EP2193339A1

EP2193339A1 - Field device having an analog output

Info

Publication number: EP2193339A1
Application number: EP08834820A
Authority: EP
Inventors: Eric Chemisky; Michael Geppert; Ulrich Hahn; Simon Rohrbach
Original assignee: Siemens AG
Current assignee: Siemens AG
Priority date: 2007-09-28
Filing date: 2008-09-26
Publication date: 2010-06-09
Also published as: CN101809413A; WO2009043821A1; DE102007046560A1; US20100302085A1

Abstract

The invention relates to a field device having an analog output, especially a measuring transducer (1) for process instrumentation having a 4-20 mA interface as the analog output (7). For digital-to-analog conversion, a digital value is split into a digital coarse portion and a digital fine portion. Depending on the digital coarse portion, a first analog signal (V1) is generated using a pulse width modulator (BUF1) having a low path filter (TP1) mounted downstream thereof, which signal is above the analog output signal (VOUT), and a second output signal (V2) using a pulse width modulator (BUF2) having a low path filter (TP2) mounted downstream thereof, which signal is below the analog output signal(VOUT). Both analog signals (V1, V2) are guided to a third pulse width modulator (SW1) which is controlled depending on the digital fine portion, a low-pass filter (TP3) being mounted downstream thereof. The invention allows provision of an analog output signal (VOUT) having high resolution and good dynamic properties. The field device is further characterized by a digital-to-analog converter that can be produced with especially little complication.

Description

description

Field device with an analog output

The invention relates to a field device with an analog output, in particular a transmitter for process instrumentation with a 4-20 mA interface as an analog output, according to the preamble of claim 1.

Such a field device is for example from the

DE 199 30 661 Al known. In process instrumentation, transducers with a 4-20 mA interface are used in a variety of ways to measure physical or chemical parameters, such as pressure, temperature or pH. These usually have a sensor, the sensor signal amplified, digitized, and then evaluated in a microcontroller and corrected for linearity and temperature behavior. The thus processed sensor signal is converted in an output circuit with a digital / analog converter into an analog output signal, in this case an output current in the range of 4-20 mA, and via a two-wire line to an evaluation device, for example a programmable logic controller in an automation network. transfer.

On the other hand, a programmable logic controller as a field device may have an analog output, for example for the transfer of a manipulated variable to a control valve as an actuator with a corresponding analog input.

To generate the analog output signal, digital / analog converters with different modes of operation are known. For example, digital / analog converters with an R2R network implemented as integrated components are available. A disadvantage of these components, however, are the associated costs and also their high power consumption. Especially for field devices that are powered by a 4-20 mA interface with their power required for operation, this can represent a significant disadvantage, since the available energy is very limited. Another possibility for digital / analog conversion can be seen to use a timer output of the microcontroller to control a pulse width modulator on which a highly accurate reference voltage out and the smoothing of the output signal is followed by a low-pass filter. However, the problem arises that a compromise must be made between the achievable dynamics and the setting accuracy of the analog signal. The frequency of the pulse width modulated signal, which has a direct effect on the dynamics, results from the bit resolution of the digital / analog conversion and the clock frequency of the microcontroller and is proportional to the product of these two variables. The timing of the microcontroller has a direct effect on its power consumption and can not be increased arbitrarily. On the other hand, the frequency of the pulse width modulated signal can not be arbitrarily reduced in order to achieve a higher bit resolution, since this is decisive for the dynamics of the generated analog output signal.

The invention has for its object to provide a field device with an analog output, in particular a transmitter for process instrumentation with a 4-20 mA interface as analog output, which is characterized by low power consumption and with which an analog output signal in high resolution and can be generated with great dynamics.

To solve this problem, the new field device of the type mentioned in the characterizing part of claim 1 features. In the dependent claims advantageous developments of the invention are described.

The invention is based on the recognition that the conflict between dynamics and accuracy of the analog output signal can be resolved by a stepwise digital / analog conversion. For this purpose, in a first stage, two first Analog signals generated with lower resolution, which are above and below the desired analog output signal. These are used in a downstream, second stage as a voltage level for generating a pulse width modulated signal whose pulse pause ratio only has to be set with an accuracy which corresponds to the resolution to be achieved compared to the coarse resolution.

Since each stage takes over a part of the resolution is at

Using a microcontroller to generate the timing signals for the pulse width modulation at the same timing a much higher dynamics achievable. On the other hand, to achieve a given dynamic, the microcontroller can now be clocked at a lower frequency, so that the

Energy consumption of the microcontroller decreases and as a result more energy is available for the actual measuring task of a transmitter. This can be used to improve the measurement accuracy of the transmitter.

Since a microcontroller is present in most field devices anyway, a particularly simple implementation of the digital / analog converter can be achieved if, due to suitable programming, it divides the digital value into a digital coarse fraction and a digital fine fraction and controls it the pulse width modulators required time signals generated.

A particularly high accuracy of the digital / analog conversion is possible in an advantageous manner when the low passes of the

Realized digital / analog converter with passive components and are dimensioned such that their input resistance compared to the Ausgangswidertand the pulse width modulators is much higher.

In order for a poor dynamic of a stage not to adversely affect the dynamics of the entire digital / analog implementation, the dynamics of the stages should be as equal as possible. This can be achieved in a simple manner in that the resolution of the coarse fraction and the resolution of the fine fraction corresponds to the same number of bits. That is to say, the digital coarse part substantially corresponds to the N most significant bits and the digital fine part substantially to the m least significant bits of the digital value, and N is approximately equal to m. Noise of the analogue output signal can be largely avoided if the production of the coarse fraction occurs with hysteresis.

With reference to the drawings, in which an embodiment of the invention is shown, the invention and refinements and advantages are explained in more detail below.

Show it:

FIG. 1 shows a basic structure of a transmitter,

FIG. 2 is a block diagram of a digital / analog converter;

FIG. 3 shows a timing diagram for explaining its functional principle and

4 shows a circuit of a digital / analog converter.

According to FIG. 1, a measuring transducer 1 for detecting a physical or chemical quantity X of a process has a pick-up 2, which converts this quantity into a measuring signal 3. In a pre-processing 4, the measurement signal 3 is amplified and digitized. In digital form, the thus preprocessed measurement signal is supplied to a microcontroller 5 which, for example, carries out a compensation of non-linearities and temperature influences and calculates the measured value to be output. In a digital / analog converter 6, the digital measured value ascertained in the microcontroller 5 is converted into an analog

Converted output signal via a 4-20 mA interface 7 for further use in a process plant, in which the transmitter 1 is inserted, is output.

For digital / analog conversion, a microcontroller μC according to FIG. 2 generates three time signals PWM ₁ , PWM ₂ and PWM ₃ . The time signals PWMi and PWM ₂ are determined according to a coarse fraction of the digital value, the time signal PWM ₃ corresponding to a digital fine fraction. A buffer BUFi, which has the function of a pulse width modulator, generates a pulse width modulated signal corresponding to the time signal PWMi, the upper

Level of a reference voltage V _re f and whose lower level is the reference ground GND. The signal is smoothed in a low-pass filter TPi, so that a first analog signal Vi is present, which is above the desired analog output signal V _O uτ. In an analogous manner, with the aid of the time signal PWM ₂ , with a buffer BUF ₂ and with a low-pass filter TP _2, a second analog signal V _{2 is} generated whose level is lower than the desired output signal V _O uτ. With the time signal PWM ₃ , which corresponds to the digital fine fraction of the digital value, a switch SWi is controlled, which thus also the

Function of a pulse width modulator has. On the switch SWi the first analog signal Vi and the second analog signal V ₂ are performed. By a low pass TP ₃ , which is the switch SWi downstream, the pulse width modulated signal is again smoothed, so that finally the analog output signal V _O uτ is present.

The following is an example of 17-bit resolution digital-to-analog conversion. The coarse fraction has a resolution of 9 bits, the fines also. A resolution of one bit, which remains at a summation of the resolutions of coarse and fine fraction compared to the resolution of the digital value is, as explained later, needed to realize a hysteresis of the coarse fraction.

The first analog signal Vi can be calculated according to the formula N.

^V i = V ref

Ni essentially corresponds to the most significant bits of the digital value and has a value range between 0 and 2 ⁹ -1.

The level of the second analog signal V ₂ can be calculated according to the formula:

N ₂ also largely corresponds to the most significant bits of the digital value and has the same value range as Ni. As explained in more detail below, a hysteresis is used to avoid noise in the analog output signal V _O uτ. For this it is determined

Ni ₌ N ₂ + 2.

With the aid of the changeover switch SW1, switching takes place in accordance with the time signal PWM ₃ between the first analog signal Vi and the second analog signal V ₂ . The level of the analog output signal V _0U T can be calculated according to the formula:

The value m corresponds to the fine fraction determined with the aid of the microcontroller .mu.C, which is used to set the timer for generating the time signal PWM ₃ . . It also has a range of values between 0 and 2 ⁹ -l Substituting in the last formula Ni by ₂ N + 2 and N ₂ through N, is obtained for the level of the analog output signal V _O uτ:

V ^V _ref

OUT 2 ¹⁷ • (m + 2 ⁸ • N). This results in a digital / analog conversion with 17-bit resolution, with the value m the least significant bits and the value N corresponds to the most significant bits.

The digital values for the coarse fraction and the fine fraction, which are determined in the microcontroller .mu.C, need not exactly correspond to the most significant bits or the least significant bits of the digital value at all times, but are set differently from them under certain conditions. With the aid of the time signal PWM3, which corresponds to the digital fine fraction, the analog output voltage is set as a function of the first analog signal Vi and of the second analog signal V2. If the value of the nine least significant bits is close to its limits, that is close to the value 0 or close to the value 2 ⁹ -l, it might happen that the digital coarse fraction and thus the two analog signals when these limits are exceeded in the absence of hysteresis Vi and V ₂ constantly switch back and forth. This would result in unnecessary noise of the analog output signal V _O uτ.

To prevent this not the values 0 and 2 ⁹ -l be used as the switching points of the digital fine fraction, but a value of 12.5% and a value of 87.5% of the total range of values of the digital fine fraction, thus for example in a range of values 512 the values 64 and 448. This is explained in more detail below with reference to FIG. FIG. 3 shows a time diagram in which a profile 31 of an analogue output signal V _O uτ, a profile 32 of a first analogue signal Vi, a profile 33 of a second analogue signal V ₂ and a profile 34 corresponding to the respective digital fine component corresponds, are plotted over time. In the left-hand area for times t <ti, the digital coarse fraction is set to the value N. The value N ₂ for setting the time signal PWM ₂ is equal to N, the value Ni for setting the time signal PWMi is N + 2. This setting of the digital coarse fraction remains constant as long as the digital fines within its limits between 12.5% and 87 , 5% of the value range is located. At time ti, the digi- Valley value such that the digital fine fraction falls below the 12.5% limit and correspondingly the analog output signal 31 approaches the second analog signal 33. As a result, at the switching point at time ti, the values Ni and N ₂ are decremented by one. At the same time, the digital fine fraction is increased by about 50% of its value range, so that no switching jump can be detected in the course of the analogue output signal 31. The new value m _{NEW of} the digital fine fraction results according to the formula:

^m NEÜ = ^m ALT ⁺ 2 ^{8 •}

Thus, m _{NEW is} about 62.5% and is below the 87.5% limit at which the values Ni and N _{2 would} again be increased by 1. It is thus realized a hysteresis, by which a noise is prevented at switching points.

At the second switch-over time t ₂ , the digital fine fraction corresponding to the curve 34 exceeds the 87.5% limit of its value range. This is followed immediately by an increment of the values Ni and N ₂ by 1 and a reduction of the digital fine fraction by 2 ⁸ . In the direct connection of the switchover, the value of the digital fine fraction is approximately 37.5% of its value range.

The division of the digital value, which corresponds to the analog output signal V _0U T, into the digital coarse fraction and the digital fines content with hysteresis of the coarse fraction is carried out in the microcontroller μC on the basis of its programming. An additional circuit complexity is advantageously not connected with it.

FIG. 4 shows a circuit 41 which is suitable for implementing the digital / analog converter. The microcontroller used is an integrated module 42 of the type MSP430 which has three timer outputs 43, 44 and 45 which are used to realize pulse width modulated time signals PWMi, PWM ₂ and PWM ₃ are used. The timing signals PWMi and PWM2 are at two Buffer 46 or 47 of the type 74LVC04 supplied, which are supplied by a diode 48 with a high-precision reference voltage. The buffers 46 and 47 are each followed by a low pass, which consists of a resistor Ri and a capacitor Ci or of a resistor R2 and a capacitor C ₂ . The resistors each have 51 kΩ, the capacitors 33 nF each. The analog signals smoothed in this manner are fed to two inputs of a Texas Instruments 3157 type switch 49. The time signal PWM ₃ is used to actuate the switch 49. The switch 49 is again followed by a low-pass filter with a resistor R3 of 150 kΩ and a capacitor C ₃ of 100 nF. This low pass finally provides an analog output signal 50 corresponding to the predetermined digital value. For the realization of the low-pass filters are deliberately passive RC filters used and no circuits with active components, since they provide a very good accuracy. It is important that the output resistance of the buffers 46 and 47 and the output resistance of the switch 49 in comparison to the input impedance of the respective downstream low-pass filter is small.

It is particularly clear from the circuit according to FIG. 4 that the digital / analogue converter with the circuit 41 shown can be produced particularly inexpensively. Despite the high overall resolution of the digital-to-analog converter, time signals PWMi, PWM ₂ and PWM ₃ with a comparatively high frequency can be used because of the stages connected in series to generate the analog output signal. This leads to a high dynamic of the digital / analog conversion.

Claims

claims

1. field device with an analog output, in particular transmitter (1) for process instrumentation with a 4-20 mA interface as analog output (7), and with a digital / analog converter (6) for generating an analog output signal at the analog output, characterized in that the digital / analog converter comprises the following components:

a device (μC) for dividing a digital value into a digital coarse fraction and a digital fines fraction,

- A first circuit with a variable depending on the coarse fraction of the first pulse width modulator (BUFi) to which a reference voltage (V _re f) is performed, with downstream low pass (TPi) for generating a first analog signal (Vi), the above the analog Output signal (VOUT) is

a second circuit having a second pulse width modulator (BUF2) which can be adjusted as a function of the coarse fraction and to which the reference voltage (V _re f) is connected, with a downstream low pass filter (TP2) for generating a second analog signal (V ₂ ) which is below the analog output signal (VOUT),

a third circuit having a third pulse width modulator (SWi) which can be set as a function of the fine fraction and to which the first analog signal (Vi) and the second analog signal (V ₂ ) are fed, and wherein the upper level of the output signal of the third pulse width modulator (SWi) is the first analog signal (Vi) and the lower level of the output signal of the third pulse width modulator (SWi) corresponds to the second analog signal (V ₂ ) and the third pulse width modulator (SWi) is followed by a low pass (TP ₃ ) for generating the analog output signal (V _OU T) ,

2. Field device according to claim 1, characterized in that the means for dividing the digital value is a microcontroller (μC), by which time signals (PWMi, PWM ₂ , PWM ₃ ) for controlling the pulse width modulators (BUFi, BUF ₂ , SWi) can be generated.

3. Field device according to claim 1 or 2, characterized in that the low-pass filters (TPi, TP ₂ , TP ₃ ) are realized with passive components such that the input resistance of the low-pass is large compared to the output resistance of the pulse width modulators (BUFi, BUF ₂ , SWi).

4. Field device according to one of the preceding claims, characterized in that the means for dividing the digital tallyvalue is formed such that the digital coarse fraction substantially corresponds to the most significant bits and the digital fines substantially the least significant bits of the digital value, wherein their number is about the same, and that the generation of the coarse fraction has a hysteresis.