Is the reference voltage source a question of whether it is needed or not? It is obviously a question of whether it is suitable or not.
It’s an analog world. Whether it’s a car, a microwave, or a cell phone, all electronic devices must interact with the “real” world in some way. To do this, the electronics must be able to map real-world measurements (speed, pressure, length, temperature) to a measurable quantity in the electronic world (voltage). Of course, to measure voltage, you need a standard to measure against, and that standard is a voltage reference. For the system designer, the question is not if you need a voltage reference, but which one to use?
A voltage reference is simply a circuit or circuit element that provides a known potential for as long as the circuit needs it. This could be minutes, hours, or years. If a product needs to capture real-world information such as battery voltage or current, power consumption, signal size or characteristics, fault identification, etc., then the signal of interest must be compared to a standard. Every comparator, ADC, DAC, or detection circuit must have a voltage reference to do the job described above (Figure 1). Comparing the signal of interest to a known value allows any signal to be accurately quantified.
Figure 1. Typical use of a voltage reference for an ADC.
Voltage references come in many forms and offer different characteristics, but ultimately, accuracy and stability are the most important characteristics of a voltage reference, since its primary purpose is to provide a known output voltage. Variations from this known value are errors. Voltage reference specifications typically use the following definitions to predict their uncertainty under certain conditions.
Table 1. High-Performance Voltage Reference Specifications (Partial)
The change in output voltage measured at a given temperature (usually 25°C). Although the initial output voltage may vary from device to device, if it is constant for a given device, it is easy to calibrate it out.
This specification is the most widely used for evaluating voltage reference performance because it shows the change in output voltage over temperature. Temperature drift is caused by imperfections and nonlinearities in circuit components and is therefore often nonlinear.
For many devices, temperature drift, TC (measured in ppm/°C), is the dominant source of error. For devices with consistent drift, calibration is feasible. A common misconception about temperature drift is that it is linear. This leads to statements such as “a device will drift less over a smaller temperature range,” when the opposite is often true. TC is typically specified using a “black box approach” to give an idea of the likely error over the entire operating temperature range. It is a calculated value based only on the minimum and maximum values of the voltage, and does not take into account the temperature at which these extremes occur.
For references that have very good linearity over the specified temperature range, or for those that are not carefully trimmed, the worst-case error can be assumed to be proportional to the temperature range. This is because the maximum and minimum output voltages are most likely obtained at the maximum and minimum operating temperatures. However, for carefully trimmed references (usually identified by their very low temperature drift), their nonlinear characteristics may dominate.
For example, a voltage reference specified at 100ppm/°C will tend to have fairly good linearity over any temperature range because the drift due to component mismatch completely masks its inherent nonlinearity. In contrast, a voltage reference specified at 5ppm/°C will have its temperature drift dominated by nonlinearity.
This is easily seen in the output voltage vs. temperature characteristics shown in Figure 2. Note that two possible temperature characteristics are represented. An uncompensated bandgap reference exhibits a parabola with minimums at the temperature extremes and a maximum in the middle. A temperature compensated bandgap reference such as the LT1019 shown here exhibits an “S” shaped curve with a maximum slope near the center of the temperature range. In the latter case, the nonlinearity is exacerbated, reducing the overall uncertainty over temperature.
Figure 2. Voltage reference temperature characteristics.
The best use of the temperature drift specification is to calculate the maximum total error over the specified temperature range. Unless the temperature drift characteristic is well understood, it is generally not recommended to calculate the error over an unspecified temperature range.
This specification measures the tendency of the reference voltage to change over time, independent of other variables. Initial drift is primarily caused by changes in mechanical stress, which usually comes from differences in the expansion rates of the lead frame, die, and mold compound. This stress effect tends to have a large initial shift, which then decreases rapidly over time. Initial drift also includes changes in the electrical characteristics of circuit elements, including the build-up of device characteristics at the atomic level. Longer term drift is caused by electrical changes in circuit elements and is often referred to as "aging." This drift tends to occur at a slower rate than initial drift, and the rate of change decreases further over time. Therefore, it is often expressed as "drift/√khr." Voltage references tend to age faster at higher temperatures.
This specification is often overlooked, but it can be a major source of error. It is mechanical in nature and is the result of changes in die stress caused by thermal cycling. After a large temperature cycle, hysteresis can be observed at a given temperature, which manifests itself as a change in output voltage. It is independent of temperature coefficient and time drift, and can reduce the effectiveness of the initial voltage calibration.
Most voltage references tend to vary around the nominal output voltage during subsequent temperature cycling, so thermal hysteresis is usually limited to a predictable maximum value. Each manufacturer has its own way of specifying this parameter, so typical values can be misleading. When estimating output voltage errors, distribution data provided in data sheets (such as LT1790 and LTC6652) are more useful.
Other specifications that may be important, depending on application requirements, include:
-
Voltage noise
-
Line Regulation/PSRR
-
Hello Elena
-
Pressure difference
-
Supply voltage range
-
Supply Current
There are two main categories of voltage references: shunt and series. See Table 2 for ADI’s series and shunt voltage references.
Table 2. Voltage References from ADI
Shunt voltage references are 2-terminal devices that are usually designed to operate over a specified current range. While most shunt voltage references are bandgap types and available in a variety of voltages, they can be considered as easy to use as Zener diode types, and indeed they are.
The most common circuit is to connect one pin of the reference voltage source to ground and the other pin to a resistor. The other pin of the resistor is connected to a resistor. The other pin of the resistor is connected to a power supply. In this way, it essentially becomes a three-terminal circuit. The common terminal of the reference voltage source and the resistor is the output. The common terminal of the reference voltage source and the resistor is the output. The resistors must be chosen so that the minimum and maximum currents through the reference voltage source are within the rated range over the entire supply range and load current range. These reference voltage sources are easy to use for design if the supply voltage and load current do not vary much. If one or both of them may vary significantly, it will usually cause the circuit to actually dissipate much more power than is required for the nominal case. In this sense, it can be considered to operate like a Class A amplifier.
The advantages of shunt reference voltage sources include: simple design, small package, and good stability under wide current and load conditions. In addition, it is easy to design as a negative reference voltage source and can be used with very high supply voltages (because the external resistor will share most of the potential) or with very low supply voltages (because the output can be only a few millivolts below the supply voltage). Shunt products provided by ADI include LT1004, LT1009, LT1389, LT1634, LM399 and LTZ1000. The typical shunt circuit is shown in Figure 3.
Figure 3. Shunt voltage reference.
A series reference is a three (or more) terminal device. It is much like a low dropout (LDO) regulator, so many of its advantages are the same. Most notably, it consumes a relatively fixed supply current over a wide range of supply voltages, and conducts load current only when the load requires it. This makes it ideal for circuits with large variations in supply voltage or load current. It is particularly useful in circuits with very high load currents because there is no series resistance between the reference and the supply.
The series products provided by ADI include LT1460, LT1790, LT1461, LT1021, LT1236, LT1027, LTC6652, LT6660, etc. Products such as LT1021 and LT1019 can be used as shunt or series reference voltage sources. The series reference voltage source circuit is shown in Figure 4.
Figure 4. Series voltage reference.
There are many ways to design a voltage reference IC. Each approach has specific advantages and disadvantages.
The buried Zener type voltage reference is a relatively simple design. Zener (or avalanche) diodes have a predictable reverse voltage that is fairly stable over temperature and very stable over time. These diodes typically have very low noise and very good stability over time if maintained over a small temperature range, making them useful in applications where the reference voltage must vary as little as possible.
This stability can be attributed to the relatively low number of components and die area, as well as the delicate construction of the Zener element, compared to other types of reference circuits. However, relatively large variations in initial voltage and temperature drift are common. Circuitry can be added to compensate for these imperfections or to provide a range of output voltages. Both shunt and series references use Zener diodes.
Devices such as the LT1021, LT1236, and LT1027 use internal current sources and amplifiers to regulate the Zener voltage and current for improved stability and to provide a variety of output voltages, such as 5 V, 7 V, and 10 V. This additional circuitry makes the Zener diode more compatible with many application circuits, but requires greater power supply margin and may introduce additional errors.
In addition, the LM399 and LTZ1000 use internal heating elements and additional transistors to stabilize the temperature drift of the Zener diode, achieving the best combination of temperature and time stability. In addition, these Zener diode-based products have extremely low noise and provide optimal performance. The LTZ1000 has a temperature drift of 0.05ppm/°C, a long-term stability of 2µV/√kHr, and a noise of 1.2µVP-P. To make it easier to understand, using a laboratory instrument as an example, the total uncertainty of the LTZ1000 reference voltage caused by noise and temperature is only about 1.7ppm, plus less than 1ppm per month due to aging.
While a Zener diode can be used to make a high-performance voltage reference, it lacks flexibility. Specifically, it requires a supply voltage of more than 7V and provides relatively few output voltages. In contrast, a bandgap reference can generate a wide range of output voltages with very little supply headroom—typically less than 100mV. Bandgap references can be designed to provide very accurate initial output voltages and very low temperature drift, eliminating the need for time-consuming in-application calibration.
Bandgap operation is based on the basic properties of the bipolar junction transistor. A simplified version of the circuit for a basic bandgap reference, the LT1004, is shown in Figure 5. It can be seen that a pair of unmatched bipolar junction transistors have a difference in VBE that is proportional to temperature. This difference can be used to generate a current that rises linearly with temperature. When this current is driven through a resistor and transistor, if they are sized appropriately, the change in the base-emitter voltage of the transistor with temperature will cancel out the change in the voltage across the resistor. While this cancellation is not completely linear, it can be compensated for with additional circuitry to make the temperature drift very low.
Figure 5. Designing a bandgap circuit to provide a theoretically zero temperature coefficient
The math behind the basic bandgap voltage reference is interesting because it combines a known temperature coefficient with a unique resistivity to produce a reference voltage that theoretically has zero temperature drift. Figure 5 shows two transistors that have been adjusted so that the emitter area of Q10 is 10 times that of Q11, while the collector currents of Q12 and Q13 remain equal. This produces a known voltage between the bases of the two transistors:
Where k is the Boltzmann constant in J/K (1.38 × 10-23), T is the temperature in Kelvin (273 + T (°C)), and q is the electron charge in coulombs (1.6x10-19). At 25°C, kT/q is 25.7mV, with a positive temperature coefficient of 86μV/°C. ∆VBE is this voltage multiplied by ln(10) or 2.3, which is about 60mV at 25°C, with a temperature coefficient of 0.2mV/°C.
Applying this voltage to the 50k resistor connected between the bases produces a current proportional to temperature. This current biases diode Q14, which has a voltage of 575mV at 25°C, with a temperature coefficient of -2.2mV/°C. The resistors are used to produce voltage drops with positive temperature coefficients, which are applied to the Q14 diode voltage, resulting in a reference voltage potential of approximately 1.235V, with a theoretical temperature coefficient of 0mV/°C. These voltage drops are shown in Figure 5. The balance of the circuit provides bias current and output drive.
ADI produces a wide variety of bandgap references, including the small, inexpensive precision series reference LT1460, the ultra-low power shunt reference LT1389, and the ultra-high accuracy, low drift references LT1461 and LTC6652. Available output voltages include 1.2V, 1.25V, 2.048V, 2.5V, 3.0V, 3.3V, 4.096V, 4.5V, 5V, and 10V. These references can be provided over a wide range of supply and load conditions with minimal voltage and current overhead. Products may have very high accuracy, such as the LT1461, LT1019, LTC6652, and LT1790; may be very small, such as the LT1790 and LT1460 (SOT23), or the LT6660 in a 2 mm × 2 mm DFN package; or very low power, such as the LT1389, which only consumes 800nA. While Zener references tend to have better performance in terms of noise and long-term stability, new bandgap references are closing the gap, such as the LTC6652, which has 2ppm peak-to-peak noise (0.1Hz to 10Hz).
This voltage reference is based on the temperature characteristics of bipolar transistors, but the output voltage can be as low as a few millivolts. It is suitable for ultra-low voltage circuits, especially comparator applications where the threshold must be less than the conventional bandgap voltage (about 1.2V).
Figure 6 shows the core circuit of the LM10, which is similar to a normal bandgap reference, in which components that are proportional and inversely proportional to temperature are combined to obtain a constant 200mV reference voltage. Fractional bandgap references typically use ∆VBE to generate a current that is proportional to temperature and VBE to generate a current that is inversely proportional to temperature. The two are combined in a resistor element in the appropriate ratio to produce a voltage that does not change with temperature. The size of the resistor can be changed to change the reference voltage without affecting the temperature characteristics. This is different from a traditional bandgap circuit in that fractional bandgap circuits combine currents, while traditional circuits tend to combine voltages, usually base-emitter voltage and I•R with opposite TC.
Figure 6. 200mV reference voltage source circuit
Fractional bandgap references like the LM10 circuit are also based on subtraction in some cases. The LT6650 has such a reference voltage of 400mV and is coupled with an amplifier. Thus, the reference voltage can be varied by changing the gain of the amplifier and a buffered output is provided. Any output voltage from 0.4V to a few millivolts below the supply voltage can be generated using this simple circuit.
Figure 7. The LT6700 supports comparisons with thresholds as low as 400mV.
The LT6700 (Figure 7) and LT6703 are more integrated solutions that combine a 400mV reference with a comparator that can be used as a voltage monitor or window comparator. The 400mV reference can monitor small input signals, reducing the complexity of the monitoring circuit; it can also monitor circuit elements that operate with very low supply voltages. If the threshold is larger, a simple resistor divider can be added (Figure 8). These products are available in small packages (SOT23), low power consumption (less than 10μA), and support a wide supply range (1.4V to 18V). In addition, the LT6700 is available in a 2mm x 3mm DFN package, and the LT6703 is available in a 2mm x 2mm DFN package.
Figure 8. Setting the upper threshold by dividing the input voltage
With all these options in mind, how do you choose the right voltage reference for your application? Here are some tips to narrow your choices:
-
Is the supply voltage very high? Choose a shunt reference.
-
Does the supply voltage or load current vary over a wide range? Choose a series reference.
-
Need high power efficiency? Choose a series reference. .
-
Determine the actual temperature range. ADI provides specifications and guaranteed operating performance for various temperature ranges, including 0°C to 70°C, -40°C to 85°C, and -40°C to 125°C.
-
Accuracy requirements should be realistic. It is important to understand the accuracy required for the application. This helps determine the key specifications. With this requirement in mind, multiply the temperature drift by the specified temperature range, add the initial accuracy error, thermal hysteresis, and long-term drift during the expected product life, and subtract any items that will be calibrated at the factory or recalibrated periodically to get the overall accuracy. For the most demanding applications, noise, line regulation, and load regulation errors can also be added. For example, a voltage reference with an initial accuracy error of 0.1% (1000ppm), a temperature drift of 25ppm/°C from -40°C to 85°C, 200ppm thermal hysteresis, 2ppm peak-to-peak noise, and a time drift of 50ppm/√kHr will have a total uncertainty of more than 4300ppm when the circuit is built. This uncertainty increases by 50ppm in the first 1000 hours after the circuit is powered on. The initial accuracy can be calibrated to reduce the error to 3300ppm+50ppm•√(t/1000hrs).
-
实际电源范围是什么?最大预期电源电压是多少?是否存在基准电压源IC必须承受的故障情况,例如电池电源切断或热插拔感应电源尖峰等?这可能会显著减少可选择的基准电压源数量。
-
What power consumption might a voltage reference have? Voltage references tend to fall into several categories: greater than 1mA, ~500μA, <300μA, <50μA, <10μA, <1μA.
-
How much current is in the load? Does the load consume a lot of current or draw current that the reference must sink? Many references can only source a small amount of current to the load, and few can sink a lot of current. The load regulation specification is a good way to account for this.
-
How Much Space Is There? Voltage references come in a variety of packages, including metal cans, plastic packages (DIP, SOIC, SOT), and very small packages such as the LT6660 in a 2mm x 2mm DFN. It is generally believed that references in larger packages will have less error due to mechanical stress than those in smaller packages. While it is true that some references perform better in larger packages, there is evidence that the performance difference is not directly related to package size. It is more likely that because the smaller package uses a smaller die, some performance trade-off must be made to accommodate the circuitry on the die. Often, the method of mounting the package has a greater impact on performance than the actual package, and paying close attention to mounting method and location can maximize performance. In addition, when the PCB is flexed, a device with a smaller footprint may experience less stress than a device with a larger footprint.
ADI offers a wide range of voltage reference products, including series and shunt references, with designs of Zener diode, bandgap, and other types. Voltage references are available in a variety of performance and temperature grades, and in almost all known package types. They range from the highest precision products to small and inexpensive products. With a large library of voltage reference products, ADI has a voltage reference to meet the needs of almost all applications.
ADI calls for satisfying everyone's pursuit of "freshness" through second-hand exchanges~
Featured Posts