I started with a ship full of max-level veterans, and I'm just a complete noob.

When Huang Kun removed the old parameter table, the first section of the atmospheric step curve of Furnace No. 3 had just passed the initial temperature, and the response delay had already exceeded the opening condition.

Jiang Ming kept the heating power constant until the readings of the two platinum resistance thermometers at the leading and trailing edges entered the stable region, and then increased the power to the second level.

Fang Xudong recorded the inlet air temperature, protective gas flow rate, and pipe pressure at fixed intervals. The three data points fell below the paper tape and were aligned with the furnace temperature curve on the same time axis.

After the outer wall of the quartz tube is heated, the leading edge sensor responds first, and the trailing edge sensor follows with a significant delay. The two curves eventually converge at a new equilibrium position.

Old Sun stood guard beside the propulsion mechanism. Although the propulsion was temporarily shut off during the no-load phase, he still turned the handwheel at the prescribed intervals to check whether the newly installed split nuts would become stiff after being heated.

Da Liu disconnected the actuator from the voltage regulator to maintain the step measurement in a pure open-loop state, while monitoring the amplifier output to prevent zero-point drift from being mixed into the thermal response.

After the three power steps were completed, Jiang Ming overlaid the heating curve with the natural cooling curve to separate the main delays of the heater, quartz tube and sensor sleeve.

The main time constant with the protective atmosphere is nearly twice as long as the old record, and the second hysteresis is concentrated between the ceramic sleeve and the platinum resistance.

"Yesterday, the controller thought the furnace temperature had changed, but the sensors were still sending back the previous change."

Fang Xudong used two pencils, one red and one blue, to mark the phase difference. The interval where the integral term continues to accumulate error falls exactly between the two lines.

Jiang Ming recalculated the critical gain and tightened the proportional action range to avoid the controller giving too large a power correction when facing delayed signals.

The integration time was increased to more than double the original time, and the rate of error accumulation decreased accordingly. After the furnace temperature exceeded the set value, the old error could exit the execution loop more quickly.

A set of RC network is added to the front end of the differential channel. Da Liu selects resistors and oil-immersed capacitors according to the calculated values. After soldering, a signal generator is used to check the cutoff range.

Fang Xudong superimposed a high-frequency small-amplitude disturbance at the input end of the bridge, and the actuator connecting rod, which had been swinging frequently, only made limited adjustments, and the furnace body could still pass through normally despite its slow changes.

Huang Kun stood next to the control panel with yesterday's records in hand. When he saw the RC network connected to the micro branch circuit, he reached out and flicked the test switch.

The actuator remains in place to short disturbances, but rotates gradually to continuous ramp signals, with the amplitude of the movement corresponding to the speed of change.

"How much is filtered out?"

"The rapid vibration in high-temperature measurement has been largely suppressed, and the furnace temperature change cycle is much longer than the filtering time, so the differential judgment can still be retained."

Huang Kun turned on the test switch and asked Fang Xudong to stick the two sets of records before and after filtering on the wall, while he himself wrote down the resistance and capacitance values ​​in a hardcover notebook.

Around 11 a.m., the time constant with atmosphere was calculated, and all three sets of corrected control parameters were locked.

Jiang Ming requested a re-inspection of the gland seal, verification of the protective gas from inlet to outlet, and that the flow meter float remain stable at the specified mark, while the pressure inside the pipe also remains within the process range.

Old Sun locked the transmission frame in the no-load test position, and gently pushed it a few times with his palm on the furnace frame. The quartz tube support and the sensor sleeve remained in their original positions.

"It can be powered on."

Da Liu closes the main switch, and the heating current rises along the predetermined slope. The automatic circuit is temporarily in a monitoring state, and will take over the control step by step after the furnace temperature exceeds 700 degrees Celsius.

The initial temperature rise curve was not much different from yesterday. After reaching 800 degrees Celsius, the controller began to reduce the power adjustment range each time, and the actuator action interval also increased accordingly.

Fang Xudong laid yesterday's oscillation curve above the new paper tape. The two lines separated around 890 degrees. The old curve continued to rise, while the new curve slowed down ahead of schedule.

At 910 degrees Celsius, the proportional circuit reduces the heating current, the integral term only compensates for long-term deviations, and the micro-branch continues to recover power according to the heating trend.

Once the furnace temperature exceeds 920 degrees Celsius, the actuator moves only a small distance each time. Although the sensor feedback has a delay, the accumulated error is already limited.

Lin Lanying stood beside the pyrometer and checked the optical readings against the temperature converted by the bridge one by one. At 930 degrees, the difference between the two was reduced to 0.12 degrees.

Jiang Ming adjusted the absolute temperature compensation knob to make the average value of the double platinum resistance thermometer coincide with the calibration point of the pyrometer, while the differential signal still independently controlled the temperature difference between the leading and trailing edges.

934 degrees, 935.2 degrees, 936.1 degrees, the slope on the paper tape gets smaller and smaller, and the regulator rod gradually approaches the equilibrium position.

When Fang Xudong reported 936.8 degrees, everyone in the lab focused on recording paper and instruments; even Lao Sun closed his toolbox.

When the furnace temperature reached 937.03 degrees, the controller recovered a small amount of power, and the curve rose to 937.07 degrees after crossing the set line.

The sustained upward surge of yesterday did not reappear. The integral term weakened rapidly after the proportional phase recovered, and the temperature line traced a narrow arc above the set value.

The curve then dropped to 936.94 degrees, the actuator supplied a small amount of power, and the next peak value stopped at 937.05 degrees.

Fang Xudong slowed down the paper tape speed by one level, expanding the time range, and marked the converted temperature next to each peak and valley with a red pencil.

"In the first twenty minutes, the highest point was +0.07 and the lowest point was -0.06."

Huang Kun looked at the data and raised his hand to stop the duty officer from approaching the control panel.

"Don't let anyone adjust it, just let it run on its own."

Lunch boxes were delivered to the door, and everyone took turns eating a few bites. Boiler No. 3 continued to run, and the gas flow and pressure inside the pipe were checked at fixed intervals.

One hour later, the furnace temperature reached a high of 937.08 degrees Celsius and a low of 936.93 degrees Celsius, with the fluctuation range remaining within the control limit.

Jiang Ming checked the actuator's action record and found that the long-term deviation had been gradually eliminated by the integral stage, the proportional correction remained narrow, and the micro-branch was not affected by high-frequency noise.

Before 3 p.m., the outside air temperature rose slightly, and the furnace temperature curve deviated upward by 0.02 degrees. The control loop then pulled the deviation back to near the set line.

Lin Lanying compared the old furnace manual control records with the records. In the past, the operator adjusted the transformer every few minutes, and the temperature around 937 degrees usually had to be increased by more than one degree.

Nowadays, the ink lines on the paper tape follow the set values ​​closely. If the scale were not magnified, the peaks and valleys would almost form a strip of limited width.

At 4 p.m., Fang Xudong replaced the paper tape with a calibrated one and marked this moment as the starting point for formal stable operation.

Jiang Ming sealed off the proportional, integral, and differential knobs one by one, and fixed the parameter positions on the control panel with lead-lined tape. Any adjustment had to be re-registered.

At the end of the first hour, the highest deviation was +0.07 degrees Celsius and the lowest deviation was -0.08 degrees Celsius, and the temperature difference between the front and rear edges remained within the specified range.

During the second hour, the protective gas flow rate experienced a slight decrease. The operator simply restored the flow rate according to the process card, and the controller absorbed the subsequent thermal disturbance on its own.

At 6 p.m., Fang Xudong pressed the time mark and cut out the paper tape for two consecutive hours, calculating the temperature at the beginning and end and all the peaks and valleys in between.

"Nine hundred and thirty-seven degrees, for two consecutive hours, with a maximum fluctuation of plus or minus 0.08 degrees."

He placed the recording paper in the center of the long table, and Lin Lanying brought the standard ruler from the Institute of Physics to recheck the paper speed, temperature scale ratio, and pyrometer correction value.

Da Liu removed the temporary test line and tightened the terminals of the formal line again, while Lao Sun checked the transmission nut after it was heated, and the handwheel could still turn smoothly throughout the entire process.

Huang Kun picked up the record sheet and read from the starting mark at 4 p.m. to the ending line at 6 p.m., only flipping through the pyrometer calibration table once during that time.

Three minutes later, he laid the record sheet flat on the table, pressed down the corner of the paper with his palm, and turned to look at Lin Lanying.

"Install the germanium rod."

Lin Lanying took out the schedule for next week's No. 3 furnace and crossed out the originally scheduled tests for the rectifier group, single crystal group, and equipment group one by one.

Huang Kun took the pen and wrote "502 Technical Group" for all the furnace positions from Monday to Saturday.

After writing the last section, he clipped the schedule back to the wooden board, and the small box containing germanium material next to Furnace No. 3 was placed next to the new process card by the duty officer.