← AXS-024

Circuit review & bench-test guide

AXS-024 — TFmini-S lidar carrier PMOD module

Document purpose#

This document explains the axs-024-lidar carrier at component level and turns the design evidence into a practical manual-review and bench-test plan. It is based on generate_design.py, the generated schematic, the README, and the manufacturer documentation for the carrier parts and the intended plug-in module (Benewake TFmini-S).

The board is an unfabricated prototype: schematic generated, ERC clean (0/0), netlist reviewed, no PCB layout, no fabricated board, no BOM export, and no bench results exist. Two connector MPNs are still TBD: in the schematic fields. Everything below is design intent to be verified, not proof that an assembled board works or is safe.

1. What the board does#

The carrier adapts a Benewake TFmini-S single-point ToF lidar to a Digilent PMOD Type 3 (UART) socket. The TFmini-S is a 5 V-powered sensor with 3.3 V LVTTL UART logic, so the carrier splits the domains: external 5 V enters on screw terminal J3 and goes only to the J2 supply pin; all signals and the PMOD stay strictly 3.3 V. The sensor streams 9-byte binary frames (header 0x59 0x59, distance, signal strength, chip temperature, checksum) at 115200 baud, 100 Hz by default.

Functional block diagram#

EXT 5 V bench supply ──> J3 (screw terminal) ──┬──> J2 pin "+5V" ──> TFmini-S VCC
                                               ├── C1 100 nF
                                               └── C2 10 µF
PMOD host (ECP5) J1                                    TFmini-S (plug-in, JST GH pigtail)
  pin 2 TXD  ──────────────────────────────────> J2 "RXD" (lidar receive)
  pin 3 RXD  <───────── R1 100 Ω ──────────────  J2 "TXD" (lidar transmit, LIDAR_TX)
  pins 5/11 GND ─────────── common ground ────── J2 GND ── J3 GND
  pins 6/12 3V3 ── (decoupling reference only; does NOT power the lidar)

2. Safety and scope boundaries#

• The TFmini-S manual states the product has no overvoltage and no polarity protection; supply must be 5 V ± 0.1 V. A reversed or overvolted J3 can destroy the sensor immediately. Verify J3 polarity with a DMM before every first connection of a new supply.
• Never feed 5 V into any PMOD pin. The only 5 V nets are J3 pin 1 and the J2 supply pin. On the unrouted PCB this separation must be re-verified after layout.
• The emitter is a near-infrared source behind an IR-pass filter. Do not stare into the transmit lens at close range, and do not remove the shell or filter.
• Use a current-limited bench supply (set ≈300 mA limit) for first power-up.
• This guide covers bench verification only; nothing here is EMC, laser-safety classification, or environmental qualification evidence.

3. Supply sequencing, common ground, and no-host behavior#

This is an external-5 V carrier; the order in which rails appear matters.

1. Ground first, always common. J3 pin 2, J2 pin 4, and PMOD pins 5/11 are one ground net. Before any power, buzz continuity from the bench-supply return through J3 to the PMOD ground pins. Never feed J3 from a supply whose return is not tied to this net (no "floating 5 V" with a separate earth path through the scope).
2. Recommended order: PMOD host on first, then external 5 V. Power the FPGA host (3V3 present, host TXD idle high or tri-stated), then enable the J3 supply. Power down in reverse: 5 V off first, then the host.
3. What happens if 5 V is applied with no PMOD host attached: the lidar runs normally and drives 3.3 V UART frames into R1 toward an open pin — harmless. The riskier case is 5 V applied while the host board is attached but unpowered: LIDAR_TX then drives ≈3.3 V through R1 into the unpowered FPGA input, forward-biasing its protection diode and back-powering the dead 3V3 rail. R1 (100 Ω) limits this to roughly (3.3 − 0.3 − V_rail)/100 Ω ≤ ~30 mA worst case, which exceeds typical per-pin injection ratings. Treat "5 V on, host attached but unpowered" as a prohibited state.
4. There is no reverse-blocking or inrush element on EXT_5V; C2 charges directly at connection. With only 10 µF this is benign, but do not add large external capacitance at J2 without rechecking.

4. Plug-in module verification (read before first power)#

This section is mandatory. TFmini-S pigtails and clone cables vary, and a swapped supply/signal wire can destroy the sensor or the FPGA.

4.1 The pin-order question — a real, unresolved discrepancy#

The carrier schematic and README assume J2 order 1 +5V, 2 RXD (into lidar), 3 TXD (out of lidar), 4 GND. The Benewake product manual (SJ-PM-TFmini-S A01, Table 5) numbers the module connector ① GND (black), ② +5V (red), ③ RXD/SDA (white), ④ TXD/SCL (green) on a GH1.25-4P connector with a Molex 51021-0400 terminal on the far end. That manual ordering matches the carrier's assumed order in neither direction (forward: GND/5V/RXD/TXD; reversed: TXD/RXD/5V/GND). The numbering convention of the physical JST GH housing versus the manual's ①–④ may reconcile this, or the schematic may simply be wrong — this must be resolved against a physical unit before the PCB is laid out, and re-verified on every assembled cable before first power.

Verification procedure, with the sensor not connected to the carrier:

1. Identify the wire colors at the sensor end: black = GND, red = +5 V, white = RXD (into lidar), green = TXD (out of lidar), per the manual.
2. Follow each wire through the pigtail (and any extension) to the plug that mates with carrier J2. Extensions can swap pairs; buzz end-to-end.
3. With the carrier unpowered, buzz J2 pad 1 to J3 pin 1 (+5 V net) and J2 pad 4 to J3 pin 2 (GND net). Confirm the red wire will land on the +5 V pad and the black wire on the GND pad in the actual mated orientation.
4. Confirm the white wire lands on the net driven by PMOD pin 2 (host TXD) and the green wire on the R1 side (LIDAR_TX).
5. Only then connect the sensor.

4.2 Plug-in key specifications (Benewake TFmini-S, manual A01)#

5. Component-by-component review#

Every physical reference designator on the schematic is covered. #FLG01– #FLG03 are ERC power flags only; they place no physical part.

Note: the host-driven TXD line has no series resistor — it is host output to sensor input, consistent with the other carriers in this line. The sensor-driven direction has R1.

6. Datasheet summary and design interpretation#

Official references: Benewake TFmini-S product page, TFmini-S Product Manual SJ-PM-TFmini-S A01 (SparkFun mirror), TFmini-S datasheet SJ-PM-TFmini-S-01 A01 (Mouser mirror), JST GH connector family. Check the current revision before procurement.

7. Expected values before bench testing#

8. Manual schematic and assembly review checklist#

• Resolve the J2 pin-order discrepancy (section 4.1) against a physical TFmini-S before layout; record the resolution in the repository.
• Pin the two TBD: MPNs (J1 header, J3 Phoenix terminal) before BOM release.
• After layout: confirm the 5 V net exists only between J3, C1, C2, and the J2 supply pin; confirm a 5 V keep-out from the PMOD edge; check J3 polarity marking on silkscreen.
• Confirm J2 pin-1 marking on silkscreen matches the GH housing polarization.
• Verify R1 is in the lidar-TX line, not the host-TX line.
• Inspect solder on the GH connector's small pads; verify the locking latch.

9. Ordered bench-test procedure#

Stop at the first abnormal result. Record board serial, sensor serial, supply, equipment, and operator.

A. Unpowered checks#

1. Complete section 4.1 (pigtail buzz-out). Do not skip it.
2. Measure resistance EXT_5V-to-GND (expect high, capacitive charging) and 3V3-to-GND at the PMOD plug. Investigate any hard short.
3. Verify J3 GND to PMOD GND continuity (<1 Ω).

B. Sensor alone, no host#

1. Leave the PMOD unplugged. Set the bench supply to 5.00 V, 300 mA limit.
2. Power J3. Expect ≤140 mA average; a red glow is visible in the transmit lens viewed from the front (per the Benewake manual's no-data checklist).
3. Scope LIDAR_TX (sensor side of R1): 3.3 V UART bursts every ~10 ms.
4. Power off.

C. Host attach and first frames#

1. Power the FPGA host first, then J3 (section 3 order).
2. Capture PMOD pin 3 with a logic analyzer at 115200 8N1: verify 0x59 0x59 sync, decode distance/strength/temp, verify checksum on every frame.
3. Point the sensor at a wall ≈1 m away: distance within ±6 cm (manufacturer accuracy for 0.1–6 m). Move to 3 m and re-check.
4. Provoke abnormal codes: cover the lens (expect Dist 65535 / Strength <100); aim at a mirror or through glass and note the documented malfunction modes.
5. Blind-zone check: a target at <10 cm should be treated as unreliable.

D. Robustness#

1. Sweep the J3 supply 4.9–5.1 V (the allowed ±0.1 V): frames must stay valid.
2. One-hour soak at 100 Hz: log frame-loss and checksum-error counts (target 0), supply current, and connector temperature.
3. Hot-unplug/replug of J3 with the host running; verify the sensor recovers and the host UART resynchronizes on the next 0x59 0x59.

10. Troubleshooting map#

11. Bench record template#

12. Review conclusion#

The architecture is honest and simple: external 5 V for a 5 V sensor, common ground, 3.3 V-only signaling to the PMOD, one series resistor on the only sensor-driven line, and local decoupling sized for the documented 200 mA transmit peaks. The principal risks are (1) the unresolved J2 pin-order assumption versus the Benewake manual's connector numbering — the sensor has no input protection, so this is a destroy-the-part class error; (2) the back-powering path into an attached-but-unpowered host; (3) two TBD: connector MPNs; and (4) the total absence of layout, fabrication, and bench evidence at this revision.