"Keywords: IRS2092, D power amplifier circuit diagram
Be
1) For overcurrent protection settings:
Current protection threshold: The current flowing through the FET includes an audio output and an LPF filtering network current, so the actual requirement value is higher than the audio current peak. I prefer to choose this: the maximum output current X1.5; 2) RDSON selection: can be selected according to the actual maximum operating temperature, can usually take the RDSON value at 85C; 3) With the current and RDSON, you can Select the protection value of the IC. It should be noted here that the FET switching wave in large current should be a very clean square, and if there is a burr, the gate resistance and the optimized decoupling structure can be appropriately increased. For DC output issues: When it is overloaded; the power supply voltage dries, but due to the non-symmetry of the transformer or load, the positive and negative power supply voltage is inconsistent, resulting in a half-wave output of the power amplifier. When this problem occurs; is usually not a power amplifier problem, you can give a slight increase in the supply voltage to increase the pond capacitor.
2) Under the general ideal engineering conditions, the shortest dead time of the half-bridge work is, the better, minimum can be set to 10 nanoseconds
3) IRS2092S should pay attention to the full-frequency power generation, it is best to select a relatively good linearity, and the modulation frequency is selected at 300 ~ 400 kHz, and the value of the L / C of the LC filter is appropriately matched, which can solve the problem. . It is recommended to select 18 ~ 22μH. If the flatness is to be further improved, L can take 10 μH, and the modulation frequency is increased to about 400 kHz.
4) PWM modulation and 咔哒 suppression problem: The excess process of the circuit does not match, the front stage power supply voltage land is caused
The voltage flowing is enlarged by the rear stage. Adjust the problem to solve the problem.
Also: IRS2092 / s has been added to the IC to disable the "click" function. There is generally no additional circuitry.
5) How to choose from MOS, gate resistance of different QG, QG Too 2092 fever is very fast: can reduce the temperature rise of IRS2092 with reduction or increase grid resistance, general; QG and switching frequency is not the main IC heat Cause, fever may more derived from the voltage overshoot generated by the switch, improve the PCB wiring and decoupling to suppress IC heat. For 300W rated output power, IRFI4020-117P will be better.
6) Make 500W power amplifier can be used with IRS2092 with IRF6785MTRPBF, and the PCB design is reasonable; THD + N can reach a level of 0.02%. IRFB4227 can also be used in this wide power, but the effect will be different.
7) For the fever of MOS tubes, MOS pipe fever is caused by a variety of reasons, first; voltage current specifications should be appropriate, excessive or too small, will increase heat. Second; use dedicated drive ICs and MOSFETs. Again; pay attention to the PCB design layout, select the appropriate installation location and specifications.
8) It is recommended not to expand when realizing greater power output with IRS2092, otherwise it is difficult to ensure accurate matching. IRS2092 is enough to make a thousand wars.
9) Class D power amplifier is because the LC filter and regulator are low-pass structure, and they have a certain attenuation of high-frequency signals, and they will hear voids. Can improve the parameters of these two parts to achieve the required bandwidth
10) IR's latest MP4, MP5 and MP7 these three DEMO versions, their frequency bands can reach 24K, under 4 ohm conditions, widely widened than the earliest demo, if they bring 8 ohm load, actually bandwidth It is far more than 48K, mainly the bandwidth of an LC filter is suppressed internally. In order to ensure the sufficient width of the audio bandwidth, it is necessary to use a relatively low requirements of its own power consumption, with a 22 michen-inductance. If you need to widen the bandwidth, just reduce this inductance value, such as 16 to 18 michen Its bandwidth can be widened, the sound amplifier itself reaches 60K, there is no problem, tested, the power amplifier can bring 3OHM or 2OHM load due to the excellent protection points and MOSFETs in the presentation board, and the MOSFET is set according to 4OHM load settings, It does not double the output power when two ohms. In addition; this circuit is currently very stable.
11) How much influence on the death time for THD?
The influence of the dead zone to THD is very big. It is actually equivalent to the wave of the string wave. If it is harmonic test, the increase is mainly three harmonics, and it is especially true for the impact of the tone. Big, we have tested, if there is a 50 nanosecond time, if no feedback is made, the value of THD can reach 2.8%. If you listen, it will feel very dry, so it is said that the dead time is THD is very important, so it is emphasized that a little bit of interlock matching and MOSFET's QGD, and MOSFET lead inductance.
12) IR2092 Carrier frequency calculation and testing at the time of beef gun:
The IR2092 is determined by the two-mesh feedback capacitance, which determines the input impedance and output impedance. These two capacitors determine its frequency band, generally, we recommend that the modulation frequency is generally limited to 48K to 100K in the bass amplifier. In fact, it is actually the subwoofer, its output frequency is usually not more than 600 Hz, 100K modulation frequency to 600 Hz is already enough. A relatively low modulation frequency can allow MOSFETs to be relatively slow to obtain lower noise, and the overall switch loss can be reduced, and the efficiency is increased to 98 percent, because the modulation frequency is lowered, and the output inductance Usually increased, according to our current experience, if it is a bass of 1 ohm, then the filtering inductance can generally choose about 60 微亨, then the bass is relatively thick.
13). Avoid noise interference mainly,
On the one hand, it is a dedicated drive and dedicated MOSFET so that its noise will be very low, our current IraudioMP4 and MP5 Demo version can directly receive tone without interference. This may be the only solution in the industry. Second, our PCB design is very distinctive, and the components line connection is very professional. It is one of the methods of solving the previous noise. Overall, the design requirements of class D power are high, the main reason is internal frequency. A relatively high, the MOSFET current will be modulated, and there is a very high RF component in the PCB. High frequency noise if diffuse; will modulate the operational amplifier to generate some background noise, which requires many techniques to solve, on the other hand, 2092 integrates high (angle) frequency noise immunity amplifier. So there are many problems in the D-amplifier, many theories have exceeded the knowledge category of traditional AB amplifiers. Class D amplifier design is very efficient, and the AB class is more about signal fidelity, thereby; increasing a lot of requirements, designed D-class among the current engineers; encountering more challenges But this will be the achievement of higher efficiency and better sound quality.
14) D-class power is a system, except for the effector, it is necessary to drive, capacitive inductance, etc., because it is working under high power conditions, its material, shielding and placement methods are very particular, first Inspiration We recommend as a shielded structure as possible, such as using PM or PQ structure, of course, is the best, even if it uses these structures, it will find that the radiation direction of the magnetic field is different, some is the front, some is downward These directions must be paid in the layout direction of the signal lines of the PCB, so that there is no magnetic field interference into the system, or the interference is minimized, so the system can achieve the best benefit.
15) About dead time selection: dead time settings and customer chosen MOSFET parameters are closely related, the reasons are two aspects, on the one hand, the MOSFET's MOSFET is different, even if the MOSFET drive resistors, but due to The QGD is different. Dossally in varying degrees, the MOSFET is turned off, so the rendering speed of the dead zone is different, and the recovery speed of each MOSFET parasitic diode is different, the equivalent payment time is not equal, setting the dead time must be longer than MOSFET The recovery time of parasitic diodes will form a relatively serious short circuit, and the power amplifier is very sensitive to such noise. Therefore, note that MOSFET's parasitic diode must be Schottky, otherwise you will find that its heat generation will be doubled, and the tone will be much harder.
16) About Schottky diode, the market is not an ideal Schottky, the ideal Schottky is required to drop 0.25V, and only 15V Schottji is close to this value. Whether it is, however, Schottky is not advisable in parallel on the MOSFET. The main reason is that MOSFET is an individual structure. Its lead inductance is usually 13nh, and the same current Schottky is usually T-220 package, The lead inductance value is close. In order to cut the current into the Schottky tube next to the MSI (MOSFET parasitic diode pressure drop is usually around 0.6 volts), all lead sensing voltages cannot exceed 0.6 volts, otherwise; parasitic diode Will open. MOSFET and Schottky are connected in parallel, from MOSFET dies, lead inductors are connected in series. For an inductance of more than 20nH, the induced voltage cannot exceed 0.6, it allows the current very very small, so when the power is slightly larger, Schottky will basically disappear, so the principle is It is not recommended to be connected to the MOSFET.
17) What is the main difference between the audio special MOSFET and the ordinary switching MOSFET?
The difference between the audio-specific MOSFET and the ordinary switch type MOSFET is three aspects. The first point is that the MOSFET for the sound is Schottky. Since its modulation frequency reaches 400K, the voltage is basically 200 volts, in this The voltage is hard switch under the voltage, and the time period has a continuation request, which is far more than the normal switching power supply to the MOSFET, so its MOSFET must be a Schottky structure. The second point is that the value of the QGD of the MOSFET is also different. The value of the special MSO is optimized, the QGD value can be reduced to ensure the output pulse width accuracy, and our ordinary switch type MOSFET, it mainly emphasizes Comprehensive value of cost and switching loss. It does not consider the loss of pulse width, the normal switch type MOSFET output is the main PWM signal, which does not do any processing, and the dedicated audio MOSFET requires the frequency band of 24K, so it's a response is very Fast. Third, the gate parasics. The MOSFET of the sound of the sound is higher than the MOSFET of the pulse. The dedicated MOSFET gate parasic internal resistance is often around 1 ohm, and this value is far less than the drive resistor, and the value of the switch type is usually more than a dozen Ohm to dozens Ohm, dedicated audio MOSFET, after controlling this value, can effectively ensure the accuracy of the pulse width, the difference between the MOSFET and the ordinary MOSFET is very, interested, compare the test results of the two Mosses, It will find that it is very obvious that it is from power consumption or from the sound quality.
18) Why is the bus voltage rise in the half-bridge circuit structure design?
The speaker is a special thing, looks; it is an ohmic, but due to its own inherent frequency characteristics, the frequency / power range of the broad spectrum is in the range of the mountainous velocity. The frequency of the output is higher than its own frequency, which exhibits a capacitive impedance; when it is induced when the frequency resonance point is low. For a horn with multiple resonance points; and is a non-complete elastic variant, the impedance changes are very complicated. These sensibility or capacitive properties indicate that the speaker stores or releases or unscrews or small energy at different times, which will form an energy cycle in the electrical loop, that is, the mechanical kinetic energy or electromagnetic energy inside the speaker will be modulated by the D class. Recycling in the power supply or filled in the speaker, this time it will find that the power supply capacitor voltage will rise or decline, and when the speaker is output, it will put the energy in the capacitor into the speaker into the speaker, as an audio The power, this time it will find the power supply is fluctuating. This is also the reason why the efficiency will be very high, which will reclaim the energy of the uncharacted effective mechanical kinetic energy and some reactive impedance components, and the AB class It is impossible to recover, as this, the damping coefficient and efficiency of the D class power are far more than the AB class. Under normal conditions, if the D class work is good, its damping coefficient can reach 110dB.
19) How does the output filter induction of Class D? What is the reason for the output filter inductance? How to solve?
The material of this type of amplifier currently has two more uses, and one is called the iron powder material, which is actually a superfine microfluceration material, mixed cast the formed ring with epoxy resin. The other is a common domestic R2KBD material. In principle, the performance of this material is more than PC40. The main reason is that the magnetic intensity of the RKBD material can reach 4,800, while the PC40 can only reach around 3,500. R2kbd material caster temperature reaches 250 ゜ C, which is much higher than PC40, from anti-frequency saturation, and reliability, domestic materials are better than imported materials.
20) What is the relationship between the driving current of the large current field effect tube and IR?
When the MOSFET is open, the most important thing is that there is a power-up mode to start. When we enter the voltage to the gate, it is like we plus the capacitor, so we open MOSFET, small current MOSFET The gate capacitance is relatively large, which affects the performance of the power release, affects the sound quality, so the large current field effect tube should take into account the driving current, and MOSFETs with appropriate QG.
IRS2092 typical application circuit diagram "
Our other product:
