JPH08330962A - Analog / digital conversion method - Google Patents

Abstract

(57) [Abstract] [Purpose] Without the need for an expensive A / D converter,
Provided is an analog / digital conversion method capable of accommodating a wide dynamic range and obtaining high resolution in a small input range. A small pseudo input signal (Vi-Vs) that does not exceed the input range is formed by offsetting a large analog input signal Vi that exceeds the input range of the A / D converter 11 with a shift signal Vs. , A digital inverse shift signal corresponding to the shift signal with respect to the pseudo input signal (Vi-Vs) 'after the pseudo input signal is converted into a digital signal by the A / D converter 11 and converted into the digital signal. Vs 'is added to obtain the digital signal Vi' corresponding to the original analog input signal.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an analog / digital conversion method, and in particular, it is possible to obtain a wide dynamic range and a high resolution in a relatively narrow range by using a popular A / D converter. A possible analog / digital conversion method.

[0002]

2. Description of the Related Art A / D converters are used in various industrial fields as converters for converting analog signals into digital signals. There are various types of A / D converters such as a feedback type pulse width modulation A / D converter and a successive approximation type A / D converter.
A type that converts a digital signal of about it is widely used at a relatively low cost.

When such a popular A / D converter is used for measuring the axial displacement of a magnetic bearing, for example, the analog output of an inductance displacement sensor or an eddy current sensor is amplified by a sensor amplifier and then this A / D converter is used. Input to D converter.
Then, the analog output of this displacement sensor is converted into a digital signal by the A / D converter, and is processed by the CPU.
The rotating shaft is levitationally supported at the target position by being fed back to the exciting current of the exciting magnetic pole of the magnetic bearing device. The popular A / D converter used for this purpose has, for example, a full scale of an input voltage of about 10 V and outputs a 10 to 12 bit digital signal converted corresponding to the input voltage.

[0004]

However, the axial displacement in the magnetic bearing device has a large movement range, for example, a few millimeters, because the rotating shaft moves from the stationary position to the target floating position during operation at the time of startup. . On the other hand, in the steady-state operation after the rotating shaft has levitated, the moving width of the rotating shaft is small, usually on the order of several μ.

As described above, the output of the displacement sensor requires a wide dynamic range because the moving width of the rotary shaft extends from several μm to several mm. The A / D converter converts this input analog signal into a digital signal and takes it into the CPU for controlling the magnetic bearing. The signal is captured by the A / D converter and C
In a digital signal processing system that processes an input signal with a PU, the A / D
Determines the input range of the converter.

However, A of about 10 to 12 bits
When the / D converter is used and the input range is set to a moving width of several mm, the resolution is lowered, and it becomes impossible to detect a signal of a minute displacement of about several μm in a steady state due to a quantization error. Also, if the maximum input is in the range of several μm in the steady state, the displacement of the rotating shaft in the steady state can be measured with sufficient accuracy, but a large displacement of about several mm at startup will be scaled out, and the measurement Is impossible.

As described above, since the conventional A / D converter determines the input range of the A / D converter from the maximum signal level of the input signal, it corresponds to a wide dynamic range and has a small input range. It was impossible to obtain high resolution at. To make this possible, A /
It is necessary to increase the number of bits in the digital processing unit of the D converter, which requires an expensive A / D converter, which leads to an increase in manufacturing cost.

The present invention has been made in view of the above-mentioned circumstances, and can cope with a wide dynamic range without requiring an expensive A / D converter and can achieve high resolution with a small input range. An object is to provide an analog / digital conversion method that can be obtained.

[0009]

According to the analog / digital conversion method of the present invention, a large analog input signal exceeding the input range of an A / D converter is offset by a shift signal so that the input range is not exceeded. A small pseudo input signal is formed, the pseudo input signal is converted into a digital signal by the A / D converter, and a digital inverse shift corresponding to the shift signal with respect to the pseudo input signal after being converted into the digital signal. It is characterized in that the signals are added to obtain a digital signal corresponding to the original analog input signal.

Further, the shift signal added to the analog input signal is a function of time set in advance so that the pseudo input signal after the addition is near zero.

Further, the shift signal is set to be zero in a steady state of the input signal.

In addition, after the pseudo input signal is formed, the pseudo input signal is amplified at a fixed ratio, and the amplified signal is converted into the digital signal by the A / D converter, and then the digital signal is digitalized at the fixed ratio. It is characterized by further comprising a step of performing a division process.

[0013]

Since the shift signal is canceled to the large analog input signal exceeding the input range of the A / D converter,
A small pseudo signal that does not exceed the input range of the D converter can be obtained. Since a small pseudo input signal is input to the A / D converter and the digital signal corresponding to the shift signal is added after digital conversion, the digital signal of the original analog input signal can be obtained. As a result, A / D
It is possible to convert an analog input signal having a wide dynamic range into a digital signal while keeping the input range of the converter small.

Since the shift signal is a function of time such that the added value with the input signal is close to zero, the pseudo signal value is always converted into an A / D converter even if the fluctuation range of the input signal during the transition is large. Can be kept within the input range of.

Further, since the shift signal is set so as to be zero in the steady state in correspondence with the analog input signal, the analog input signal is not affected by the shift signal in the steady state and has a small input range. It can be operated. As a result, in the steady state, the quantization error can be reduced and analog / digital conversion can be performed with high resolution.

Also, since the pseudo signal is multiplied (amplified) by K times before conversion by the A / D converter, and digital division is performed by 1 / K after conversion, quantization is performed by the A / D converter. It is possible to reduce the quantization error at that time and improve the resolution.

[0017]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the accompanying drawings.

FIG. 1 is a block diagram showing an analog / digital conversion method according to an embodiment of the present invention. The analog input signal Vi is, for example, an analog output of a displacement sensor of the magnetic bearing device, and this signal is converted into a digital signal by the A / D converter 11 and taken into the CPU 10 to be subjected to digital control processing. In this embodiment, a generator (not shown) for the shift signal Vs and a subtracter 12 are provided, and the input signal Vi
If A exceeds the input range of the A / D converter 11, A
The shift signal Vs is added so that the / D converter 11 does not scale out. That is, the shift signal Vs is the input signal Vi
From the analog input pseudo signal (V
i-Vs) is formed. Further, this pseudo signal is amplified K times by the amplifier 13 which is a multiplier, and the A / D converter 11
Is input to

Here, the shift signal Vs added to the input signal Vi is such that the added value of the input signal Vi and the pseudo signal canceled by the shift signal Vs is near zero in the input range of the A / D converter 11. , A function of time of the shift signal Vs is set in advance. This time function predicts and determines the transient response of the measurement object from past experience in advance. Alternatively, a function may be used in which the input signal is monitored and the shift signal is set to zero when the input analog signal Vi has a magnitude within the minimum range of the A / D converter based on the information. Furthermore, the function of time may be set to a function that predicts a time point at which the input analog signal Vi falls within the minimum range of the A / D converter and thereafter makes the shift signal zero. . If such a discontinuous function that causes the function of time to suddenly become zero causes a problem in the stability of the entire control loop, for example, a quadratic function is used to gently approach zero. You may do it.

The amplifier 13 amplifies a pseudo signal output obtained by canceling the shift signal Vs from the input signal Vi so that the input range of the A / D converter 11 is effectively used. For example, when the input range is 10V and the pseudo signal is about 1V, this is K = 10 by the amplifier 13.
By multiplying twice, the input range of the A / D converter 11 can be used in full scale.

A control CP is provided on the output side of the A / D converter 11.
U10 is connected. The control CPU is provided with the digital dividing means 14 formed by the computer program, and the multiplier 13 analog-multiplies the pseudo signal (Vi-Vs) by K. Perform division processing. Reference numeral 15 is a digital adding means for adding the output of the A / D converter 16 from the output of the dividing means 14. The digital adding means 15 is also an arithmetic means in the control CPU 10 similarly formed by a computer program. The inverse shift signal Vs' is a digital signal, which is also a function of time, calculated corresponding to the analog shift signal which is a function of time. Therefore, when the digital reverse signal Vs' is added to the digital pseudo signal (Vi-Vs) 'by the adding means 15, the digital input signal Vi' is obtained.

FIG. 2 shows the waveform of each part for explaining the operation of the analog / digital conversion method of one embodiment of the present invention. FIG. 2A shows, as an example, the waveform of the analog input signal Vi which is the output of the displacement sensor in the magnetic bearing device. The magnetic bearing is started at time T0 in this embodiment. At time T0, the position of the rotary shaft is the stationary position, and the displacement sensor output at this time is -V0. It is started at time T0, the rotary shaft levitates, the rotary shaft moves to the target floating position, and becomes a steady state at time T1. Then, the operation in the steady state, that is, the operation in which the rotary shaft is levitationally supported at the target levitation position, continues until time T2. The displacement of the rotating shaft during this period is extremely small, about several μm. And at time T2,
The operation in the steady state is ended, the rotation of the rotary shaft is stopped, and the rotary shaft returns to the stationary position. Therefore, the output of the displacement sensor starts from almost zero at time T2, and again at -V0 at time T3.
Becomes This voltage transient change, from zero to -V0
Is a few mm in the moving width, which is extremely larger than the moving width of the rotating shaft of several μm in the steady state.

FIG. 2B shows the shift signal Vs. The shift signal Vs is an analog signal that cancels the input signal Vi so that the input signal Vi does not scale out in the A / D converter 11. The shift signal Vs cancels the input signal Vi mainly in a transient state, and is preferably zero within a range in which the addition signal does not exceed the input range of the A / D converter 11. In the steady state, it is set to zero so as not to affect the input signal Vi. Therefore, at time T0, the output is + V0, which decreases with the passage of time, and at time T1, the rotating shaft reaches the position in the steady state, so the output is set to zero. Zero output continues in the steady state.
Then, from the time T2 when the rotating shaft stops rotating to stop the operation of the magnetic bearing device to the time T3, the shift signal linearly increases from zero to the voltage + V0. Then, after the time T3, the output voltage + V0 is continued because the rotating shaft is at the stationary position.

FIG. 2C is a pseudo signal in which the input signal Vi is canceled by the subtracter 12 with the shift signal Vs. This signal is amplified (multiplied) K times by the amplifier 13, and the A / D
It is input to the converter 11. This signal is a pseudo signal Vi in the transient state T0 to T1 and T2 to T3 as shown in the figure.
-Vs. From T1 to T2 in the steady state, the shift signal Vs is zero, which is a fine vibration signal corresponding to a displacement of several μm. The input signal Vi largely fluctuates in the transient state at the time of starting and stopping, but since the shift signal Vs is added, the fluctuation width after the addition becomes small as shown in the figure,
Within the small input range.

The signal K (Vs-Vi) after this amplification is A /
The digital signal is converted by the D converter 11. Then, in the digital division means 14, the amplification factor K of the multiplier (amplifier) 13 is set.
Is digitally divided by. Therefore, the output of the dividing means 14 is a digital signal (Vi-Vs) '. On the other hand, the reverse shift signal Vs' is the analog shift signal Vs shown in FIG. 2B converted into a digital value by calculation.

The digitized reverse shift signal Vs' is
The digital adder 15 outputs the output of the divider 14 (Vi-V
s) 'is digitally added. Therefore, the output of the addition means 15 is the sum of (Vi-Vs) 'and Vs', and the digital signal Vi' corresponding to the analog input signal Vi is output and taken into the magnetic bearing control processing unit 16. Be done. The shift signal of the digital signal Vi 'is A
Since the signal is subtracted before the A / D conversion and added after the A / D conversion, the waveform converted into the analog value is similar to that shown in FIG.

As described above, in the steady state, the signal level is small, and the input signal is amplified and A / D amplified to improve the resolution.
The converter 11 can perform A / D conversion processing. or,
Even if the input signal exceeds the input range of the A / D converter 11 in the transient state, it is canceled by the shift signal Vs so that the input range of the A / D converter 11 is not exceeded even in the transient state. D conversion processing can be performed, and a digital signal Vi ′ corresponding to the input signal Vi can be obtained as an output. Therefore, it is possible to construct a digital signal processing system having a wide dynamic range, a small quantization error, and a high resolution without using an expensive A / D converter having a large number of bits.

In the above description of the embodiment, the sensor output for measuring the displacement of the rotary shaft in the magnetic bearing device is A / D converted as an example, but the gist of the present invention is the displacement of the magnetic bearing device. It is needless to say that it can be applied not only to the output processing of the sensor but also to various A / D conversion applications in industry.

Further, the shift signal Vs is subtracted from the input signal Vi to cancel the shift signal Vs, thereby forming a pseudo signal suitable for a small input range. You may think. In this case, the subtraction process after conversion into a digital signal is the subtraction process. This is just a matter of definition of addition and subtraction.

Further, in this embodiment, the pseudo signal (Vi-Vs) is used.
Is amplified K times by an amplifier and then input to the A / D converter. However, if the input signal is sufficiently large, this amplifier is not always necessary. In this case, 1 / K division processing after conversion into a digital signal is also unnecessary.

[0031]

As described above, according to the present invention, the A / D
In a system that takes a signal into a converter and converts it into a digital signal, an input signal that exceeds the input range of the A / D converter is canceled by a shift signal so as not to scale out, and an inverse shift signal is generated after A / D conversion. Is to be added. Therefore, the input signal having a large fluctuation width at the time of transition is canceled by the shift signal and becomes a pseudo signal having a small fluctuation width, so that the dynamic range of the A / D conversion can be expanded.

By setting the shift signal to zero for the input signal having a small fluctuation width in the steady state, the A / D conversion process can be performed with high resolution. Therefore, it is possible to economically realize an A / D conversion that can obtain a wide dynamic range in a transient state and a high resolution in a stationary state by using a popular A / D converter.

[Brief description of drawings]

FIG. 1 is a block diagram illustrating an analog / digital conversion method according to an embodiment of the present invention.

FIG. 2 is a time chart of each part in the above converter method, in which (A) input signal Vi, (B) shift signal Vs,
(C) shows these pseudo signals (Vi-Vs).

[Explanation of Codes] 10 CPU 11 A / D Converter 12 (Analog) Adder 13 Amplifier (Multiplier) 14 (Digital) Division Means 15 (Digital) Subtraction Means

Claims (4)

[Claims] 1. A small pseudo input signal that does not exceed the input range is formed by offsetting a large analog input signal that exceeds the input range of the A / D converter with a shift signal, and the pseudo input signal is generated. The digital signal is converted by the A / D converter, and the digital inverse shift signal corresponding to the shift signal is added to the converted pseudo input signal to correspond to the original analog input signal. An analog / digital conversion method, characterized in that the obtained digital signal is obtained. 2. The analog signal according to claim 1, wherein the shift signal for canceling the analog input signal is a function of time set in advance so that the pseudo input signal after the cancellation is near zero. Digital conversion method. 3. The analog / digital conversion method according to claim 1, wherein the shift signal is set to be zero in a steady state of the input signal. 4. Forming the pseudo input signal, amplifying the pseudo input signal at a constant ratio, and converting the amplified signal into the digital signal by the A / D converter, and then digitally performing the digital signal at the constant ratio. The analog / digital conversion method according to claim 1, further comprising a step of performing a division process.