Skip to content

Software Guide

The SanjANALYZER main window is organized into labeled regions A through I. The sections below describe each region and its controls.

[diagram/image in source — not reproduced here: Figure 1, the annotated Thermal Image Analyzer window layout]

(A) Load a Thermal Image

SanjANALYZER works with two file types, depending on whether the imaging session was Steady State or Transient:

  • Steady State data files use the .fdi extension.
  • Transient data files use the .tti extension.

Both file types are the raw data from which thermal images, cross-sections, and area data are generated, and both appear in the file selection control.

To load a thermal image, double-click the file that contains the image of interest. The CCD image and the thermal image load and appear in their respective image display windows.

(B) CCD Image Display

The CCD image display shows the image of the device. Three image tools are available:

  • Hand (grab) tool — moves the image around within the display window.
  • Zoom tool (magnifying glass) — zooms in on close-up areas of the image. Shift-click to zoom back out.
  • Area select tool (square) — used with the merged image to crop the thermal image to the region of interest.

(C) Primary Thermal Image Display

The thermal image display shows the view selected from the tab controls at the top: Thermal Image, Merge, or 3-D VIEW.

Thermal Image view. This is a calculation of thermoreflectance data using the masking information or the single coefficient you selected. The ROI mean and thermal cross-sections are generated from this view.

Merge view. The merged image (Figure 2 in the source guide) is for display and saving only — the underlying temperature values live in the thermal image, not the merged output. The merge tool converts the data to a single-precision RGB image for compatibility with standard image formats, and uses three sliders:

  • The first slider sets the percentage of CCD image overlaid on the thermal image, which helps pinpoint the exact location of hotspots.
  • Gain and offset sliders both apply to the thermal image. At 100% gain / 0% offset, the display is unchanged relative to the thermal image; adjusting them lets you emphasize different features.

Additional Merge view controls:

  • Crop from CCD Image checkbox — crops out the part of the thermal image far from your region of interest. Select the rectangle tool on the CCD image and draw a box around the region to keep. This is especially useful for AVI movie files, since cropping reduces file size by eliminating portions of the image without thermal information.
  • Fix Temperature Scale checkbox — used mainly when creating an image series movie or output images. With this checked, the previously set scale does not change when a new file loads. Recommended workflow for an image series: find the hottest image in the series, set the scale to the desired level, check this box, go back to the first image in the series, then run the process series tool. The scale is then fixed across the whole movie instead of auto-scaling per image.
  • Show Voltage / Delay / Color bar checkboxes — display the acquisition conditions and the color-to-temperature mapping on the image.

3-D VIEW. SanjANALYZER includes a 3-D plot of the thermal data (Figure 3 in the source guide). Select the region of interest in the Merge view tab before viewing the 3-D plot. Options include:

  • Smoothing
  • Show mesh or show surface
  • Different color palettes
  • Adjustable plot scale

Navigation controls in the lower right of the 3-D display include preset views (top button: overhead X-Y plane; next two buttons: X-Z and Y-Z planes) plus freehand rotation. Left-click and drag to rotate the view; right-click and drag vertically to zoom in and out.

[diagram/image in source — not reproduced here: Figure 2 (Merge View) and Figure 3 (3-D SanjVIEW)]

(D) Region of Interest (ROI) Mean and Statistics (Thermal View)

The region select tools use a predefined shape or a freehand region to determine the mean temperature of a selected area. The thermal image's temperature information is calculated using the masked or single-value thermoreflectance coefficients. Several ROI geometries are available. When you select an ROI or cross-section in the thermal image, the corresponding area is also shown on the CCD image. Selected ROI regions can be saved and reloaded for comparison or reuse.

  • Limit Bars. A newly loaded thermal image is auto-scaled to its highest and lowest data points, so a few outlying points can obscure the image. Use the mouse to set limit bars that coerce the data to the limits you select. This scaling also affects the Merge and 3-D VIEW plots.
  • Color Palette Selection. Several color palettes are available for the thermal image display; the selected palette also appears in the merged image and is preserved when the thermal image is saved.

(E) Parameters & Graphs

Experimental parameters saved with the thermal image are displayed here; the available parameters differ depending on whether the session was Steady State or transient.

Thermal Cross Section Graph. Select the line tool and draw a line on the thermal image to view cross-sections. The resulting data appears in the cross-section graph. If masking is in use, some data points may not appear because they were masked out as invalid. Right-click the plot to change its legend. A drop-down to the left of the plot changes the x-axis from pixel count to distance — select the correct objective used during imaging to get accurate distance scaling. Cross-sections can be saved in ASCII (tab-delimited) format using the Save button and path.

ROI Series Graph. Plots the temperature of a selected region as an image series is processed, so you can track how different regions' temperatures change as voltage or delay changes. Use the drop-down selector to set the x-axis to time or voltage, and choose a linear or logarithmic x-axis. "Chart scale fact[or]" scales the chart by the chosen number.

(F) Image Calibration

Point-to-point calibration calibrates each individual point in the thermoreflectance image against a calibration image file (Figure 4). Image shift matters a great deal here — even a slight shift can cause significantly different Cth values, especially at the edges. Point-to-point calibration requires a Cth map saved from the TCAT software; the Cth(x,y) map assigns a thermoreflectance coefficient to each (x,y) point in the thermoreflectance image.

Point-to-point options:

  • TCAT file — where you load the point-to-point calibration image.
  • Cth Noise Filter — the Cth map can be noisy, which can amplify noise in the final thermoreflectance image (a small coefficient, e.g. ~1E-6, can produce a large amount of noise since the thermoreflectance coefficient is in the denominator).
  • Image Alignment — the software aligns the point-to-point calibration image with the captured thermal image. Auto Align does this automatically; you can also align manually by shifting the Cth X/Y.

Coefficient masking / single coefficient. Masking lets you apply multiple thermoreflectance coefficients and eliminate shadows and other areas without valid thermal data. It takes some practice to get right. If you don't need masking, select Single Coefficient to apply one thermoreflectance coefficient to the entire thermal image.

The mask display shows the current mask alongside the CCD image used to build it. Sliders A1/A2 and B1/B2 set the limits of the coefficient applied to each of two surfaces; pink and green regions show where each coefficient applies. Regions outside the limit bars are treated as NaN (not a number) and excluded from cross-sections and ROI calculations.

To make a mask:

  1. Note that masking for two coefficients requires knowing both thermoreflectance coefficients in advance.
  2. Select Make Mask from the selector.
  3. Slide A1 from 0 (darkest points) to the right until shadows appear in the mask display.
  4. Slide A2 so the A1–A2 band represents the dark surface that should receive the first thermoreflectance coefficient.
  5. Slide B1 and B2 to apply the bright thermoreflectance coefficient to the bright material (B2 is often left at 100%).
  6. Confirm the resulting mask in the display window — its scale reflects the coefficients now applied to the thermal image.

The mask display also has an area zoom tool (selectable below the graph display) to help fine-tune masking in the region of interest.

[diagram/image in source — not reproduced here: Figure 4, Point to Point calibration screen]

(G) Save Images/Data

Use this control to save images, cross-sections, ROI data, and the area-mean graph created by auto-processing (Figure 5). Output is ASCII text compatible with common plotting tools. Check append data file to save a series of cross-sections to a single data file — also useful when comparing different ROI areas within a thermal image.

Image output options. The CCD image, thermal image, merged image, and 3-D VIEW can each be saved in a variety of formats via the Save Path control. When you save an image, it is converted to an RGB image for compatibility with standard image formats — keep in mind that these RGB outputs are for visualization only and do not preserve temperature data. If you need the thermal image data for use in another program (for example, custom calculations in MATLAB), export in ASCII format instead; ASCII files can be reloaded the same way and preserve the numeric thermal information. CCD data is exported in unit-less gray levels, which also depends on averaging time.

[diagram/image in source — not reproduced here: Figure 5, Save Image/Data panel]

(H) Data Series

An image series must already be acquired and saved via the acquisition module before you can process it here. With this control you can:

  • Save an image series as a collection of images (in any output format)
  • Generate a time-vs-ROI-temperature plot
  • Generate a voltage-vs-ROI-temperature plot
  • Save a series of images as an AVI movie
  • Save a series of cross-sections as ASCII (tab-delimited) output

Clicking Process Series sequentially loads each image in the series, starting from the currently selected image, and generates a movie or a graph. You must specify the number of images in the series. Use the Save Path control (see section G) for the resulting data plots, image series, or AVI movie.

The region mean series graph plots the selected area mean (see section E) for each image in the series. Its x-axis can automatically plot voltage or delay, selectable via a drop-down, with a linear or logarithmic scale option. A separate speed setting (slow / medium / fast) controls the playback visualization during auto-processing.

(I) Data

The advanced post-processing features in this section are unique to SanjANALYZER-PLUS and are accessed from Data on the main toolbar.

Hyperspectral thermoreflectance image. For devices made of several different materials within the region of interest, hyperspectral or multi-spectral thermoreflectance calibration gives a fuller picture of device thermal behavior — a common need with modern, materially complex device designs. Hyperspectral calibration improves thermal resolution and accuracy for these devices (Figure 6).

Report generation. Generate and save a report in Word, Excel, or a standard format by selecting the format type. You can choose which data to include — title, notes, CCD image, thermal image, Merge view, and graphs — and preview the report before generating it.

Transient CAL. Use Transient CAL to calculate the average temperature difference between a Gate and a Source region:

  1. Use the Coefficient Mask to carefully select the Gate or Field Plate (R1) region on top of the channel — these pixels should give the most accurate, reflective region for the thermal measurement.
  2. Hold Ctrl and left-click to select R2 on a nearby calibration region — typically the Source or Drain pad.
  3. Enter the Low Range and High Range values for the stable heating region after the device pulse, where temperature is assumed uniform and continuous over short distances.
  4. Enter the Average R1-R2 starting from (s) time — the point (close to the end of the device heating pulse) from which the temperature change between R1 and R2 is averaged. Click Process Data to show the scaled R1 data relative to R2.
  5. Confirm that this matches the decay curves for R1 and R2; the tool outputs the Scaled Cth and the R1/R2 ratio.
  6. The resulting Avg R1-R2 value is typically the temperature difference between the Gate and Source region.
  7. To get the final peak junction temperature, add the Avg R1-R2 temperature change to the Steady State R2 temperature obtained from the Steady State measurement.

Image processing. Additional processing controls are available under Data (Figure 7):

  • Pre-Filter / Averaging Factor (binning). When a thermal image is faint — due to small temperature changes or a poor thermoreflectance coefficient — software binning can help. Binning spatially averages neighboring pixels, lowering spatial resolution but improving the signal-to-noise ratio of a cross-section plot. Binning options are 1x, 2x, 4x, or 8x.
  • Pre-Filter / Interpolation. Raises interpolation from zero-order to bi-linear to smooth the image; higher interpolation adds pixels to smooth out hard edges.
  • Filters (Thermal View). Each filter trades spatial resolution for a lower noise level:
  • Low Pass Filter (smooths the image)
  • Median Filter (filters out outlier noise)
  • Shadow Filter (filters out dark spots)
  • Edge Filter (filters out edges)
  • Convolution Filter
  • Edge Method. Choose the algorithm the software uses to eliminate edges.

[diagram/image in source — not reproduced here: Figure 6 (Hyperspectral Analysis) and Figure 7 (Image Processing panel)]