Practical Electron Microscopy and Database

An Online Book, Second Edition by Dr. Yougui Liao (2006)

Practical Electron Microscopy and Database - An Online Book

Chapter/Index: Introduction | A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z | Appendix

Electron Voltage Imaging (EVI) and Electron Voltage Probing (EVP)

Electron Voltage Imaging (EVI) and Electron Voltage Probing (EVP) are electron beam probing (eBP) techniques. Both of these techniques utilize the interaction between an electron beam and the surface of a device under test (DUT) to detect electrical signals and analyze the device’s behavior:

  • Electron Beam Probing: This technique involves directing an electron beam at a semiconductor device, generating secondary electrons that carry information about the local electrical properties (voltages) of the device. Electron beam probing allows non-contact measurement of electrical activity in devices, which is especially useful for advanced semiconductor nodes where physical probing would be difficult or impossible.
  • Electron Voltage Imaging (EVI): EVI is a form of electron beam probing that is primarily used to map signal frequencies across a device’s surface. It works by analyzing the modulation of the secondary electron signal as the device switches between different voltage levels. EVI applies a frequency-based approach using Discrete Fourier Transform (DFT) to visualize electrical activity across the device, which is useful for understanding clock signal integrity and diagnosing scan chains.
  • Electron Voltage Probing (EVP): EVP is another electron beam probing technique, but it focuses on capturing time-domain waveforms of electrical signals. EVP records voltage changes over time, allowing detailed tracing of signal paths within the device. This technique is particularly valuable for real-time analysis of signal behavior and for debugging complex signal integrity or timing issues.
The physical principles underlying both EVI and EVP are based on the same fundamental phenomenon, known as voltage contrast. This principle involves detecting surface potentials on a DUT by using an electron beam in a SEM. When the electron beam is directed at the surface of the DUT, the interaction of the beam with the sample generates secondary electrons. The emission of these secondary electrons is influenced by the local electric field or surface potential of the DUT. Areas with different potentials (voltages) will emit secondary electrons differently, and this variation in emission is what creates the voltage contrast signal.

EVI and EVP had been used to capture signals in modern-day integrated circuits, [1] for instance, timing analysis, particularly for low duty cycle signals, such as those used in power-saving architectures. In this way, timing and phase relations of signals can be measured using an electron beam, which is important for debugging design issues where signal integrity and timing are critical. 

Two-point transformation sampling scheme [1], where two electron pulses sample the signal at specific intervals to capture the timing and phase information, can be used to track and analyze the phase relation between the DUT and the electron beam.​ The two-point transformation sampling scheme is used to enhance signal detection in low duty cycle signals during electron beam probing. The method allows for more accurate and faster sampling of signals with varying duty cycles, particularly in circuits where traditional methods like Discrete Fourier Transform (DFT) might struggle due to low signal-to-noise ratios (SNR) at lower duty cycles:

  • Signal Sampling at Two Points: The scheme involves sampling the signal at two distinct points in time. One sample is taken at the peak of the signal, and the other is taken at a point where the signal is at its minimum or close to zero. The difference between these two samples is used to represent the signal at that location.
  • Improved Signal Detection for Low Duty Cycles: Low duty cycle signals are those where the signal is active only for a small portion of the total time (e.g., 4% of the time). These signals are often hard to detect using traditional methods like DFT because they generate weak spectral responses. The two-point transformation method improves the SNR for these low-duty cycle signals, making it easier to detect them.
  • Phase Locking for Accuracy: The system is phase-locked to the device under test (DUT), meaning the electron beam pulses are synchronized with the signal from the DUT. This allows for precise sampling at the correct points in the signal cycle.
  • Fast Data Acquisition: Since only two samples are needed to capture the essential features of the signal, data can be acquired much faster compared to traditional methods like DFT, which require more points to perform the transformation. This speed improvement allows for better averaging and even further enhancement of the signal-to-noise ratio.
  • Application: This method is particularly beneficial for detecting and analyzing signals in circuits that employ low duty cycle clocking schemes, such as those used for power-saving in devices like GPUs or systems employing power throttling mechanisms.

 

 

 

 

 

 

 

 

 

[1] Neel Leslie, James Vickers, Jennifer J Huening, Xianghong Tom Tong, Patrick Pardy, Electrical Event Capture with an Electron Beam Probing System, ISTFA 2023: Proceedings of the 49th International, https://doi.org/10.31339/asm.cp.istfa2023p0164, November 12—16, 2023, Phoenix, Arizona, USA.