The prototype was finally assembled, and the next step was real testing. They requisitioned an old, unused test bench in the corner of the project area, where a standard test spindle was fixed. Liu Li and Zhou Wei carefully installed their somewhat bulky "balancing device" near the drive end of the spindle.

Connect the wires, check, and power on.

The laboratory was so quiet you could hear each other breathing. Sun Mei stared intently at the oscilloscope, which displayed the signal from her homemade inductor probe; Zhou Wei's hand rested on the switch on the control box, where his optimized control algorithm was running; Liu Li stood beside the test bench, ready to respond to any unexpected mechanical issues.

"Start!" Zhou Wei pressed the button.

The main motor of the test bench emitted a deep hum, driving the spindle to begin rotating slowly. The speed gradually increased.

"500 RPM, signal stable," Sun Mei reported.

"1000 revolutions per minute, the vibration amplitude is starting to increase, which is in line with expectations." Zhou Wei stared at the vibration data on the monitoring screen.

"The balancing device is in normal standby mode." Liu Li checked the electromagnet and slider mechanism.

When the spindle speed climbs to nearly 70% of the operating speed, the vibration of the spindle intensifies significantly, and the amplitude of the waveform on the oscilloscope increases sharply.

"Vibration exceeds limits! Initiate the balancing program!" Zhou Wei ordered, while simultaneously manually activating the control algorithm.

In an instant, the electromagnets inside the device were energized and de-energized sequentially according to the algorithm's instructions, attempting to drive the small mass blocks to the calculated compensation positions.

However, the situation spiraled out of control in an instant!

The test bench emitted violent, irregular vibrations, accompanied by a piercing metallic scraping sound! The waveform on the oscilloscope was not suppressed; instead, it fluctuated wildly as if it had gone mad, with amplitudes several times greater than before! The entire device seemed not to be suppressing vibrations, but rather frantically amplifying them!

"Emergency shutdown!" Liu Li shouted.

Zhou Wei quickly cut off the power. The test bench shook for a few more seconds due to inertia before slowly coming to a stop. A faint, burnt smell of circuit overload filled the air.

The first bench test was a complete failure.

The three of them stood around the still slightly trembling experimental table, their faces grim.

"The control response is too drastic!" Zhou Wei spoke first, his tone urgent. "The compensation command given by the algorithm is too large and the timing is wrong. Instead of offsetting the vibration, it has created positive feedback and is 'resonating' with the vibration!"

“The probe signal itself has a lot of noise and obvious delay,” Sun Mei pointed to the raw signal recorded on the oscilloscope. “Based on such input, it is difficult for the algorithm to make an accurate judgment.”

“There’s also a problem with the mechanical parts,” Liu Li said as she inspected the inside of the device. She found that a mass block slider was slightly stuck under severe vibration. “The reliability of the moving parts is not high enough at high speeds.”

The problems were far more numerous and complex than they had imagined. Deficiencies in the mechanics, sensing, and control systems intertwined and amplified each other, leading to this disastrous outcome.

The laboratory was a mess, and the atmosphere was heavy. The first prototype they had painstakingly built proved to be so fragile and vulnerable in the face of reality.

"What should we do?" Sun Mei's voice carried a hint of frustration.

Liu Li stared at the failed device, remained silent for a moment, then bent down to pick up a screw that had fallen to the ground and clenched it tightly in her hand.

"What else can we do?" She raised her head, her gaze sweeping over Zhou Wei and Sun Mei. Though exhausted, she was unusually determined. "Find the problems and solve them one by one. Optimize the mechanical structure, adjust the control algorithm, and improve the signal processing. So what if we fail once? It's not like we haven't failed before."

Her words shattered the stagnant air like a stone. Zhou Wei vigorously wiped his face: "Yes! At least we know where the problem lies! Feedback delay, algorithm parameters, mechanical reliability... the direction is clear!"

Sun Mei took a deep breath and looked at the oscilloscope again: "I saved all the data from this failure, especially the signal characteristics before and after the loss of control. This is the most valuable analysis data."

The disastrous failure of the first experiment did not break them; instead, it was like a bucket of ice water, waking them from their initial blind optimism and making them more clear-headed and resilient as they threw themselves into the next round of even more arduous improvements.