Features

● Bandwidth: 350MHz-8P

● High-output current: ± 70ma

● conversion rate: 2100V/μs, 5VP-P

123] ● Differential gain/phase: 0.12%/0.05 °

● Low static current: ± 4mA

● Low input bias current: 1.2 μA

● Rising time : 1.9ns, 5VP-P

● Settlement time: 9ns, 0.1%

Application

● Broadcast/HD TV equipment

● High-speed digital communication

● Pulse/RF amplifier

● High -speed analog signal processing

● Line driver (50 , 75 ) Big electric device

● CRT output stage drive

● Active filter

Explanation

OPA623 is a current feedback computing amplifier, specially distinguished by high distortion It is designed with precision broadband systems such as video, radio frequency and intermediate frequency circuits, and communication equipment.

New circuit design, plus complex bipolar technology, has achieved obvious performance in single -chip integrated circuit technology.

The current feedback is optimized, with broadband bands, good pulse response, flatness flatness, low distortion, and work under low static current of ± 4mA.

It provides a large signal bandwidth of 350MHz under 2.8VP-P output voltage, and a conversion rate of 2100V/μs. At the 30MHz bandwidth of 0.05DB, the flatness flat makes it suitable for HDTV design. Another feature of the computing amplifier is its high output current ± 70mA, enabling it to use the amplifier to use the amplifier as a line driver in the video router, allocate amplifier, analog and digital communication device to connect 75 cable.

The power supply voltage of OPA623 is ± 5V, which is designed for extended industrial temperature range (-40 ° C to+85 ° C), and provides plastic SO-8 and 8-pin plastic impregnation packs.

pin configuration

Input protection

For MOSFET devices, prevent static electricity damage from electrostatic damage. The necessity has long been recognized by people, but all semiconductor devices should be protected by this potential destructive source. OPA623 integrates the ESD protection diode in the film, as shown in Figure 1. These diodes do not need outsideThe Ministry protects the diode, which will increase the capacitance and reduce the communication performance.

As shown in the figure, all input pins of OPA623 are internal protection of any power supply through a pair of back -to -back diode to prevent static discharge. When the input voltage exceeds any power of about 0.7V, these diode starts to turn on. This happens when the signal source still exists and the amplifier loses the power. The diode can usually bear the continuous current of 30mA without damage. However, in order to ensure long -term reliability, the diode current should be limited to about 10mA as much as possible from the outside.

The design of the internal protection diode can withstand 2.5kV (using human models) and will provide sufficient ESD protection for most normal operating procedures. However, static damage may cause subtle changes in the input characteristics of the amplifier, not necessarily damaged the device. In the precision amplifier, this change will significantly reduce the offset and drift. Therefore, when dealing with OPA623, it is strongly recommended to take anti -static measures.

Typical performance curve

VCC u003d ± 5VDC, RL u003d 100 , IQ u003d ± 4mA, RIN u003d 150 , tamb u003d +25 ° C, unless otherwise explained.

Performance discussion OPA623 requires very low static static static static Power achieves its excellent communication performance by using current feedback topology. This broadband single -piece computing amplifier is designed for the gain of up to 20V/V. Power and cost here are the primary consideration. OPA623 consumes only 40MW under the power supply of ± 5V, but maintains a large signal bandwidth of 350MHz under VOUT u003d 2.8VP-P and 2100V/μs conversion rate. Thanks to the current feedback structure, OPA623 provides a stable operation without compensating capacitors, even at the unit gain.

OPA623 has 0.12%and 0.05 ° low difference in gain and phase errors at 4.43MHz to meet the performance and cost requirements of large -capacity broadcasting and HDTV applications.

OPA623's large signal bandwidth, high conversion rate, excellent pulse response and high drive capabilities are very suitable for broadband RGB video applications, RF instruments, and even high -speed digital communication systems.

For most circuit configurations, OPA623 current feedback operations amplifier can be treated like a traditional operational amplifier. Like the voltage feedback computing amplifier, the feedback network -controlled closed -loop gain connected to the inverter input. However, for the current feedback computing amplifier, the impedance of the feedback network also controls the opening gain and frequency response. OptionalSelect the feedback resistance value to provide almost constant closed -loop bandwidth within a wide range of gain, and make a flat gain adjustment with the frequency.

Explanation

Broadband computing cross -guidance amplifier (OTA) and output buffer are the main modules of current feedback computing amplifier. The simplified circuit diagram is shown in Figure 2. OTA consists of a complementary unit gain amplifier and a subsequent current mirror. The input buffer is connected to the input end of the computing amplifier. The voltage on the high impedance+IN terminal is transmitted to the -N terminal with low impedance. The current mirror will reflect the high impeda OTA output that flows into or out of the+IN terminal at a fixed ratio to the high impedance OTA output directly connected to the complementary output buffer. It is designed to drive low impedance transmission lines or loads. The output of the buffer is not limited or protected by current.

As shown in Figure 3, the feedback of the current form is applied to the low impedance inverter input end through R2, and the size of the R2 | R1 determines the opening of the operation amplifier. The ring is increasing.

The hybrid model shown in FIG. 4 describes the communication behavior of a broadband current feedback amplifier without internal compensation. The opening frequency response of the various R2 values u200bu200bshown in Figure 5 is determined by two time constants. The component R and C between the current source output and the output buffer constitute the open -loop polar TC that dominates. The signal delay time of modeling in the output buffer combines several small phase movement time constant and delay time. They are distributed in the amplifier and also exist in the feedback circuit. As shown in Figure 5, increasing R2 | | R1 will lead to a decrease in the opening gain. The ratio of two time constants TC and TD also determines the product of the best closed frequency response GOL GCL:

However Two time constant TC and TD are fixed by designed by computing amplifiers. However, the external changes in the feedback circuit R2 | | R1 allows the changes of the open -loop gain GOL and a closed -loop gain GCL. This keeps the product GOL*GCL constant, which is the theoretical condition for the best flat frequency response.

When the driver is highly capacity, this change may be beneficial. Through the external settings, the circuit gain can also be optimized to a large -scale capacitance load. As shown in Figure 7, the closed -loop gain is+2V/V, and the capacitance load is as high as 47PF.

It should be noted here that higher -opening gain (generated by lower feedback resistors) will also produce lower distortion.

Through the external control of the operating characteristics of the computing amplifier, the dynamic behavior can be adjusted according to the specific application requirements, and the opening of the loop gain selection provides almost a constant closed -loop bandwidth. As shown in Figure 6, the various gains have a variety of gains. Best flat frequency response. This behavior is the opposite of the operation of the operation of the internal compensation stable unit, and the bandwidth and closed -loop gain of the latter become inverse.Compared with the high output level and high gain, the bandwidth and conversion rate will be greatly limited.

Generally speaking, lower feedback resistors will generate wider bandwidth, more frequency response peaks and more pulse response overlap. Higher feedback resistance can cause override response, almost or no peak and overwhelming.

The component pipe foot and layout capacitance, as well as the trace lines and wire plates from the resonant IC circuit, can cause a hundred -meter oscillating. This very high frequency oscillation can cause excessive increases of power currents, thereby destroying the equipment.

Resistance (100 to 250 ) connects, close to high impedance, and input damping LC circuits non -inverse phase to generate safe operation.

Thermal factors

OPA623 does not need a radiator in most environments. However, the use of heat sinks will reduce the internal heat rise, which makes it colder and more reliable. Under extreme temperature and full load conditions, radiator is needed. The power consumption in PDL is given by PDL, and the power consumption is represented by PDL (PDL is static power consumption). Although PDQ is very low (40mW at VCC u003d ± 5V), be careful when applying signals. For high -speed computing amplifiers, a more accurate method for determining power consumption is to measure the average static current under several typical load conditions. The power consumption of OPA623 is affected by the signal type, plus signal frequency, output voltage, load resistance and signal conversion duplication rate. Figure 8 shows the relationship between the average power current and the sine wave frequency under different output voltage. Figure 9 shows the relationship between the average power current and the repetitive frequency of the application square wave signal.

Circuit layout

The physical layout of the printing circuit board has the high frequency performance of the computing amplifier OPA623. A big influence. Here are some suggestions. Oscillation, ringing, low bandwidth, low bandwidth and peak values u200bu200bare typical problems that plague high -speed components.

Add a 100Ω resistor to the high resistance input terminal to prevent excessive input -end oscillation. The power supply is very close to the device pin. The pyrone -type capacitors (about 2.2 μF) are used in parallel with 470pf ceramic sheet capacitors. It is recommended to use the surface paste because of their low inductance. Although OPA623 works under low static current, during the steep transition, high -charged discharge current will flow. PC board traces of the power cord should be wide to reduce impedance and inductance.

Make a short low -induced trajectory. The entire physical circuit should be as small as possible.

use low impedance ground planes on the side of the component to ensure that there are low impedance grounding in the entire layout.Such as the input terminal of the amplifier, these nodes are sensitive to the strange capacitors.

do not recommend the use of sockets because they increase significant inductance and parasitic capacitors.

use low inductance and surface installation components. The circuit using OPA623AU surface installation element will provide the best communication performance.

Insert prototype board and winding board cannot work properly. A clean layout is necessary to use radio frequency technology. There is no shortcut.

make the feedback trajectory as short as possible. Reverse input is sensitive to the strange capacitance that causes the peak in response to the frequency. The strange capacitors of the inverting input end increase the gain under high frequency.

Application information

Accurate pulse response and high conversion rate enables OPA623 to be used for digital communication systems. Figure 12 shows the circuit schematic diagram of the output amplifier with a gain of+2V/V. The amplifier can drive 75 coaxial cable with a high -speed data stream of 140Mbit/s. Figure 13 is a binary 0, Figure 14 is binary 1, which shows the pulse shielding of the CCITT suggestion G.703 and the corresponding pulse of OPA623. The signal coding of the file rate of 139.264mbit/s is CMI, the signal amplitude is 1VP-P, and the amplitude limit is ± 11DB. Of course, OPA623 can also be used for HDB3 encoded 34MBIT/S, 155Mbit/S, STM-1 and 155Mbit/SBB-ISDN transmission system.

Figure 13: According to CCITT's suggestion G.703, which corresponds to the pulse shield of binary 0.

Note: (1), the maximum ""steady -state"" amplitude shall not exceed 0.55V limit. If the over -adjustment of other transients does not exceed 0.05V, the overrun and other transients are allowed to fall into the dotted area, and the range is 0.55V and 0.6V. It is studying whether it is possible to relax the over -adjusting volume that may exceed the steady level.

(2), all measurements for using these masks should be used to communicate signal AC into a oscilloscope input terminal that is not less than 0.01 μF. The nominal zero level of the two masks should be aligned with the trajectory of the oscilloscope without input signals. Then apply the signal to adjust the vertical position of the trace line to meet the limit of the mask. For the two masks, any such adjustment should be the same and should not exceed ± 0.05V. It can be checked by removing the input signal again and verifying the trace line in the ± 0.05V range of the zero level.

(3), each pulse in the coding pulse sequence should meet the restrictions of the relevant mask, and it has nothing to do with the state of the front pulse and the posterior pulse. For actual verification, such asThe 139264kHz timing signal that is associated with the interface signal source is available, and it is preferred to use it as a timing benchmark for oscilloscope. Otherwise, you can test whether the relevant mask can be tested through the ALL-0S and ALL-1S signals. (In practice, the signal may contain a frame of each REC.G.751.) (4), for these masks, the ups and duration of the rise and attenuation should be at -0.4V and 0.4V of 0.4V Measurement, and should not exceed 2NS.

Figure 14: According to CCITT's suggestion G.703, the pulse shielding corresponding to binary 1.

Note: (1) The maximum amplitude of ""steady -state"" should not exceed 0.55V limit. Allowing over -rushing and other transients to fall into the dotted area with a boundary of 0.55V and 0.6V, provided that they do not exceed the steady -state level 0.05V. The possibility of relaxation exceeding the steady -state level is being studied.

(2), all measurements for using these masks should be used to communicate signal AC into a oscilloscope input terminal that is not less than 0.01 μF. The nominal zero level of the two masks should be aligned with the trajectory of the oscilloscope without input signals. Then apply the signal to adjust the vertical position of the trace line to meet the limit of the mask. For the two masks, any such adjustment should be the same and should not exceed ± 0.05V. It can be checked by removing the input signal again and verifying the trace line in the ± 0.05V range of the zero level.

(3), each pulse in the coding pulse sequence should meet the restrictions of the relevant mask, and it has nothing to do with the state of the front pulse and the posterior pulse. For actual verification, if the 139264kHz timing signal is available to the interface signal source is available, it is preferred to use it as a timing benchmark for oscilloscope. Otherwise, you can test whether the relevant mask can be tested through the ALL-0S and ALL-1S signals. (In practice, the signal may contain a frame of each REC.G.751.)

(4), for these masks, the ups and duration of the rise and attenuation should be at -0.4V and 0.4V of 0.4V Measurement, and should not exceed 2NS.

(5), the reverse pulse will have the same characteristics. Note that the zero -level timer tolerance of negative jump and positive transition is ± 0.1ns and ± 0.5N, respectively.