower than the general diode. Schottky Barrier Diodes are considerably more expensive than generic rectification diodes, but are relatively cheap in "Digit". When I went (December 2007) he was 100 V1A 5 stable, and he was 200 yen (model numbering). Shot Key Barrier Diode for Current Amplification Stage Powered He bought a 100 V 3a (it was 200 V, model number, unit price).
Output element
The output uses 2SJ200 / 2SK1529 (MOSFET). This also digits with digits. Heat sink was unknown model number, but he used 475 yen. The power supply block capacitor uses two of 25V 33000UF. This block capacitor also purchases two 35 capacitor bands due to a diameter of 35 mm. When the above 15V-0V 15V-0V 2A is rectified and smoothed, it was about 15 * √2 to 21 V, so I thought that it was okay at 25 V withstand voltage, but it was actually rectified with a Schottky barrier diode and measured without load And it was 24 [V] and the last minute. ([2008/01/20] When measured during operation, it was about 21.4 [V].)
Differential amplification circuit device
His 2SK389GR and 2SA1349BL used for differential amplifiers used a few years ago, and he used some of them. However, when I looked up again, it was so high. . . Such a small element is the same as the output element. Such varieties are already finished production, and the price may be determined by the number of markets. But as me, I'm a little careful to buy some such high elements. Since the second or later he will replace the 2SK389GR with his 2SK30A and try to review the current value. This time because the power supply voltage is low, 2SA1349BL is 2SA798G (2008/1/25, 157 yen for Wakamatsu commerce, and this is still cheap), or it is likely to be used enough per 2SA872A. It is a variation of the element I'm worried, but I checked it with 100 pieces of 2Sk30a and checked out and checked the IDSS, and the elements that show almost the same value on the tester Yes. I think that IDSS is only a guide to know the variation of the device, but I think that it may have a characteristic close enough for practical use. The price is also about 100,000 yen if he is 2SK30A (2SK389GR is about 3 pieces 1500 yen).
Case
The case used an aluminum chassis called "FL-8" in Fuji Chassis. W275 × H40 × D150, T = 1.0. This was also purchased with "Digit". Because there is no bottom lid, all parts need to be fixed on the top of the case. This side is the most difficult difficulty production, but the absence of the bottom is conversely improved and the maintenance is good (1155 yen) and cheaper (1155 yen). The rubber foot him a 10 mm x 10 mm square sticking type to four corners of the bottom.
others
As I made a lot, I decided to try a little about what I did not do before:
All discrete
Paince chassis
Moth
One by one
So it is a circuit diagram (click and display another window).
Constant voltage power supply circuit
The constant voltage power supply is a great deal of the impact of Mr Kubota. On my own, power tr is troublesome to drive the base current, so he is the dimension that he uses MOSFET to output. Zener diode uses RD7.5E (VZ = 7.5V), but when I look at the data sheet now, he seems to be less change in Zener voltage change due to temperature per RD6.2E. Since the Zener voltage changes slightly by the current flowing, a constant current circuit is provided. Here we used his IDSS of his 2SK330Y. IdsSS Measured Values 2.3 When using one of [MA], he got a voltage of about VZ = 7.1 [V]. The IdsS of the FET changes at ambient temperature, but looking at the RD7.5E data sheet, it appears that the change in Zener voltage near the number [MA] is small. However, the Zener voltage itself that occurs has some variations for each element. This time we do not have variable resistors for voltage adjustment, so take 2SK330Y and RD7.5E to get the desired voltage. The value of the resistance to determine the base voltage of the error amplification TR (2SC1775AE) is strange, but his 4.7K resistance he has a large amount of 4.7K (= 9.4K) or three (= 14.1K) It is obtained by connecting in series.
Constant voltage power supply circuit transformer
The transformer for constant voltage power supply circuits is 18V-0V 18V-0V. Connecting 18V and 0V in the middle and is a center tap. I wrote in an overview, but because the transformer capacity is quite small, it is necessary to be careful about the current consumption of the voltage amplifier circuit. The constant voltage power supply is set to 20V, but when we consumed 40 [MA] to test, it was good because only 30 [MV] decreased. The equivalent internal resistance was 30 [MV] / 40 [MA] = 0.75 [Ω]. ([[2008/1/20] The transformer is still set to 18 [V] and it seems better to use it.)
Constant determination of constant voltage power supply circuit
The output voltage is determined by the voltage between the base and emitter voltage of the Zener voltage and the error amplification Q5 and the resistance value (R3 and R5) dividing the base of the error amplification Q5. The base potential of Q5 is about 8 [V] with the zener voltage and the base emitter voltage of Q5. It is a bit different from the circuit diagram, but when the current of 1 [MA] is performed now, R5 = 8K [Ω]. Assuming that Q5 HFE = 500 (almost actual value), the current flowing into the base of Q5 is 2.3 [MA] / 500 because constant current 2.3 [MA] flows to the Collector of Q5, but this current is The value can be regarded as small enough for 1 [MA] flowing to R3 and R5. Moreover, the current that flows into the basis of Q5 is designed to flow to R5 that can be ignored. In this way, we think that the current flowing into the base of Q5 is ignored and almost 1 [MA] is also flowed to R3. Since the voltage applied to R5 is about 8 V, the value of R3 is determined to determine the value of R3 and R3, which is equal to the output potential. This time, since the output potential is 20V, if R5 = 12K [Ω], almost 20 [V] will be obtained.
When the above story is summarized, the Current I of VB / R3 flows to R3 and R5 by the base potential VB of Q5. Therefore, the output voltage Vo is obtained as Vo = i * R5 + i * R3 = (Vb / R3) * R5 + (Vb / R3) * R3 = Vb * (R3 + R5) / R3. In the actual circuit, R5 = 14.1K, R3 = 9.4K, so VO = 8 * (9.4 + 14.1) /9.4 = 20 [V]. [2008/1/20] I wrote above, but I think that it is better to set it to 18V because the transformer is 18V 40mA.
NFB control
Error amplification TR operates as an emitter ground inverting amplifier. In other words, when the base potential decreases (with a ripple component etc.), the collector potential drops, so the potential output from the MOSFET also decreases. Conversely, when the base potential is lowered slightly, the output potential operates to rise. Such negative feedback (NEGATIVE FEEDBACK, NFB) control keeps the output potential almost constant.
Capacitor for oscillation prevention
It may oscillate because it is NFB control. It is a 220 [UF] capacitor at the output to prevent this. It seems to be difficult to oscillate as the capacity is increased, but if it is too large, the frequency characteristic is likely to be worse. 220 [UF] may be bigger.
Capacitors for reducing residual ripple
47 [UF] capacitors connected in series to FETs seems to be working to escape the ripple component to earth. 47 There is room for confirmation whether [UF] is the best (Borrowed from the circuit of Mr. Kubota).
The 10 [UF] capacitor connected to the base of the error amplification TR has the effect of reducing the residual ripple component of the output, and it seems that "the capacitor" is absolutely necessary to ensure stability and performance (Design of a flat transistor circuit P238). Here isLet's follow it directly.
Capacitor for noise reduction
A 100 [UF] capacitor next to the Zener diode is applied to absorb white noise generated from the Zener diode. There is room for confirmation whether 100 [UF] is the best (this also borrowed from the circuit of Mr. Kubota).
This is such a place for constant voltage power supply circuits.
Contained constant voltage power supply produced
Moth
Amplifier circuit
The amplifier circuit is an NFB with a differential amplification circuit 2-stage + current amplification stage 1 stage. I think it is the most basic configuration among practical amps. Examples use MOSFETs instead of TR. In this configuration, he does not make a heat runaway (as much as I know) if the MOSFET does not heat the Bias circuit of TR (as much as I know), so I think he is easier than TR.
About differential amplifier circuit
The differential amplifier circuit is configured using two TRs, and it seems that two TRs have the effect of offset the non-linearity with two TRs from the nature of amplifying only two input differences. Therefore, I think that it is necessary to use TRs that have a good characteristic to bring out the performance of the differential amplifier circuit. However, it is impossible to prepare two TRs that have a completely equal characteristic, and slight variation will appear as an offset voltage or strain. This time he adopted a product with two transistors simultaneously manufactured on one chip so that the variation of TR characteristics minimizes. The first differential amplifier circuit was composed of dual FET (2SK389GR) and the second stage with 2SA1349BL (dual transistor). This time, we expect the operation improvement of the two-stage differential amplifier circuit, and the load is constant current load (3.5 [mA]) with constant current diode, but it does not understand well as the effect.
Drive power of output?
MOSFET's drive power may be insufficient. MOSFETs seem to be heavy at high frequencies (unconfirmed), so the electrical characteristics in the high frequency may not be very good. However, the feeling you heard is a wonderful sound from low band to high range. I actually heard my friend, but said that "the high-frequency growth is wonderful." He will try to add a MOSFET drive stage to the Unit 2 (but I do not feel too much).
First Differential Amplifier Circuit
He uses FET in the first different differential amplifier circuit, but this reason is still not well known. If the current flows to the base, it feels that the current flows out, but if it is, it is said that he should use FET without using TR. I also feel like it will be. However, when he is FET, the second stage is likely to come out with the problem of bare gain. Ma I feel like I'm experimenting with myself.
Constant current circuit of the first differential amplifier circuit
The common source of FET's differential amplifier circuits are constant current in 2SC1775AE 3 stones (Q9 to Q11). This he makes TRs closely attached to each other and adhering with an adhesive, so the ambient temperature is almost the same. Normally, it is usually similar to the same amount of R19, R20 and R23 to 3 K, and the current flowing to the left and right side is made as much as possible, but this time the capacity of the transformer is small, so the two stones on the left (Q9, Q11) Suppress the current to be slightly vented to 1 [MA] and spend about 3 [MA] to Q10. Q11 has read the description "Into Temperature Compensation" somewhere. It corresponds to 2 diodes in Q9 and Q11, and the fixed current circuit of the differential amplifier circuit is a circuit that does not perform two diodes in series in series, but it is not necessary to know why it is necessary. (I feel that one is needed ...).
Where the wiring has almost completed
Earth wiring and ham noise
When the ground was wired in accordance with the circuit diagram, the ham noise was slightly heard, so I changed the ground with the above schematic, but it was not clear which one was better. If you change the stress of the ground, think that it is a practical scope (→ see the two units and see the "Earth wiring path").
When I get my ears close to the speaker, I still hear Ham Noise, but I gave up on the scope of practical use. In fact, you can hear a minute spectacular sound that is larger than Ham Noise.
Attenuator
Also, at the time, ATTENUATOR (attenuator) was attached to the input, assuming that line output such as TV is directly connected. Using a rotary switch of 2 circuit 6 contacts, -8 [db] / STEP. It was high enough that the number of contacts is high, and it was troublesome to solder in a lot of resistance. -0 [db], -8 [db], ..., - 40 [db], but since the gain of the amplifier itself is +20 [db] (10 times), the signal at the end of the input is -20 [DB] to +20 [DB] can be set in 8 [DB].
Six-stage attenuation (8 dB / STEP)
Although it is called attenuation, etc., it is only that the variable resistance of A curve (log scale) is replaced with individual resistors. However, the calculation is quite tired. When a contact is connected to the amplifier's input, the resistance value from the signal input to the contact point to the contact point to GND is to be determined in order to determine the resistance value in order, but actually calculations Since the perfect resistor can not be obtained without custom-made, it can be obtained, so you can actually get it, and use a resistor close to the calculated value as much as possible, so if you use this value, use this value as it uses this value. The calculated value of the resistance of the resistance is in order to obtain a resistance value as close to the target value, as the resistance value to be actually used is ... ". Place the calculation method here. Although it is a brave calculation method, if it is about six stages, it will be done immediately if you calculate it. If the number of steps increases more, you will want to create a calculation program.
It feels like I actually tried using Attenuator, but if it is 8 [db] / STEP, the volume of the volume may have been too big. Since the perspective is six steps, the next smaller sound will be a larger sound if you connect the line output. So this machine is basically recommended to use in -0 [dB]. (Smans> Friends) Next is 7 [db] / STEP. (^ _ ^;
Sense
Analog circuits are compensated. As you can not move "to move" like digital circuits, you can not be careful with part of the circuit. I'm out of mind without getting out of mind.
Moth
Secondary
So I made the second machine and I made it. It took a long time to say.
2 units. The power switch and the LED are reversed from the 1st unit due to the left-handed specification
So it is a circuit diagram (click and display another window). Circuit diagram of Unit 2
Power supply circuit for voltage amplification
This time, the power supply for voltage amplification was set to approximately ± 18 [V] (measured value is +18.21 [V] and -18.20 [V]). The resistance of the 2SJ76 / 2SK213 gate is changed to 100 → 1K, but the power supply made a mistake as it made a circuit diagram. It is not that there is a basis for being 100 [ω], so is it good? I think there is almost no influence. An amp that I want to be Tada (← hey). I'm writing while listening to music with two units now, but it's a surprise sound.
First Differential Amplifier Circuit
In the amplifier's first differential amplifier circuit, we stopped 2SK389GR and he used 2 2SK30ATM-GR. I used things that IdSS showed the same value in the tester, but if you actually move it, the operating point (the measured value of the voltage drop of the drain resistance) is as much as 0.03 [V]. In the schematic, both drain resistance are 1.5k [Ω], but in fact, it is adjusted by putting a resistance of about 22 [ω] in series. But there is no problem in practical use even if you do not adjust.
It is an influence on the sound of the deviation of the characteristics of the two-element characteristics used for the differential amplifier circuit, but clearly, "There is a noise in which the characteristics of the element of the first stage are missing It is not a level that I understand. I am confident that the amplifier using his 2Sk30ATM-GR with his 2Sk30ATM-GR and the amplifier using 2SK389GR is switched by the switch, and I am confident that I can not distinguish it (but I can not boast) . Anyway, it is good because 2SK30ATM-GR has a sufficient practical and satisfactory sound quality.
Temperature compensation effect of Q11?
It is a transistor Q11 that seemed extra in the constant current circuit in the first stage, but this is probably compensated for the temperature change of the base emitter voltage of the second stage differential amplifier circuit Q7 / Q8? I thought.
For the time being, Q9 shall offset and compensate for the temperature change of the voltage between the Base and Emitter of Q10. If there is no Q11, when the temperature around Q7 / Q8 increases, the voltage between the base and emitter voltages of Q7 / Q8 decreases, so the common emitter potential of Q7 / Q8 falls. Therefore, since the voltage applied to the common emitter resistance is increased, it changes in the direction in which the current value is increased.
However, if Q11 has a decrease in temperature rise with temperature rise, the emitter potential of Q10 rises as a result. Then, since the current flowing to R23 increases, the current flowing to R10 / R11 also increases, and the voltage applied to each is also increased. Therefore, the potential of the Q7 / Q8 common emitter also changes in the direction in which it also increases. That means it should be offset to the extent that the amount will increase + increase (I think it will not be a good thing). From the above discussion, Q11 certainly thought that there was a temperature compensation effect and remained without removal.
Two-stage differential amplifier circuit
2SA798G was used for the second different differential amplifier circuit. Here is that he is less cheaper than 2SA1349BL, because he has a low pressure resistance compared to 2SA1349BL. However, I will do 1 157 yen. 2SK389GR → 2SK30ATM-GR × 2 The difference in sound of 2 Sometimes I think that I do not know the difference in sound even if I change to 2SA798G → 2SA872AE x 2. Just because he has bought 10 2SA798G, he also uses 2SA798G.
Output drive stage
In the Unit 1 he was worried about the drive capabilities of the MOSFET, so this time, the constant current element that was at the load of the second stage differential amplifier circuit was removed, and I tried to put an output driver instead ( 2SA1360 / 2SC3423). Current of 10 [ma] or more flows in this drive stage, so it is connected to a power supply for output rather than a voltage for voltage amplification.
Unfortunately, the environment of the first unit is no longer (I gave up the speaker), but I can not compare it, but I have no impression that the sound drivers have been dramatically improved. But maybe it gets a little better? Well, that's the degree. However, since the various places are changed other than here, the influence on the sound driver's sound is unknown.
Anyway, it is nice that the heat generation of the power transformer for voltage amplification has decreased by removing the constant current device of the two-stage differential amplifier circuit. It may have the effect of lowering the set voltage to the regular 18 [V], but it is about to "warm" during operation.
Bias circuit
Although the story is back and forth, the output driver produces a voltage loss of about 1.2 [V], so it is necessary to set the driver + output stage bias voltage in consideration of this. Q19: 2N for the occurrence of bias voltageI am using 3906 (I would like to keep the stone I thought that I could not use it as it is. The stone used here simply causes a voltage drop of numbers [V], so if the current is also 10 [MA], it is sufficient (this time it will flow up to about 6 [MA]), unless there is a commitment I think that a small signal transistor can be used. When the temperature compensation of the output element is compensated, I think it is good to select the shape of the shape that is easy to attach to the output element or heat sink.
2N3906 was calculated by 0.7 [V] because the base emitter voltage is about 0.7 [V] at room temperature. Since the resistor R28 connected between the base and emitter is 1k [ω], a current of 0.7 [V] / 1000 = 0.7 [MA] flows to R28. The current flowing into the base of 2N3906 is micro, so ignore it, R27 also flows 0.7 [MA]. Since R27 is 6.8k [ω], it is calculated for the voltage of 0.7mA * 6800Ω = 4.76 [V] to both ends of R27. The sum of the base emitter voltage between the Base emitters of Q19 0.7 + 4.76 = 5.46 [V] is the voltage generated between the emitter collectors of Q19. The actual value should be slightly lower than 5.46 [V] because a little current flows on the base. This measured value was about 5.3 [V] at the time of power ON, and about 5.15 [V] during operation (the voltage between the base and emitter of the transistor decreases when the ambient temperature increases).
The voltage of 5.3 [V] generated in the bias circuit is to cause the MOSFET to flow idling current, which results from the output voltage. The bias transistor is connected under the + side output in the Collector of Q7, so the bias voltage affects the amplitude of the signal on the negative side. Why is it to affect the minus side, and the + side has a voltage loss already by the 2SA 798G common emitter resistance (measured value about 1.6 [V]). Since the + side already has this loss, adding 5.3 [V] losses, the + side becomes loss of 6.9 [V], and the signal + side is 18.2 [V] -6.9 [V] = 11.3 [V] -6.9 [V] = 11.3 [ V] will be headed. However, to output 10W, the peak voltage is required about 12.7 [V] on one side, so this is a calculation that can not be removed by 10W.
Therefore, it is still useless with a negative side without loss, but it is still 18.2-5.3 = 12.9 [V] and quite a bit. In fact, it may not come out 10W. It is good separately (do not output 10 W first).
If idling current is not flowed to the MOSFET, the bias voltage can be set lower lower, but if the MOSFET does not flow to some extent, it can be read from the data sheet that the linear area can not be used (the low current area is Characteristic curve tight). The current value of the measure that is considered to improve the linearity of the MOSFET is about 200 [ma]. This time, the bias voltage of 5.3 [V] has been set, but approximately 4 [V] is used as the drive stage, and about 4 [V] will be between the output MOSFET's gate. Assuming that this voltage equally applies to 2SJ200 and 2SK1529, approximately 2 [V]. The measured value of the MOSFET's idling current was about 220 [ma]. However, this value varies with the left and right channels, so if both are aligned with the same time, R28 is adjusted with fixed resistance + variable resistor. I hate to use variable resistors (because the aging deterioration is most intense in the part), I adjusted with a fixed resistor in the meantime.
Earth wiring path
It is also an earth wiring path that I felt more important than the difference in sound by the element. Ham noise comes out with a terrible route path. After that I hope you use a thick line as much as possible. When I moved in a test, I used a good thin wiring, but I felt that the bass was quite weak. I just noticed that I was noticed. But earth is a line where all the current join and returns, so I think it will be thin and "Hung clogging".
This time, the grounding route of ground was trial and error. But after all, I went to almost the same path as 1 unit. The center taps of two power supply transformers are connected to each other. The shorter wire is dropped from the ground (Earth written as "A" as "A" of the larger transformer. As described in the circuit diagram, the power amplification circuit ground is connected in the order of 0 [V] to 1 to 4, as described in the schematic. I think that it would be like this if the circuit is assembled to the board, but I think that it is not exactly the same. After that his 0 [V] terminal of the speaker was connected to both the "5" ground point and the center tap of "A". "A" did not feel particularly needed (or it may not be necessary), and if "5" does not connect to "5", ham noise has increased. When both "5" and "A" are connected to the speaker, the ground wiring forms a loop, but the increase in Ham Noise was not understood. I do not know if this is the best route (it is probably not the best), Ham Noise is the same level as one unit, and it is a level that can not be heard that the ear is not brought close to the speaker. It may be slightly heard that it is an efficient speaker. This time you can hear the sound of the power transformer for the output stage even more than Ham Noise.
others
R9, C14, C18, C19 is for phase compensation, especially when he does not have C18. If the prototype circuit is assembled with a breadboard, the contact connection may have a C component, which may move even if there is no phase compensation part, but if it actually solder, no matter thereof I think sure to ensure it. C18 capacitors were at least 100 [PF] on both TRs in Unit 1, but the Unit 2 is also operating stable as only 47 [PF] to remove the output. C18 feels like affecting the sound quality, so if you use TRs of 50 [PF] so as not to put this, you may want to make strange sounds with external C (It will be a sound of TR in place). Looking at web pages, it can be seen that the phase compensation capacitor corresponding to this C18 tends to use an expensive capacitor (dip mica). For example, in the audio circuit, an abnormally high capacitor or resistance is used, but even if there is a difference in sound due to that, I think (high part is a specific data that sound is good) Because I have never seen it). 2SA606 / 2SC959 just used to be 50 [PF], and it was commonly used by Mr. Kaneda's amp. Such circumstances and he have not been produced by 2SA606 / 2SC959, now it has become a very expensive TR, which is a very expensive TR. 2SA606 / 2SC959 had some father with some of the father (Kaneda amplifier made), and I have picked up, but I'm sorry, but I don't feel that HFE is all broken. However, is COB with a recent TR with 50 [PF]? . .
10 [Ω] and 1 [uh] at the output are the phase compensation of the wiring from the output terminal to the speaker, but because the effect is not understood well, it is bypassed with the second unit. Although it is judgment that it does not have to move, it is judged that the bypass can be removed immediately for when there is a problem such as oscillation. Because it is bypassed with a thin line, it may still work a little.
If you bring the speakers close to the output stage power transformer, cause electromagnetic interference and sound from the speaker (nearby if you are near ...). However, he is using the amplifier from about 50 [cm] speaker. EMCs and VCCIs are considered to be "not talking".
Backside of the second unit. It is dirty because it failed to mask the bottom painting, but it doesn't usually see it.
I'm a lithuning room (laughs). The source compared to PS, SB-Live! (PC), and iPod shuffle, he was the best iPod shuffle. Even in this setting, it is possible to play as much reproduction that I would warner to a different dimension. "
Our other product:
