PROCEEDINGS

of the Tomsk Order of the Red Banner of Labor Polytechnic Institute named after S. M. Kirov

Vol. 171
1969

AUTOMATED PROBLEM MACHINE

Yu. K. Petrov, V. M. Razin

(Presented by the scientific seminar of the Faculty of Automation and Computer Engineering)

At the Department of Mathematical and Computing Instruments and Devices of the Faculty of Automation and Computer Engineering of the Tomsk Polytechnic Institute, an automated problem machine has been built. The device is designed for a series of 15 problems. Each problem, together with four answers, one of which is correct, is printed on 24×36 film and inserted into the device. With such an arrangement, rapid replacement of the instructional material is ensured. The problem and four answers are projected onto a screen, and the student is invited to press the button with the number of the correct answer. After the button is pressed, the machine advances the frame and gives a signal of a correct answer if the answer is indeed correct. Confirmation of the correct answer by a signal is an instructional factor, which makes it possible to call the device a tutor. The machine gives, on a five-point system, a grade for the solution of all 15 problems. The device includes a printing unit that records the time spent solving the problem, the number of the correct answer, and the number of the button pressed by the student. The printed information makes it possible to draw conclusions about the degree of mastery of each of the 15 problems, about the correctness of the formulation of the problem itself, and about the quality of the lecture material on which the solution of these problems is based. It is possible to compile a program of instruction in a strict logical sequence, so that the solution of subsequent problems is conditioned by mastery of the preceding material. The printing unit makes it possible to work through such programs on the basis of statistical documentary material.

Existing teaching machines require the participation of an instructor in the learning process. The automated problem machine is intended to operate under conditions in which no instructor is present. The instructional material is supposed to be replaced as it is mastered, on the basis of the printed information. A photograph of the automated problem machine is shown in Fig. 1.

The automated problem machine was built as part of state-budget work, to a considerable extent by students of specialty 0608. In the course of manufacturing the automated problem machine, two diploma projects were completed and six course projects were defended. The work on manufacturing the automated problem machine was financed by the institute’s scientific department.

The principal elements from which the device is assembled are: a 6N1P electronic tube, a TG1 0.1/1.3 thyratron, and RPN-type relays. The device has stabilized power supplies of +250 V and −170 V. For powering the relay windings, an unstabilized 24 V direct-voltage source is provided.

The electronic part of the device is mounted in 20 cells; the cells are standardized as far as possible. The cells contain eleven symmetrical flip-flops of the time counter and the counter of correct answers. Ten controlled thyratrons are used to drive the printing device. At the moment of printing, depending on the magnitude of the control potential, a thyratron is either ignited or not ignited. If a thyratron is ignited, then through the relay winding a capacitor of capacitance 0.5 μF, previously charged to 250 V, is discharged; the relay armature raises the needle, and a dot is marked on the tape.

Fig. 1.

The number of the correct answer and the indication of the fact that an instruction on the rules of operation with the automatic problem-setter is projected onto the screen of the device are encoded by means of light and dark spots, which are located on the left side in each frame. The light code is projected onto photoresistors.

The electronic part of the device contains circuits that convert the light code into voltages of 100 V and 200 V, depending on the illumination of the photoresistor.

The electronic part of the device contains two multivibrators, which operate in a self-oscillating mode and have an oscillation period equal to 3 sec. Each multivibrator consists of a capacitor and a thyratron, which is ignited as soon as the voltage on the capacitor reaches a certain value. In the multivibrator mode, cold-cathode tubes may also operate; however, in devices with a long relaxation period, preference should be given to hot-cathode thyratrons, since in the latter there is a greater difference between the ignition voltage and the extinction voltage, thereby increasing the charging time of the capacitor. One of the multivibrators sets the time intervals and, together with the time counter, consisting of 7 flip-flops, constitutes the electronic clock. The other multivibrator in the control system produces a “psychological pause,” during which the learner can clearly examine the signal of the correct answer or the mark. In addition to the devices listed, the electronic part of the device contains inverters, pulse amplifiers, monostable multivibrators, and two decoders on semiconductor diodes.

The functional diagram of the automatic problem-setter is shown in Fig. 2. In Fig. 2, the numeral 1 denotes the control device; 2 — projection device; 3 — electronic clock; 4 — printing device; 5 — decoder of the number of the correct answer; 6 — decoder of the number of the pressed button; 7 — push-button device; 8 — decoder of the correct answer; 9 — device issuing the correct-answer signal; 10 — indicator of the end of the teaching program; 11 — counter of correct answers; 12 — decoder of marks; 13 — mark-issuing device.

For the projection device, the objective lens, condenser, and reflector of an ordinary photographic enlarger are adapted. The frames with the problems are inserted into 16 pockets arranged around the perimeter of an aluminum disk, which, by means of a solenoid drive, is rotated by one sixteenth of a full revolution; in this way the frame is advanced. A rubber tube is fitted onto the shaft with the disk, ratchet wheel, and gear for fixing the frame. With the aid of the rubber tube, simultaneously with the advance of the frame, the tape is pulled forward by four millimeters, and thus the position is prepared for printing information about the solution of the next problem.

The printing device outputs information in the form of binary numbers. The presence of a perforation on the tape corresponds to the code for one; the absence of a perforation on the paper tape corresponds to the code for zero. The two leftmost positions on the paper tape are intended for the number of the correct answer; the next two positions represent the number of the pressed button; the next five positions correspond to the time for solving the problem; and the last tenth position separates the series of 15 problems that make up the complete program.

Fig. 2.

The decoder of the correct answer consists of four coincidence circuits with two inputs, assembled on D7Zh semiconductor diodes. The information at the input of the decoder is supplied from electronic circuits with photoresistors in a paraphase potential code. At the corresponding input of the decoder for the number of the correct answer, a voltage of 250 V is set; at the remaining outputs of the decoder, a voltage of 90 V is set. The circuits with photoresistors create control potentials for controlling the tyratrons that print the number of the correct answer.

The decoder of the number of the pressed button is built on the contacts of an RPN relay and controls the tyratrons that print the number of the pressed button.

If the number of the correct answer coincides with the number of the pressed button, then the high voltage from the decoder of the correct answer ignites the SG-2 gas-filled tube, and in this way the signal of a correct answer is issued. If the number of the pressed button and the number of the correct answer do not coincide, then the low voltage of the corresponding decoder cannot ignite the gas-filled tube, and the signal of a correct answer is not issued. At the moment the gas-filled tube is ignited, a positive pulse arises across its cathode resistance; this pulse starts a KIPP relay, by the trailing edge of whose positive pulse a unit is recorded in the counters of correct answers. Information from the counter of correct answers, by means of inverters, is converted into paraphase code and is fed to the input of the diode decoder of the grade. The diode decoder of the grade is assembled on 8 D7Zh diodes and controls four gas-filled tubes, which correspond to grades 2, 3, 4, 5. The decoder issues grade 3 if the number of correct answers is greater than 7, grade 4 if the number of correct answers is greater than 11, and grade 5 if the number of correct answers exceeds 13. In any state of the counter, a high voltage appears only at one of the four outputs of the decoder and igni-

one of four gasotrons lights up, and thus one of four marks is issued.

The control unit performs the following operations: it stops the time count, issues the correct-answer signal and the mark for 3 sec, extinguishes the mark after three seconds, produces a pulse to ignite the thyratrons, extinguishes the marks, advances the frame, removes the relay from self-locking, and switches on the time count. The control unit includes a multivibrator, two KIPP relays, two pulse amplifiers, four thyratrons controlling the operation of the relays, and a device with a photoresistor for indicating the end of the training cycle.

The control unit is started in four different ways and performs three work cycles that differ from one another. If the learner presses one of the buttons for the correct answer, the corresponding relay contacts disconnect the clock multivibrator from the power source and switch off the control-unit multivibrator. Other relay contacts connect the output of the correct-answer decoder with the same number to the anode of the gasotron, and the correct-answer signal is issued. After three seconds the multivibrator thyratron is ignited. A positive pulse from the cathode resistance of the thyratron ignites the printing thyratron and the KIPP relay with a relaxation period equal to 500 msec. This time is needed for the printing relay to operate. The trailing edge of the positive pulse of the KIPP relay is amplified and used to start the thyratron of the drive of relay 8, which unlocks the entire relay system, the thyratron of the solenoid drive of the KIPP relay for extinguishing the registers. A narrow positive pulse from this KIPP relay is amplified and applied to the zero inputs of the triggers. For the solenoid drive a capacitor of 150 μF is used, which discharges through the thyratron and the solenoid winding.

The control unit can be started by means of the “start” button or when the clock overflows. In this case, at the moment when the last seventh trigger of the time counter is set to zero, a pulse is generated which starts the thyratron of the drive of the corresponding relay. In both cases the entire cycle of the control unit is carried out, except for printing and issuing the correct-answer signal.

The fourth work cycle of the control unit differs sharply from the three listed above. If an instruction is projected onto the screen, then by means of a special light signal a circuit with a photoresistor is actuated, which produces a potential permitting ignition of the thyratron of the drive of the seventh functional relay. The thyratron is ignited by a pulse from the first clock trigger and switches on the seventh functional relay, consisting of two independent relays. The relay cannot be started directly from the photoresistor; otherwise false operations will occur during frame advance. The seventh relay, with its contacts, connects the output of the register-extinguishing amplifier to the zero outputs of the triggers of the correct-answer counter, disconnects the control input of the solenoid-drive thyratron, permits the issuing of the mark, and starts the multivibrator of the control circuit. As a result of the operation of the control circuit, the information in the correct-answer counter (and in the clock counter) is extinguished, one half of the seventh relay is removed from self-locking, and at the same time the mark is extinguished. The contacts of the two halves of relay 7 form a logic circuit by means of which a pulse for igniting the thyratron of the unlocking relay is supplied only in the case when both halves of relay 7 are in the same state. The control circuit places the halves of relay 7 in different states, and thereby useless operation of the unlocking relay is prohibited.

Let us explain some features of the fourth cycle of the control circuit. The device gives a grade for the solution of all 15 problems of the training program; therefore, during 15 cycles of operation of the control circuit, information must accumulate in the correct-answer counter, and in the 16th cycle, when an instruction is projected onto the screen, the information in the correct-answer counter is erased.

The grade decoder is permanently connected to four nixie tubes, but since a large positive potential is applied to the cathodes of the nixie tubes through the contacts of relay 7, the nixie tubes do not light. At the moment when relay 7 operates, the inhibiting potential is removed and the grade signal lights up. As soon as one half of relay 7 is released from self-locking, the inhibiting potential is restored and the grade signal goes out. The problem-machine spares the students’ self-esteem: it does not leave a poor grade on public display, but erases it after three seconds.

If a student has stopped solving a problem somewhere in the middle of the program and has forgotten to switch off the machine, then the machine itself advances the frames and stops in the position in which an instruction is projected onto the screen.

A machine operating without service personnel must possess the greatest possible simplicity of control. All control of the problem-machine consists in pressing buttons, while certain psychological factors are taken into account. Suppose a student has pressed a button and has not received a signal indicating a correct answer. In such a case he may be inclined to press another button immediately; however, such actions will not alter the operation of the control device: the buttons are interconnected by means of relays in such a way that switching on any one of them blocks the operation of all the others.

It is assumed that the problem-machine will replace a course of practical classes in one of the subjects. However, the possibilities of the problem-machine are considerably broader. With the aid of this device one can study the process of thinking itself. Thus, with the aid of the printing mechanism, a large amount of statistical material can be collected on the time required to solve a given problem. On the basis of this statistical material, certain correlations may be found between the difficulty of a given problem and the time required to solve it, and conclusions may thus be drawn about the course of the thought process in solving problems.

In conclusion, the authors express their gratitude to the students of the Tomsk Polytechnic Institute, specialty 0608: V. I. Mitkovsky, B. S. Nedelkin, E. P. Parfenov, P. A. Sedelnikov, T. V. Serbina, Yu. T. Serbin, and G. P. Sukhinina, who took an active part in the construction of the problem-machine.

REFERENCES

1. I. I. Grebel, A. M. Dovzhenko. Automatic Devices for Instruction, Kiev, 1965.