Sensor and coordinates

Coordinates, scan ranges and origin finding in the GUI all work in the "logical coordinates of the sensor". To use the GUI correctly, first understand how the coordinates are defined on this page.

There are two kinds of sensors, strip type and pixel type, selected with [Type] in the sensor editor (since the 2026-10-16 version). The type changes the definition of the logical origin, the Find Origin and Full Calibration procedures, the default scan range, the stage's [Go to strip] / [Go to pixel], and the analysis type. Unless otherwise noted, the following describes the standard sensor (strip type). For the pixel type, see Pixel sensors.

Geometry of the standard sensor

The standard sensor is an HPK strip AC-LGAD with 500 µm pitch and about 10 mm length. X runs along the strips, Y across the strips, and Z along the optical axis (focus).

Along X, starting from the edge of the sensor frame (GND) on the readout side, the regions are:

Region	Width	Logical X [µm]
⓪ Between frame and strip metal end	52 µm	0 → −52
① Stub	380 µm	−52 → −432
② Bonding pad region	200 µm	−432 → −632
③ Exposed strip	9318 µm	−632 → −9950
④ End gap	98 µm	−9950 → −10048
Beyond	—	Next strip region (not read out)

• X = −10 mm is the middle of the end gap ④ (the boundary between two strip regions).
• Along Y, six strips C0–C5 are arranged at 500 µm pitch, and C1–C4 are read out (Readout channel map).

Logical coordinates and logical origin

Positions shown on screen, Go-to moves and scan ranges are all in logical coordinates (in µm).

Axis	Definition of the logical origin (0)	Direction
X	Edge of the sensor frame (GND)	Strips extend to negative X
Y	Electrode center of the reference CH (C1 by default)	More negative for higher CH numbers (C2 = −500 µm …)
Z	In-focus position (focus)	—

The logical origin is determined automatically by origin finding, and the controller coordinates are reset to 0 at that position (R:W). All stage moves are relative moves (M:W) to logical coordinates (absolute moves A:W are not used because they were buggy).

Machine coordinates

Machine coordinates are measured from the mechanical origin (H:W) and are shown below the logical coordinates on the stage panel.

• R:W resets the controller coordinates to 0, so the GUI accumulates "the machine coordinates of the logical origin" and saves them in logs/stage_frame.json (reset to 0 on H:W, and the previous coordinates are added on each R:W). This value is kept across GUI restarts.
• Tracking is lost if you do M-ORG from the joystick JS-301, restart the controller, or interrupt a homing; the panel then shows "Machine coordinates: unknown". Run [Mechanical Home (H:W)] from the GUI again.
• Machine coordinates can change with each M-ORG, so use them only as a guide for coarse positioning.

Sensor tilt and correction coefficients

The sensor is mounted slightly tilted, so three coefficients are used to correct for it.

Coefficient	Meaning	Default for the standard sensor
dZ/dX	Tilt of the focal plane along X	−0.02 µm/mm (nearly flat)
dZ/dY	Tilt of the focal plane along Y	+3.79 µm/mm
dY/dX	In-plane rotation of the strips	+7.67 µm/mm

Without the rotation correction, Y shifts by about 72 µm over the full strip length (about 9.4 mm).

• The screen display and the [Edit coefficients...] input use µm/mm; the keys in sensors.json and the scan log (dzdx, dzdy, dydx) use µm/µm.
• The coefficients are taken in this order of priority:
  1. Values saved in the sensor profile (Full Calibration or manual input)
  2. For the standard sensor only, the measured values in scan_geometry.py (camera measurement, 2026-08-24)
  3. 0 for any other uncalibrated sensor
• The active coefficients and their source are shown both on the [Alignment] tab of the stage panel and in the [Tilt correction] box of the auto scan panel.

Sensor dictionary (sensor profiles)

So that boards or sensors can be swapped without editing code, the information for each sensor is collected in one profile and saved in logs/sensors.json. Switching with [Sensor] in the toolbar replaces the default scan range, tilt coefficients, saved positions and the scope CH to readout pad map with those of that sensor.

Item	Content
Name, note	Name shown in the list, and a free-text note
Type	Strip / Pixel
Geometry (strip)	Strip pitch, electrode (metal) width, offset from the frame edge to the start of the exposed strip, exposed strip length, end gap
Geometry (pixel)	Pixel pitch, bonding-pad size (the signal-free square; the ruler for Find Origin)
Readout	For each scope CH, used or not and the pad name; reference CH. Strip: electrode center Y (reference CH = 0). Pixel: pixel (column, row)
Default scan range	Start, stop and step of the fast axis Y and the slow axis X
Full calibration points	Strip: X points, Y points (in the form "Y:CH"), Y offsets for Z measurement and X-edge detection, half-width of the Y sweep. Pixel: pad centers of the readout pixels other than the reference (automatic), Y offset for Z measurement, half-width of the Y sweep
Tilt coefficients	dZ/dX, dZ/dY, dY/dX
Saved origin	Machine coordinates of the logical origin. Updated automatically on each origin finding
Saved positions	Name and coordinates (logical or machine coordinates)

As of 2026-11-01, sensors.json on the control PC contains the following 7 profiles.

Sensor (ID)	Type	Status / use
Standard strip (500 µm pitch, 10 mm) (strip500_std)	Strip	Contains measured values (scan_geometry.py). Normally use this one
4col (1000 µm pitch) (4col_1000)	Strip	Template with unconfirmed dimensions except the pitch. Correct them to measured values before use
Variable pitch (500 µm section) (varpitch_500)	Strip	Same as above
Standard strip 2x2 measurement (500 µm pitch, 10 mm) (2x2_500_m_10_mm)	Strip	Profile for the scan across 2 columns (the segment boundary) in the central region of the 2×2 segmented sensor (SEG scan scan_20261008_151544_2dYX in the analysis note matome, 5 µm rows)
pixel#3 2x2 measurement (pixel_3_2x2)	Pixel	Measurement of the HPK S16694 pixel sensor read out as 2×2 (2026-10-15). Created with an older version and migrated to the pixel type automatically when the 2026-10-16 version started
pixel#4 2x2 measurement (pixel_4_2x2)	Pixel	Measurement of pixel#4 of the same type read out as 2×2
pixel#4 1x4 measurement (pixel_4_1x4)	Pixel	Measurement of 4 pixels of pixel#4 read out in one row (C1→C4 one pitch apart in the +X direction)

The GUI's initial profiles (when it is first started on a new PC) are the first three above and the pixel-type template "Pixel 2x2 (500 µm pitch, 100 µm pad)" (pixel_2x2_500, the layout of pixel#3 on 2026-10-15).

Pixel sensors

For the pixel type, the bonding pad (a metal-covered square that gives no signal) is used as the ruler instead of the "frame edge and electrode center" of the strip type.

Quantity	Definition
Logical origin	The +X/+Y corner of the reference CH's pad. The pad spans x, y ∈ [−pad size, 0]
Pixel center	Same as the pad center. Center of the pixel at (column, row) = (−pad/2 − column × pitch, −pad/2 − row × pitch)
Column, row	The reference CH is (0, 0). Columns increase by 1 toward −X, rows toward −Y
4-pixel corner	(−pad/2 − pitch/2, −pad/2 − pitch/2)
Default scan range	600 × 600 µm from the pad corner, in 10 µm steps

In the standard 2×2 layout (pitch 500 µm, pad 100 µm), the C1 pad center is (−50, −50), C2 is the −X neighbor, C3 the −Y neighbor, C4 the diagonal neighbor, and the 4-pixel corner is at (−300, −300) µm.

• The Find Origin and Full Calibration procedures also switch to the pixel versions (Find Origin for pixel sensors).
• The move button in the [Positions] tab of the stage panel becomes [Go to pixel], which can move to "Beside pad (signal)", "Pad center (no signal)" and "4-pixel corner".
• The adjacent 2ch pairs and midpoint positions for TUNE optimization are also determined from this geometry.
• sensor_type=pixel is written in the scan log header, and the [Analysis] tab automatically selects the pixel analysis (Analysis).
• The analysis (tt_pixel.py) detects the pad positions automatically from the amplitude, so a small offset of the origin does not change the analysis results.

Registering a new DUT

Strip typePixel type

The standard strip opened in the editor

[Type] switched to Pixel (switched on the standard strip profile for illustration; columns and rows are still 0)

The bottom of the dialog shows the sensor's tilt coefficients and their source, the saved origin (machine coordinates), and the number of saved positions. The screenshots were taken in simulation, so the numbers differ from the real setup.

1. Select a similar sensor with [Sensor] in the toolbar.
2. Press [Edit...].
3. Update [Name] and [Note], and select the [Type].
4. Update the geometry.
   • Strip: pitch, electrode width, frame edge → strip start offset, exposed length, end gap
   • Pixel: pixel pitch, pad size
5. Update the readout (use flag and pad name for each CH, reference CH).
   • Strip: electrode center Y. For equal spacing, use [Fill Y from pitch].
   • Pixel: (column, row) of each CH. For 2×2, [Standard 2x2 layout] fills in C1 (0,0), C2 (1,0), C3 (0,1), C4 (1,1).
6. Press [Derive from geometry] to regenerate the scan range and calibration points.
7. Press [Save As New Sensor]. To just fix an existing sensor, press [Save].

Note

A sensor saved under a new name does not inherit tilt coefficients, saved origin or saved positions. Run a Full Calibration first on a new sensor.

Matching the X origin to the real dimensions

Origin finding takes "the X edge where the signal disappears (boundary between pad region and exposed strip) + the design offset of 632 µm" as the frame edge. If the real dimensions differ from the design, adjust "Frame edge → exposed start" in the sensor editor.