-
2022-09-21 17:24:28
AD8305 is 100 decibel range (10 mAh to 1 mAh) to a number converter
Features
Optimized optical diode interface; measurement for more than 50 years; 0.1 decibel from 10 mAh to 1 mAh; single power or dual power operation (total voltage 3V to 12V) Full pair ratio capacity; nominal slope is 10 MV/DB (200 MV/DECADE); optional adjustment of the nominal cutting distance of 1 na (set up from the external resistance) slope and cutting distance; Quick response time of current level; miniature 16 -core chip scale (LFCSP 3 mm × 3 mm); low power consumption: ~ 5MA static current.
Application
Optical power measurement; wide range of base bands compressed by numbers; current voltage ratio measurement; light absorption measurement.
General description
AD8305 [1] is a cheap micro -pairing converter used to determine the optical power in the fiber system. It uses an advanced implementation of classic transliteration (based on connection) technology, providing a large dynamic range in a multi -purpose and easy -to -use form. The single power supply voltage is sufficient between 3V and 12V; you can choose to use dual power. Low static current (usually 5 mAh) allows use in battery power supply.
The input current IPD of 10 mAh to 1 mA on the input pipe's foot is the optimal proportion of NPN transistor's collector current, which converts the current into a voltage with precise number relationship (VBE (VBE To. The second converter is used to process the reference current (Iref) applied to PIN Iref. These input nodes are slightly higher than the ground (0.5 V). This is usually applicable to the optoelectronic diode application that does not require ground. Similarly, it is easy to consider this bias voltage when Iref is generated. The output of the front end can be obtained on the PIN VLOG.
The basis of the output is 200 MV/Decade (10 millivolta/decibels) in name. Therefore, the 100dB range corresponds to the output change of 1V. When the voltage (or caps output) is applied to ADC that allows external reference voltage, the 2.5V AD8305 voltage at PIN VREF can use the output output to improve the zoom accuracy. The appropriate ADC includes AD7810 (serial 10 digits), AD7823 (serial 8 -bit) and AD7813 (parallel, 8 or 10 bits). For other values of the slope, it can be provided using a simple external resistance network.
The number of numbers (also known as the reference current) is nominally positioned by the use of 10 μA current IRef generated from the outside. Iref is provided by a 200 kΩ resistor. V) and reference to input Iref (voltage 0.5 V). By changing the resistance, the cutting distance can be adjusted in a wide range. AD8305 can also be compared with numbersMolecular currents are applied to input, and the parent current is applied to Iref.
Provide a buffer amplifier for driving a large amount of load, which is used to increase the basic slope of 10mv/dB to a higher value, as a precision comparator (threshold detector), or to achieve low low Pass filter. The rail -to -rail passing level can be swinging within the 100mV range of the positive and negative power rail, and the peak current source capacity is 25mA.
Small signal bandwidth decreases with the decrease in the current level. The increase in low -frequency noise spectrum density increases and increases, which is a basic aspect of the hyper -linear pair converter. At the 10NA level, the bandwidth of the AD8305 is about 50kHz, and it increases proportionally to the IPD, with a maximum value of about 15MHz. Using the buffer ballast can realize the low -pass filter of the high tritoly, solve the problem of increasing noise level at low current.
AD8305 can be used in the 16-drawing LFCSP packaging, and is specified at the temperature of -40 ° C to+85 ° C.
Typical performance features
v 5V, v 0V, R 200kΩ, T 25 ° C, unless otherwise explained.
The overall structure
AD8305 solves various interface conditions to meet the needs of the optical fiber monitoring system, and it is also useful in many non -optical applications. These notes explain the structure of this unique translation of the number of amplifiers. Figure 33 shows the simplified schematic diagram of the key element.
Optical diode current I received at the PIN input terminal. Due to the low bias voltage of the JFET computing amplifier, the voltage of the node is basically equal to the voltage on the two adjacent protective tube feet (VSUM and Iref). The transistor Q1 converts the input current I to the corresponding number of the number of numbers, such as equal form 1. In the case of the normal single power supply voltage, the limited positive value of the Q1's collector should be partially voltage. The interior is set to 0.5 V, which is one -fifth of the 2.5 V reference voltage appearing on the pin VREF. The resistance of the VSUM pin is 16 kΩ in nominal; the voltage is not a general partial voltage source.
AD8305 also supports the use of optional negative power voltage V at the pin VNEG. When V is -0.5V or more negative, VSUM can be grounded; therefore, inpt and Iref assume the potential. This allows the operation of a voltage input to the number converter, including one or two inputs in one or two inputs. Note that the resistance is required to set the I to keep the intercept value. It should also be noted that the set electrode-transmitting pole voltage of Q1 and Q2 is now full V, and the impact caused by self-heating will be greaterly losing.Error errors are caused under the current.Compared the reference v of Q1's input related V with the reference V of the second crystal tube Q2 working under I. This is generated outside, with a recommended value of 10 μA. However, other values within decades can be used when the legal consistency decreases slightly (see Figure 3). 理论BJT(双极结晶体管)的基极发射极电压可以用方程式1表示,它立即显示了其基本对数性质:[123 ]
Among them: IC is its collector current, which is a fixed-standard current, usually only 10-17 A.
KT/Q is a thermal voltage, which is directly proportional to the absolute temperature (PTAT), with 25.85mV at 300K.
The current I has never been accurately defined, and it shows a stronger temperature dependence, which changes about one billion times between -35 ° C and+85 ° C. Therefore, to use BJT as an accurate number element, these two temperature dependence must be eliminated.
A pair of BJT (one working under the optoelectronic diode current I, and the other work under the current I) can be written between the basis transmission pole voltage:
Therefore, the uncertainty and temperature -related saturated current IS in Formula 1 has been eliminated. In order to eliminate the temperature changes of KT/Q, this differential voltage is basically simulated. In fact, it puts a variable in the equal form 2. The output of this process also includes the conversion from the voltage mode to the current mode. It is a middle temperature correction current:
Among them, i is an accurate, stable temperature stable target setting standard Current, it determines the slope of the function (the current changes in each ten years). For AD8305, i is 44 μA, so all the values of i and I have nothing to do with the temperature of 44 mAh/ten years. The current then converts the output of the voltage mode, V, and the label is 200 millivolves/ten years.
Obviously, for i i, the output should be zero. For smaller input current values, the output should be negative. To avoid this, I need I need to be as smaller as possible. However, this is unrealisticly using such a small reference current to 1 nail. Therefore, when the PIN VRDZ is directly connected to the VREF, the offset voltage is added to V to move it up 0.8V. This will move the cut -off to the left for 40 years, from 10 μA to 1 NA:
Among them, i is the operation value of the intercept. To disable this offset, the pipe foot VRDZ should be grounded, and then only i is i. Because the value of i lt; I will cause negative VLOG, a negative power supply with sufficient value is required to adapt to this (see the negative power supply part).
Voltage V is generated by applying I to the internal resistance of 4.55 kΩ. The internal resistance is combined by parallel combination between parallel combination between the 6.69 kΩ resistor and the ground and the 14.2 kΩ resistor and the VRDZ pin. When the vlog tube foot is removed and the cut -off is re -positioned by the ground VRDZ, the output current, i generates voltage at the foot of the vlog tube
where v 200 mv// Ten years, or 10 MV/DB. Please note that any resistance load on VLOG will reduce this slope, and due to the variability of the resistance on the ship, it will also cause the overall marking uncertainty. Therefore, this approach is not recommended.
When VLOG uses dual power (V and V), it may also swing in the ground. When V -0.5V or larger, input pins and Iref can now be positioned on the ground by simply grounding vsum.
Management interceptor and slope
When a single power supply is used, VRDZ should be directly connected to VREF to allow the input current range throughout the five years. As mentioned earlier, this will introduce a 0.8V accurate offset voltage at the foot of the VLOG tube, which is equivalent to 40 years, which generates a number of conversion functions, which can be written:
[ 123] Therefore, the effective interception current I is only one -tenth of i. When the recommendation value I 10 mAh is used, it corresponds to 1 millowte.
A resistance can be reduced by installing a resistor on the vlog tube. In view of the incorrect proportion of the on -board resistor to the additional resistance, it is strongly recommended not to do this. In addition, few people want to reduce the basic sloping rate of 10 millivoli/decibels; if necessary, it should be performed under the low impedance output of the buffer to avoid this disorder and allow the use of higher slope.
AD8305 buffer is essentially a non -promised computing amplifier, with rail -to -orbit output width, good load driving capacity, and unit gain bandwidth greater than 12MHz. In addition to allowing the introduction of gain, the standard feedback network is used and the slope voltage V is increased. The buffer can also be used to achieve multi -polar low -pass filter, threshold detector and various other functions. More details about these can be found in the AD8304 data table.
Response time and noise considerationThe response time and output noise of AD8305 are basically a function of signal current I. For small current, bandwidth is proportional to i, as shown in Figure 15. The output low -frequency voltage noise spectrum density is the function of i (Figure 17), which will increase for smaller I. For detailed information about the noise and bandwidth performance of the cross -line amplifier, see the AD8304 data table.
Power order
Some applications may lead to large input signal current ( gt; 1 mAh) before AD8305. In this caseBelow, it is recommended to achieve power sorting, so that the AD8305 is powered on before the photoelectric diode or current source.
In applications that cannot be achieved in power sorting, VSUM should be driven from the outside. In applications that are not easy to obtain in low impedance dual source, the circuit shown in Figure 34 can be used.
The 2N2907 crystal tube used in FIG. 34 is a common PNP -type switch transistor. The choice of R and R makes the voltage of the bottom of the transistor ~ 0.5V.
Generally, V × [R/(R+R)] should be equal to about 0.5 V. Set R 5 kΩ and R 1 kΩ, which will cause the additional static current of 3 V power to 500 μA under normal operation. By selecting a higher crystal tube than 2N2907, a larger resistance value can be used for the compressor network.
Given a typical 0.7V voltage. When AD8305 is closed and a large input signal is being applied, the voltage at VSUM is ~ 1.2V. Once the AD8305 is powered on, the voltage at VSUUN is pulled down to its nominal value 0.5 V. The circuit in FIG. 34 Tests the positive electrode power supply of 3 V to 5 V within the entire temperature range of AD8305. C, R are the 11 -in -compensation network and C are the barrier -based containers recommended on VSUM.
If the number of board space restricting the number of external circuits AD8305 can eliminate the transistor in Figure 34 and connect the resistor division directly to VSUM. In this case, the bias voltage at VSUM and INPT is set by the resistance value selected for the distributor instead of the internal bias settings of the AD8305.Application
AD8305 is easy to use in optical monitoring systems, and in similar cases, the wide range current will be converted to its equivalent current, which is expressed in decibels. The basic connection of measuring a single current input is shown in Figure 35, which also includes various non -necessary components.
The 2V voltage difference between VREF and INPT pin and the external 200 KΩ resistor R provided 10 μA reference current I to the pin IREF. Connecting the pin VRDZ to the VREF can increase the voltage at VLOG by 0.8V, effectively reduce the interception current i by 104 times to position it at 1NA. For i, other values that can be used widely can be used from less than 100 na to greater than 1 ma. The impact of these changes is shown in Figure 5.
In the estimated stability of the intercept, any temperature change in R must be considered. In addition, when using a very low I value, the total noise increases. In the fixed cut -off application, the use of large reference current is almost not good, because this will only compress the low -current end of the dynamic range from a single power operation, which shows 5 V. Here is 5 V. It is recommended to be in vsum and groundUse a capacitor to minimize the noise on this node and help provide a clean reference current.
Since the basic labeling of VLOG is 0.2V/DECADE, the oscillation of the cushioning output end 4V corresponds to 20 years, so it is usually necessary to improve the slope to better use the rail voltage range. To facilitate explanation, the circuit in FIG. 35 provides a total slope of 0.5 V/ten years (25 MV/DB). Therefore, using i 10 μA, V running from 0.2 V at the time of I 10 NA to 1.4 V when I 1 MA, and the buffer output runs from 0.5 V to 3.5 V, corresponding to the dynamic range of 120 dB (electricity, that is, that is, that is, 60 DB light power).
The optional electric container from VLOG to grounding forms a single low -pass filter combination with the 4.55 kΩ resistor of the pin. For example, using 10NF C, -3DB angle frequency is 3.5kHz. This filter is useful in minimizing output noise, especially when I was very young. The multi -pole filter is more effective in reducing total noise; the AD8304 data table provides examples.The dynamic response of the entire input system was affected by the external RC network of two input terminals (INPT, Iref) ground. These are necessary to stabilize the input system throughout the current range. Due to the extensive changes in frequent, the bandwidth changes with the change of the input current. The RC network adds one zero to the input system to ensure the stability of the entire input current level. The network value shown in Figure 35 is usually sufficient, but when the optical diode capacitor is high, some experiments may be required.
Although the two currents are similar, they need to be careful when the current ( lt; 100 na) and temperature ( lt; 0 ° C) need to be used under the operation. It is recommended to modify the RC network to 4.7 nf and 2 kΩ to measure 10 NA at -40 ° C. By checking the transient response in the representative current level, the capacitor value can be adjusted to provide a rapid rise and decrease time with acceptable settlement. To fine -tune the network, the resistance value should be adjusted.
Calibration
The nominal slope and cutting of AD8305 are 200 mv/decate and 1 na, respectively. These values are unspeakable, and only the slope may change as high as 7.5%with the temperature. Therefore, it is recommended to perform simple calibration to improve accuracy.
FIG. 36 shows the improvement of the accuracy when using two -point calibration method. To perform this calibration, two known current I and I should be applied within the scope of linear work from 10 mAh to 1 mA. The output obtained, V and V, and calculating slope m and intercept B are calculated separately, respectively.
Use two known optical power P and P to perform the same calibration. This allows calibration to calibrate the entire measurement system and provide incidents at the same timeThe simplified relationship between light power and V voltage.
The unregulated error line in FIG. 36 is generated when the slope of the measuring output is 200 MV/DECADE, and in fact it is 194 MV/Decade Essence Correction of this difference reduces the measurement error by 3 decibels.
Most applications using negative power supply
AD8305 only requires a single power supply from 3.0V to 5.5V. However, in order to provide more generalability, dual power can be used, as shown in Figure 37.
When the input transistor (q1 in Figure 33) has sufficient negative bias pressure on its emission pole, the use permit of the negative power V allows the node to be placed on the nodes of the conformity of the power V. On the horizon. When V -0.5 V, the default conditions of Q1 and Q2 are the same when the vsum is grounded. This deviation does not require accurate, and you can use the source of unclear definition.
However, the source needs to support static current and INPT and Iref signal current. For example, the use of a positive bias of about 0.7V or the Schartki potential voltage with a slightly higher than 0.5V may be convenient. The effect of power on the range and accuracy of the power is shown in Figure 10.
In the case of harmonious node ground, AD8305 is now available as a voltage input pair amplifier at the voltage input at the IREF input, input or denominator input, the method is to insert the resistance from the voltage source to the relevant pipe foot. Instrument. The overall accuracy of the small input voltage is limited by the transition of the input end voltage of the JFET computing amplifier.
The negative power supply also allows output to swing in the ground to allow the intermediate value corresponding to the intermediate value of I. However, the voltage V still refers to the ACOM pin. Although the negative voltage does not swing under the default working conditions, it can be done freely. Therefore, the supply of a resistor from VLOG to the negative voltage supply reduces all values of VLOG, thereby increasing the value of VLOG, thereby increasing the value of VLOG, thereby increasing the value intercept. The disadvantage of this method is to reduce the slope due to the diversion of the external resistance, and the ratio of the ratio of the internal resistance to the external resistance is not clear, resulting in errors of the slope and intercept.
Delivery ratio application
Generally, the ratio of two currents needs to be determined, for example, in the absorbance measurement. These are usually used to evaluate the attenuation of non -active optical elements, such as optical filters or variable optical attenuators. In these cases, the reference detector is used to measure the incident power of entering the component. Then use the second detector to measure the existing power and calculate the ratio to determine the attenuation coefficient. Because the AD8305 is basically a ratio measurement device, the well -measured system of molecules and denominator (I and I) is almost the same, it can greatly simplify such measurements.
FIG. 38 illustrates the AD8305 pair ratio capacity in the measurement rate of light. Here is a reference detector two poleThe reference current I provides a proportional proportion to the optical reference power of the optical reference power. The second detector measurement to the proportional transmission signal power. The AD8305 calculates the number of these two current ratios, such as the equivalent 11, and re -expressed the power item in the equivalent 12. Both equations include the end factor of 10000 introduced by Pin VRDZ output offset by PIN VRDZ. If the real (non -offset) pairing ratio shown in Formula 4 is the first choice, VRDZ should be fixed to eliminate offset. As mentioned earlier, the negative power supply is allowed to allow the V tap output to swing at the ground at the pins VNEG, and the input pin Inpt and Iref are also allowed to set it to the ground potential. Therefore, the AD8305 can also be used to determine the number of two voltages.
FIG. 38 also shows how to use two external capacitors and a resistor to achieve a second -order Sallen key low -pass filter. Here, the angle frequency is set to 1kHz, and the filter Q is selected to provide the best flat (no super -adjustment) pulse response. This frequency should be zoomed in or downward, just zoom capacitors according to the appropriate factors. Note that the resistance required to implement the filter is a 4.55 kΩ output resistance on the PIN VLOG. Although this is not exactly the same as the proportion of external resistors, the external resistance may slightly change the Q value of the filter, in most cases, the effect of pulse response can be ignored. Note that the gain (× 2.5) of the buffer is an integral part of the design of this explanatory filter; usually, the filter can be redesigned for other closed -loop gain.The transmission characteristics can be represented by light power. If we assume that these two detectors have the same response, the relationship is:
Use unit pair (AB) loga+logb and define the attenuation as -10 as -10 × Log (P/P), the overall transmission characteristics can be written:
Figure 39 shows the linear IN-DB relationship between the absorbance and circuit output in Figure 38.
Reverse input pole
Some applications may need to be connected to the source current instead of the exchange current, such as the cathode side connected to the optical diode. Figure 40 shows the use of current mirror image circuits. This allows monitoring cathode light power at the same time, and data recovery paths that use transmissions to block large verses in anode. The improved Wilson reflector provides a very close current and high output resistance of the unit. FIG. 41 shows the measurement and transmission function and legal consistency of the AD8305 combined with the current mirror interface.
During the characteristic process of AD8305, the device is considered as a precision current input number pair to the number of pairs to the number of pairs of the input pairs The converter, for several reasons, through lighting photovoltaic twoThe polar tube generates accurate light current. The test current is generated by using a good -calibrated current source (such as Keithley 236) or a high -value resistor from the voltage source to the input pin. When using a very small input current, you need to be very careful. For example, the three -axis output of the current generator is used with the protective device connected to the VSUM. The input record on the PC board is protected by connecting adjacent records to VSUM.
These measures are necessary to minimize the risk of leaking current paths as much as possible. When the nominal bias on the input pin is 0.5 V, the leakage path resistance of the 1 gallal to the ground will be subtracted by 0.5 NA from the input. For the source current of 10 NA, this is equivalent to the error of -0.44 dB. In addition, in the process of characterization, the output resistance at a very high output resistance at the input pin and long cables allows 60Hz and RF emission to introduce a large number of measurement errors. Careful protection technology is essential for reducing the receiving of these strange signals.
The main feature settings shown in FIG. 42 is used to measure V, Static (DC) performance, number consistency, slope and intercept, pipe foot vsum, INPT, INPT The voltage and buffer offset and V drift with the temperature at the IREF. In order to ensure stable operation within the full current range of the I and temperature limit, the filter component with C1 4.7 NF and R13 2 KΩ at the place where Iref is inserted into Iref. In some cases, a fixed resistor is used between the pin VREF and Iref to replace the precision current source. For dynamic testing, including noise and bandwidth measurement, more special settings are needed.
FIG. 43 shows the configuration for measuring buffer bandwidth. The AD8138 evaluation board includes the provision of the offset VLOG at the buffer input end, and allows the use of a single power supply to measure the entire range of i. The input impedance setting of the network analyzer is 1 MΩ.
FIG. 44 The settings shown in FIG. The frequency response measurement of the part of the number amplifier. The AD8138 output offset to 1.5 V DC, and adjust to the depth of 5%in the frequency. Select R1 (within a range of 1.0 gallial) to provide i. The buffer is used to remove VLOG from the measurement system.
The structure in FIG. 45 is used to measure noise performance. The battery provides a power supply voltage and input current to minimize the introduction of the effects of noise noise and grounding circuit. The entire evaluation system, including the current setting resistor, is installed in a closed aluminum shell to provide additional shielding for external noise sources.
FIG. 46 shows the settings for the measurement of pulse response. As with bandwidth measurement, VLOG is directly connected to BFIN, buffering plattifier configurationIncrease for unit.The output of the buffer is connected to the TDS5104 oscilloscope through a short cable, and the input impedance setting is 1 MΩ.LeCroy's output is shifted to generate the initial base current of the given value R1, and then the pulse generates ten years of current jump.
Evaluation Committee
AD8305 has an evaluation board, which is shown in Figure 49.It can be configured as a variety of experiments.The buffer is set up to the unit, which provides a slope of 200 MV/Decade, and the cutting distance is set to 1 NA.Table 4 describes various configuration options.