JPS5819945A - Multichannel analog/digital converting circuit - Google Patents

Abstract

PURPOSE:To reduce the number of parts and reduce the fault rate and the cost, by using only one A/D converter for many analog inputs for the purpose of inputting measure data to a computer. CONSTITUTION:When a correction signal 11 is transmitted from a computer 4, the output of an amplifier 5 is given to an A/D converter 3 to produce a digital signal and is inputted to CPU4. Stored contents are converted to an analog signal by a D/A converter 8 and are fed back as a signal, which corrects the offset in the direction where the output signal reduced to zero, to the amplifier 5. Next, the correction signal 11 is transmitted from the CPU4 to input amplifiers 1A- 1Z, and the input signal is put into the state of full scale in the input amplifier 1A, and the output becomes a digital signal and is inputted to the CPU4. Its contents come to an analog signal by a D/A conveter 7 and are fed back as an amplification degree adjusting signal 15 to the amplifier 5 so as to attain the amplification degree a prescribed value.

Description

[Detailed Description of the Invention] The present invention provides a method for inputting measurement data into a computer.
The present invention relates to an improvement in a circuit that successively converts a large number of analog inputs into digital quantities using only one analog-to-digital converter (hereinafter abbreviated as an A/D converter).

In recent years, computers have been widely used in the field of measurement value processing. The objects of measurement, such as temperature, pressure, and speed, are usually detected as analog electrical signals, which are amplified to an appropriate level using an amplifier, and then converted to digital signals using an A/D converter, which is then sent as data to a computer. is entered as . The input data is processed by a computer,
It is also used for recording and control after being organized.

By the way, A/D converters are expensive, and when measuring hundreds to thousands of points, the cost of using an A/D converter for each analog signal is enormous. This means that the price exceeds the price of the computer. Therefore, in order to reduce the cost of the measurement system, the number of expensive A/D converters is reduced, and several 100 analog signals are switched using relays, etc. 2, and multi-channel A is converted one after another by one A/D converter. /D conversion method is widely used industrially. Although this method is an extremely excellent method for reducing costs, the drawbacks of this method have become a problem as the accuracy required for industrial measurement has become higher in recent years.

The cause of the problem, which will be explained in detail later, is that there are errors in the offset and amplification of the input amplifier for each analog signal, errors in the thermal electromotive force of the relay, etc. This makes it difficult to improve measurement accuracy because it mixes into the signal. Therefore, in order to reduce these errors, it has been put into practice to add a correction circuit to each input circuit, but using a large number of correction circuits requires a large number of parts, which increases the cost and requires multi-channel conversion. Not only does it go against the goal of circuits to reduce costs, but it also has disadvantages such as lower reliability due to an increase in failures and lower operating rates due to repeated correction work in a short period of time, so it is not a fundamental countermeasure. do not have.

Another method used is to correct errors using computer calculations, but it is difficult to perform such correction calculations on a large number of input signals each time a measurement is made.
This forces the computer to do extra work, reducing the efficiency of data collection. In other words, this method cannot be called a fundamental countermeasure either.

The present invention eliminates the above-mentioned drawbacks and makes it possible to perform highly accurate measurements, and does not cause a significant increase in price, decrease in reliability, or decrease in operating rate or data collection efficiency. This is a feature. The present invention will be explained in more detail below.

Figure 1 is a diagram showing an example of the circuit configuration of a multi-channel A/D conversion circuit that has been used up to now. In detail, the part 18 surrounded by a broken line excluding the CPU in the figure is the A/D conversion circuit part, and the CPU 4 is the computer part. be. Analog signal input terminal A in the figure,
Each input signal applied to B, -...Y,z is input to an input amplifier IA, IB,...IY.

Each channel is amplified to an appropriate size by lZ. Relays 2A, 2B, .PHI. . . 2Y, 2z are channel connection/disconnection switches arranged for each channel, that is, a channel closing device group. In the case of FIG. 1, the measurement is carried out as follows. First, relay 2A is closed by relay open/close signal lO from CPU 4, and the analog signal of channel A is selected and sent to A/D converter 8. child\
The signal is converted into a digital signal and inputted to the CPU 4. Thereafter, relay 2A is opened by relay opening/closing signal lO from CPU 4, and the operation for channel A is completed.

Subsequently, relay 2B is closed, and the next digitized signal of channel B is input to CPU 4 in the same process. By repeating such operations, signals of all channels are input to the CPU 4.

This circuit has a simple configuration and is effective when high accuracy is not required for measurement. However, when high accuracy is required, various problems occur because the circuit itself generates errors and changes the input signal.

To explain this in more detail, the input amplifier IA.

1B, . . . IY, 1z generates some output signal (offset) even when the input signal is zero, and its magnitude changes depending on the ambient temperature and elapsed time. In addition, the input amplification 51
The amplification degrees of A, IB, . Furthermore, the contact resistance, thermoelectromotive force, etc. of each contact of the relays 2A to 2z change depending on the surrounding conditions, and these all cause errors.

Measures conventionally used to eliminate these sources of error and perform highly accurate measurements include the input amplifier IA-12
A method of adding an offset correction circuit to each of the IA-12 and adjusting the amplification so that the offset becomes zero, a method of adding an amplification adjustment circuit to each of the human power amplifiers IA-12 and adjusting the amplification, There are methods such as using a mercury relay, which is also stable, and if you implement these methods and make careful fine adjustments, the error will decrease. However, the major disadvantage is that the fine adjustment is done manually and requires a long time, during which time the side chambers are interrupted. Furthermore, if a correction circuit is added to each human power amplifier, the number of parts will increase significantly, which has the drawback of significantly increasing not only the price but also the failure rate. A more fundamental problem is that these measures are ineffective against changes in ambient temperature or fluctuations in error over time, which require fine-tuning again. becomes. That is, in order to carry out highly accurate measurements, it is necessary to frequently repeat fine adjustment and calibration operations, which has the major disadvantage of increasing the number of measurement steps and significantly lowering the operating rate of the measuring device.

As mentioned above, another conventionally used countermeasure is a method in which errors are corrected using computer calculations, but this method does not correct errors in each circuit by circuit, but rather corrects errors necessary for measurement operation. Insert a calibration operation at a certain frequency, accurately grasp the error of each channel at that time, store it in the computer memory, and correct the obtained measurement results by calculation according to the stored contents. It's a method.

This method is excellent because it does not increase the number of parts, cost, or failure rate much, and the operating rate of the measuring device is also good because the calibration work is done by computer within a short time without manual labor. However, when the scale of measurement increases and thousands or tens of thousands of measurements are processed online in real time in a very short period of time, the computer repeats the supplementary IF calculation for each channel every time the measurement is made. A major drawback is that the collection efficiency is reduced and the measurement cannot be completed within the required time.

Now, the present invention will be explained. FIG. 2 is a side view of the configuration of a multi-channel A/D conversion circuit embodying the present invention, and the portion excluding the CPU 4 is the A/D conversion section. In FIG. 2, respective input signals are applied to the analog signal input terminals A, B, . . . Y, z, and the circuit of FIG. 2 performs both calibration and measurement operations.

First, the calibration operation will be explained. First, when the calibration signal 11 is sent out from the computer 4, the input amplifier IA enters a state where the input signal is zero. Next, relay 2A is closed by a relay open/close signal from CPU 4, and the output of LA is connected to summing amplifier/variable gain amplifier (AMP) 5. The output of the amplifier 5 is applied to the A/D converter 3 to become a digital signal and input to the CPU 4. The output signal of the amplifier 5 at this time should originally be zero since the input signal is zero, but it becomes a certain value due to the addition of the offset error and the thermoelectromotive force error of the relay.

The CPU 4 writes this error output into one of the storage elements of the memory (RAMII').The contents of the memory are digital.
The signal is converted into an analog signal by an analog (D/A) converter 8 and fed back to the AMP 5 as an offset supplementary signal 14 in a direction where the output signal becomes zero. Next, the calibration signal 11 is sent from the CPU 4 to the human power amplifier IA-12, and the input amplifier IA receives the human power signal at full scale (full scale).
cale) state, and its output becomes a digital signal and is input to the CPU 4. Since this value includes an error due to variation in the amplification factor, the CPU 4 writes a signal corresponding to the error into the next memory element of the RAM 6.

The content is turned into an analog signal by the D/A converter 7 and is fed back to the AMP 5 as an amplification degree adjustment signal 15 so that the amplification degree becomes a specified value. Thereafter, the two relays are opened due to disconnection of the relay opening/closing signal 10 from the CPU 4, and the calibration operation of one circuit is completed.

Next, by applying the calibration signal 11 from the CPU 4 again, the input amplifier IB becomes in a state where the input signal is zero, and then the relay 2B is closed by the relay opening/closing signal IO, and the error correction failure is performed in the RAM 6 in the same process as above. The offset error is corrected, and then the gain is corrected.

Similar steps are repeated to complete the calibration of all channels.
do.

Next, the second measurement operation will be explained. First, relay 2A is CP
It is closed by the relay opening/closing signal lO from U4, and the analog signal of channel A is selected and input to AMP5.

At the same time, an address signal 12 is sent from the CPU 4 to the memory RAM 6 to select the storage element corresponding to channel A, and the stored error correction information 5 is sent to the D/A converters 7 and 8 for offset correction, which is an analog signal. This becomes a signal and an amplification degree adjustment signal and is sent to the AMP5. Therefore, the input analog signal is corrected for errors in the AMP 5, and a signal containing no errors is input to the A/D converter 8, which is then input to the digital signal output 1icpu4. At this point, the relay 2A is opened by the relay opening/closing signal IO from the CPU 4, and the measurement for channel A is completed. Next, relay 2 is activated by relay open/close signal lO.
When B is closed, the error-corrected signal is input to the CPU 4 in the same manner as above, and by repeating the above operation, the signals of all channels are input to the CPU 4.

During actual measurement, a calibration operation is first performed for each channel, and then a measurement operation is performed for each channel, but by inserting the calibration operation at an appropriate frequency into continuous measurement operations instead of performing them alternately, It is possible to suppress errors due to ambient temperature or the passage of time to a low range.

The above-mentioned feature of the method of the present invention is that the adjustment is automatically performed by commands from the CPU 4, whereas in the past, manual adjustment was required for measurements that required high precision, which required a long time. Therefore, it can be measured in an extremely short time.
Instead of adding a correction circuit to each channel as in the conventional method, one common part (two correction functions are consolidated) for all channels, so the circuit is simple and the number of parts does not increase. low failure rate and price increase,
Since the corrected signal is input to CPU4, the CPU
U4 is that there is no need to perform correction calculations, and therefore data can be collected efficiently. These features eliminate the difficulties of conventional methods and produce extremely large practical effects.

The example shown in FIG. 2 is a method in which the amplification adjustment signal is fed back to the summing amplifier/variable gain amplifier (AMP') 5, but in another embodiment, the amplification adjustment signal is fed back to the A/D converter 3. There is a method of controlling the gain of the A/D converter by feeding back the voltage to the reference voltage terminal of the A/D converter. It is clear that the latter method can achieve exactly the same effect as the former.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing a configuration example of a conventional multi-channel A/D conversion circuit, and FIG. 2 is a configuration example diagram of a multi-channel A/D conversion circuit according to the present invention. A-Z...Analog signal input, IA-12...Input amplifier, 2A~22-...Relay, 8...A/D converter,
4 am am = r 7 computers (CPU), 5...
・Summing amplifier and variable gain amplifier (AMP), 60-・
Memory element group (RAM), ? , 8...D/A converter, IO...Relay open/close signal, 11--Calibration signal, 12--Address signal, 18--Multi-channel A/D conversion Circuit, ■4-
...Offset correction signal, 15...Amplification degree adjustment signal. Representative: Ogaki, 1 person from outside the university

Claims (3)

[Claims] (1) A group of input amplifiers that amplify multi-channel analog input for each channel, and a group of channel switches that open and close according to switching signals from the computer and switch the output of the input amplifier group to the summing amplification/variable gain amplifier in channel order. , one addition amplifier and variable gain amplifier, and an A/D converter that converts the analog signal input from the addition amplifier and gain amplifier into a digital signal and sends it to an external computer;
A RAM that stores error correction information from the computer at a specified address using an address signal from the computer.
a memory, a first D/A converter that converts the offset correction signal output from the RAM memory into an analog signal, and an RA.
Equipped with a second D/A converter that converts the amplification adjustment signal output from the M memory into an analog signal, and sets the channel open/close signal input from the computer and the input signal of the input amplifier to zero and full scale states. A multi-channel analog-to-digital conversion circuit characterized by performing a measurement operation for each channel and a calibration operation inserted before several measurement operations using a calibration signal and a memory address signal, and sending the output to a computer. (2) The multi-channel analog-to-digital conversion circuit according to claim 1, characterized in that the analog offset correction signal and the amplification degree adjustment signal are applied to a summing amplification/variable gain amplifier. (3) Use the analogized offset correction signal to correct the addition amplification/variable gain amplifier, and use the amplification adjustment signal to
2. The multi-channel analog-to-digital conversion circuit according to claim 1, wherein the multi-channel analog-to-digital conversion circuit is provided for adjusting the reference voltage of the /D converter.