1. Intro
I independently diagnosed and restored our lab’s Kurt J. Lesker sputter deposition system, which had been inoperable since 2019 due to unresolved mechanical and software faults. With no institutional knowledge remaining, I volunteered to take on the challenge based on my experience with equipment repair and systems troubleshooting.
2. How I Fixed It
The restoration process spanned approximately 30–40 hours over two months, unfolding in multiple stages:
Software Access & System Familiarization
- Gained administrative access to the control laptop, which required navigating system software and bypassing root file restrictions.
- Located and interpreted the startup scripts, vacuum interlock logic, and communication protocols between the system control software and the physical devices.
Power Supply and Vacuum Startup Issues
- Identified that the power supply would not initialize due to failed vacuum threshold readings.
- Learned the operational sequence for pulling vacuum using a dual-pump setup (roughing pump and turbo pump).
- Used oscilloscope and multimeter readings to monitor signal handshakes, validate power traces, and cross-reference against the startup logic in the code.
Mechanical Investigation
- Performed mechanical teardown and inspection of:
- The turbo pump, ensuring it was spinning up and sealing correctly.
- Both DC and RF sputter heads, which required disassembling shielding and argon gas lines.
- Flowmeters and solenoid valves, to verify whether required argon flow rates were achieved.
Root Cause Discovery
After weeks of signal tracing, hardware inspection, and software mapping, I identified that the system was halting due to insufficient coolant flow, which prevented interlocks from progressing to the "power-on" state. Upon further disassembly, I discovered:
- The DC sputter head was heavily corroded and partially clogged, which throttled coolant flow below threshold.
- The flowmeter, though functional, was giving correct readings that triggered a valid shutdown response due to this restriction.
Replacing and cleaning the flow path components allowed the system to proceed through its interlock logic and fully boot for the first time in years.
