JPH0546562B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to a current adder circuit that drives a load with an intermediate value of multiplexed control device outputs.

[Technical Background of the Invention] In recent years, control devices used to control processes that require high reliability are often multiplexed for the purpose of improving the reliability of the control devices themselves. A multiplexed configuration has a regular device and a standby device, and there is a method of using the standby device in the event of a failure of the regular device, and a method of constantly operating multiple control devices and obtaining the most reliable representative value from the outputs of those control devices. A selective control method is used. Generally, the former method includes
A fault detection circuit that detects equipment failures is essential, and switching from regular equipment to standby equipment is performed when a fault is detected by the fault detection circuit, so the fault detection circuit is required to have high fault detection capability. Ru. However, it is impossible to detect all failures, and the higher the failure detection ability, the more complex the failure detection circuit becomes, and failures of the failure detection circuit itself become a problem.

On the other hand, the latter method requires at least three control devices to select the most likely representative value, and also requires a selection circuit to select the representative value. If the device is triplexed, an intermediate value selection circuit is usually used as the selection circuit. In this case, the intermediate value selection circuit selects the intermediate value of the outputs of the three control devices as a representative value and drives the load to be controlled, so even if one control device fails, control will continue normally. becomes possible. However, even with this method, there is only one intermediate value selection circuit, so if the intermediate value selection circuit fails,
Normal control is no longer possible. That is, no matter how many control devices are multiplexed, the reliability of the selection circuit that selects the most likely representative value from among the outputs of each control device and narrows it down to one becomes a problem. For this reason, it is necessary to increase reliability by simplifying the selection circuit as much as possible and reducing the number of components, and a selection circuit commonly used for this purpose is a current addition circuit.

FIG. 3 is a configuration diagram showing an example of the configuration of a conventional current adding circuit. The outputs _S1 of the three control devices 1 (subscripts A, B, and C are added in the figure to distinguish them, and the same applies to the others) are input to the isolation amplifier 2, and the outputs of the isolation amplifier 2 S ₂ is converted into current output S ₃ by current converter 3, and each current output
S _3A , S _3B , and S _3C are all connected to one end of the load 4 .
The other end of the load 4 is connected to a feedback resistor 5 common to each current converter 3, and the terminal voltage _S5 of the feedback resistor 5 is fed back to the current converter. isolated amplifier 2
Although not necessarily required, it is usually provided to electrically isolate the control device side and the load side and to operate the control device 1 stably against external noise and the like.

Next, a conventional configuration example shown in FIG. 3 will be explained. The operation of isolation amplifier 2 and current converter 3 is as follows (1)
It is shown by the formula and the formula (2).

S ₂ =S ₁ (1) S ₃ =K ₃ (S ₂ −S ₅ ) (2) Note that K ₃ in the above equation (2) represents the gain of the current converter 3. Next, if the current supplied to the load ₄ is S4, the following equation (3) is obtained.

S ₄ = S _3A + S _3B + S _3C ……(3) Furthermore, the terminal voltage S ₅ of feedback resistor 5 is the feedback resistance value.
When it is Rf, it is expressed by the following equation (4).

S ₅ = RfS ₄ ...(4) From the above equations (1) to (4), each control device output S _1A , S _1B ,
The relationship between S _1C and load current S ₄ is as shown in equation (5) below.

S ₄ =K _3A S _1A +K _3B S _1B +K _3C S _1C /Rf(K _3A +K _3B +K _3
C )...(5) [Problems with the background art] However, in general, the characteristics of the current converter 3 are
As shown in the figure, it has saturation characteristics, and when the possible range of current converter output _S3 is 0 to 100 (%),
The linear range in which the above formula (2) holds true is only 100/K ₃ (%) in S ₂ -S ₅ . Therefore, in order for the above formula (2) to hold true, it is necessary that |S ₂ −S ₅ |<100/K ₃ (%) ...(6) hold true. If |S ₂ −S ₅ |
If K deviates from the range of equation (6) above, K ₃ becomes extremely small as shown in FIG. When the conditions for the above equation (6) to hold are calculated using the above equations (2), (3), and (5), at least |S _1A −S _1B |<100/K ₃ (%) |S _1B −S _1C ｜<100/K ₃ (%) ｜S _1C −S _1A ｜<100/K ₃ (%) ...Two of the following equations (7) need to hold true. Therefore, for example, if the control device 1c fails, the control device output S _1C becomes
If the range of equation (7) is exceeded, K _3C ≒ as described above.
It becomes 0. Therefore _{, from} the above formula ( _{5), the fluctuation rate for S 4 ( S 4 −S 4 ′} ) _/ S ₄ is the above
From equations (5), (7), and (8), (S ₄ −S ₄ ′)/S ₄ <1/3×100/K ₃ (%) ...(9). Generally, the gain K ₃ of the current converter 3 is large, and if we assume that K ₃ = 100, the control device output
No matter how S _1C fails, the load current S ₄ is 0.3
(%) or less, which means that sufficient selection has been made. It is also clear that the larger K ₃ is, the better in order to make a more complete selection.

However, as K ₃ increases, the linear range of the current converter 3 becomes narrower as described above, and even when all the control device outputs S ₁ are normal, the two The output S ₃ of the current converter 3 will be saturated. Now, a case where the control device output S _1B fails while the current converters 3 _A and 3 _C are saturated will be explained based on FIG. 5. At time t ₁ in Figure 5
If S _1B fails in the decreasing direction, the current converter output
S _3B goes in the negative saturation direction, and naturally the load current S ₄ also decreases. As a result, S _1C −S ₅ decreases, and the current converter output S _3C , which had hitherto been negatively saturated, recovers, and finally |S _1C −S ₅ |<100/K ₃ (%) is satisfied. The load current _S4 settles within the range. However, since the current converter 3 generally has poor response characteristics in the saturation region compared to the linear region, there is a non-negligible time delay (indicated by Δt in the figure) before it recovers to the linear state.
The current converter output S _3C does not recover instantly. Therefore, the load current S ₄ remains fluctuating during Δt. Therefore, even if one series of the control device 1 fails, the load current will temporarily fluctuate, which is unfavorable in terms of control.

To avoid this problem, the current exchanger gain
It is sufficient to make K ₃ small so that all current exchangers 3 enter a linear state within the normal variation range between the control device outputs S ₁ . However, if K ₃ is made small, the selection characteristics will deteriorate due to the above equation (7). Therefore, in the prior art, the control device 1
Reducing transient fluctuations in load current when one series fails and improving steady-state selection characteristics are contradictory demands, and it is impossible to satisfy both. Furthermore, since the feedback resistor 5 is single-layered, and a failure of the feedback resistor causes a failure of the whole, the reliability of the whole is determined by the reliability of this feedback resistor, and the reliability of the whole is greater than the reliability of the feedback resistor. There were problems such as the inability to raise the

[Object of the invention] The present invention solves the problems of the prior art described above,
It is an object of the present invention to provide a highly reliable current adding circuit that achieves both reduction of load current fluctuation when one control device series fails and sufficient selection characteristics.

[Summary of the Invention] For this reason, the present invention provides multiplexed amplifiers and current exchangers corresponding to the multiplexed control device, and each amplifier has a respective control device that responds to the deviation between the respective control device output and the actual load signal. A signal obtained by adding the deviation between the amplifier output and the intermediate value of each amplifier output is input, while each multiplexed current converter is provided with a feedback resistor for each amplifier output and the load signal is fed back. It is characterized by what it did.

[Embodiment of the Invention] FIG. 1 shows the configuration of a current adding circuit according to an embodiment of the present invention. In the figure, the same reference numerals as in FIG. 3 indicate the same or corresponding parts. The difference from FIG. 3 is that the load drive signal (hereinafter referred to as the control device output) S 1 and the load current are output from the control device 1 to drive the load 4 between the first isolation amplifier 2 and the control device ₁ . A first arithmetic circuit 7 is provided that calculates the deviation from the actual load signal S ₆ obtained after insulating S ₄ by an insulating amplifier 6, and the output of this first arithmetic circuit 7 is passed through an amplifier 8 to the first insulator. An intermediate value selection circuit 9 is provided which inputs the input point to the amplifier 2 and the output S ₈ of each amplifier 8, and selects an intermediate value thereof, and outputs the intermediate value selection circuit.
The deviation between _S9 and the amplifier output _S8 is calculated by the second arithmetic circuit 10.
, and the output of the second arithmetic circuit 10 is added to the first arithmetic circuit 7 as a correction signal, and the feedback resistor 5 is provided for each current converter and multiplexed. Note that COM1, COM2, etc. in the figure indicate independent power supply commons.

Now, assuming that the gain of the amplifier 8 is _K8 , the range from the control device output _S1 to the first isolated amplifier output _S2 is expressed by the following formula.

S ₈ = K ₈ {S ₁ − S ₆ + mid(S _8A , S _8B , S _8C ) − S ₈ } ...(10) S ₂ = S ₈ ...(11) Here mid(S _8A , S _8B , S _8C ) are each signal S _8A , S _8B ,
S Represents the intermediate value of _8C . The equation from the first isolated amplifier output S ₂ to the current converter output S ₃ is the same as equation (2). However, each current converter 3 has an independent power supply common, and one end of each feedback resistor 5 is also connected to the same independent power supply common as each current converter. Although the output current is once added and supplied to the load 4, it does not flow into the commons other than the current converter that outputted it, but always flows into the common of the current converter itself that outputted it, so the feedback resistor 5
What is fed back to the current changer 3 is the terminal voltage of the feedback resistor corresponding to the current value outputted by itself. Therefore, S ₅ =RfS ₃ ...(12). On the other hand, the actual load signal detected by the second isolation amplifier 6 corresponds to the load current and is expressed by the following equation.

S ₆ = RlS ₄ ... (13) S ₄ = S _3A + S _3B + S _3C ... (14) Here, Rl is the resistance value of the load 4.

First, from equations (2) and (12), S ₃ =S ₂ /Rf (15). This indicates the input/output characteristics of current converter 3, but since it is not included in (15), K ₃ is a large value and the output S ₃ does not saturate and is always stable relative to the input S _2. It is linear. On the other hand, (10) and (11)
From the formula, for example, if S _2A = mid (S _2A , S _2B , S _2C ), then S _2A = K _8A (S _1A − S ₆ ) ...(16) S _2B = S _1B − _{S 6} + S _2A = S _1B +K _8A S _1A −K _8A S ₆ ...(17) S _2C = S _1C −S ₆ +S _2A = S _1C +K _8A S _1A −K _8A S ₆ ...(18). Therefore, from equations (13), (14), (16), (17) and (18),
_S1A ,
The relationship between S _1B , S _1C and S ₄ is as shown below.

S ₄ = S _1A + αS _1B + βS _1C /Rl ……(19) Here It is. Therefore, when examining the degree of influence when the output S _1C of the control device 1C becomes abnormal, the load current S ₄ when S _1C is abnormal, that is, when S _1C = 0 in equation (19)
When the value of is S ₄ ′, it can be seen that (S ₄ −S ₄ ′)/S ₄ <1/3×100/K ₈ (%) ……(21). Therefore, if K ₈ is set to a large value, sufficient selection can be made as in the conventional configuration example. Also, from the above formulas (17), (18) and (19), K ₃
If is large, S _2B = S _2A + (S _1B − S _1A ) S _2C = S _2A (S _1C − S _1A ) ...(22). Therefore, |S _1B −S _1A | and |S _1C −S _1A |
If the difference is within the range of normal control device output S ₁ , then S _2B ≒ S _2C ≒ S _2A , and the outputs S ₂ of each current exchanger have approximately the same value. Also, for this reason, saturation naturally does not occur.

In this way, even if the gain K ₈ of the amplifier 8 is set to a large value, saturation does not occur anywhere on the circuit, it is always in the linear region, and the current converter output S ₄ is balanced throughout. Therefore, even if, for example, the control device output _S1A becomes abnormal, it will be instantly corrected by other series as shown in FIG. Moreover, for this reason, the value of K ₃ can also be made larger than in the conventional configuration example, and it is possible to provide better selection characteristics than in the conventional configuration.

[Effects of the Invention] As described above, according to the present invention, it is possible to achieve both reduction in load current fluctuation when one control device series fails and sufficient selection characteristics. Furthermore, since the feedback resistors are also triplexed, failure of one feedback resistor does not cause a failure of the entire system, greatly improving reliability.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of a current adding circuit according to an embodiment of the present invention, FIG. 2 is a signal diagram showing the operation of the current adding circuit of FIG. 1, and FIG. 3 is a diagram showing the configuration of a conventional current adding circuit. 4 is an input/output characteristic diagram showing the operation of the current exchanger of FIG. 3, and FIG. 5 is a signal diagram showing the operation of the current adder circuit of FIG. 3. 1... Control device, 2... First isolation amplifier, 3
...Current converter, 4...Load, 5...Feedback resistor,
6... Second isolation amplifier, 7... First arithmetic circuit, 8... Amplifier, 9... Intermediate value selection circuit, 10
...Second arithmetic circuit.

Claims (1)

[Claims] 1. Calculates a deviation signal between a load drive signal and an actual load signal that each multiplexed control device performs control calculations and outputs in order to drive a load, and adds a correction signal to this deviation signal. a multiplexed first arithmetic circuit that adds and outputs the output; a multiplexed amplifier that amplifies the output of each of these first arithmetic circuits; each has its own feedback resistance, and the output of each of the amplifiers is a multiplexed current converter that converts into a current signal; a multiplexed intermediate value selection circuit that simultaneously inputs the outputs of each of the amplifiers and selects an intermediate value thereof; a multiplexed second arithmetic circuit that respectively calculates the deviation of the output of the value selection circuit and outputs each of these deviation signals as the correction signal to the first arithmetic circuit; At the output of each current converter,
Further, the other end of the load is connected to a feedback resistor of each of the current converters, and a signal corresponding to the current flowing through the load is added to the first arithmetic circuit as the actual load signal. circuit.