← AXS-005

Circuit review & bench-test guide

AXS-005 — Rotary encoder PMOD module

Document purpose#

This document explains the axs-005-encoder prototype at component level and turns the design evidence into a practical manual-review and bench-test plan. It is based on the implemented schematic generator (generate_design.py), the module README, the sensor-line conventions in ../README.md, and the verified Pmod-spec notes in ../../pmod-common/PMOD-SPEC-NOTES.md.

The board is a schematic-complete prototype only: ERC is clean (0 errors / 0 warnings) and the netlist has been reviewed, but no PCB has been laid out, fabricated, or bench-verified. Every number below is a design target to verify on a first article, not proof that an assembled board works.

1. What the board does#

A 12-pin PMOD Type 1 (GPIO) module carrying an Alps Alpine EC11E15244G1 rotary encoder (30 detents, 15 pulses per revolution) with an integral shaft push switch. Phase A, phase B, and the switch idle high through 10 kΩ pull-ups to the PMOD 3.3 V rail and are pulled to GND as the contacts close. Light 10 nF snubbers take the worst RF edge off the A/B contacts; the real debouncing and quadrature decoding are deliberately left to the FPGA fabric — that is the lesson the module teaches (level L2).

There is no IC on the board. The module draws essentially zero quiescent current; the only current paths are the three pull-ups when contacts close (about 0.33 mA each).

Functional block diagram#

PMOD J1 (Type 1 GPIO)                         SW1  Alps EC11E15244G1
  pin 6/12  3V3 ──┬── R1 10k ── ENC_A_RAW ──── A
                  ├── R2 10k ── ENC_B_RAW ──── B      C (common) ── GND
                  └── R3 10k ── ENC_SW_RAW ─── S1     S2 ────────── GND
                                    │  │  │
       C1 10nF ── ENC_A_RAW ── GND  │  │  │   (snubbers on A/B only)
       C2 10nF ── ENC_B_RAW ── GND  │  │  │
                                    │  │  │
  pin 1  ENC_A  ◄── R4 100R ── ENC_A_RAW
  pin 2  ENC_B  ◄── R5 100R ── ENC_B_RAW
  pin 3  ENC_SW ◄── R6 100R ── ENC_SW_RAW
  pin 5/11 GND

2. Safety and scope boundaries#

• This is a 3.3 V logic module. Power it only from a PMOD host socket (AruviX ECP5 rev-A or any 3.3 V Digilent-compatible socket) or from a bench 3.3 V supply with the current limit set low (50 mA is ample).
• Do not plug the module into a host socket that switches its power pins to 5 V. The circuit itself would survive (passives and contacts only), but the host-side assumptions in this guide are 3.3 V LVCMOS.
• The plug is polarized only by convention: pin 1 is the rightmost pin of the upper row viewed from the module top with the connector toward you (Pmod spec 1.2.0 Figure 2, verified in PMOD-SPEC-NOTES.md). A reversed or offset insertion can put 3.3 V on signal pins. Check orientation before first power; the 100 Ω series resistors limit but do not eliminate miswiring current.
• The encoder contacts are rated 10 mA / 5 V DC max and the switch 0.1 A / 5 V DC max. Never wire these contacts to switch anything directly; they are signal contacts only.
• Mechanical: the EC11 shaft takes real hand torque. Until the PCB layout and its mounting are proven, support the board when turning the knob so the plug and solder joints are not the mechanical ground path.
• A working bench result is not a lifetime claim. The encoder is rated 15,000 rotational cycles and the switch 20,000 operations; contact behavior (bounce, chatter, resistance) degrades toward end of life.

3. Signal behavior and design intent#

1. At rest, R1–R3 hold ENC_A_RAW, ENC_B_RAW, and ENC_SW_RAW at 3.3 V. The host sees logic high on PMOD pins 1, 2, 3 through R4–R6.
2. Rotating the shaft closes/opens the A and B contacts against the grounded common C in quadrature: 15 complete A/B cycles per 360° revolution, with 30 detents — one detent per half cycle.
3. Pressing the shaft closes S1 to the grounded S2 and pulls ENC_SW low.
4. Falling edges are fast (contact closure to GND). Rising edges are RC: 10 kΩ × 10 nF ≈ 100 µs time constant on A/B. This is deliberately about 20–30x shorter than the specified contact chatter (3 ms max), so it removes RF-fast edges without hiding the bounce the HDL lesson must handle. ENC_SW has no snubber: the push switch bounces hard on purpose.
5. Quadrature arithmetic for this exact part: 15 pulses/rev × 4 edges/cycle = 60 quadrature edges per revolution, and 60 / 30 detents = 2 counts per detent in a 4x decoder.

Which phase leads for clockwise rotation is defined by the Alps product drawing. The fetched Alps product page confirms two-phase A/B output but presents the direction diagram graphically, so the CW/CCW lead assignment below is a bench-verification item, not a cited fact: the common EC11 convention is A leads B for clockwise (viewed from the shaft). Capture it and write it down for the HDL driver.

4. External interfaces#

There are no test points on the schematic; probe the resistor pads. The "RAW" side of R4/R5/R6 (encoder side) shows the contact truth; the PMOD side shows what the host sees.

5. Component-by-component review#

5.1 Connector and encoder#

Key datasheet parameters for SW1 (from the fetched Alps product page): 30 detents, 15 pulses/rev; detent torque 10 ± 7 mN·m; rotational chattering (bounce during rotation) 3 ms max and switch bounce 2 ms max, specified at R = 5 kΩ; encoder contact rating 10 mA / 5 V DC max with a 1 mA minimum operating current; push-switch rating 0.1 A / 5 V DC with a 500 µA / 5 V minimum; insulation resistance 100 MΩ min at 250 V DC; rotational life 15,000 cycles, switch life 20,000 operations; operating range −40 to +85 °C.

Review finding (accepted, watch on bench): the 10 kΩ pull-ups give only 3.3 V / 10 kΩ ≈ 0.33 mA contact wetting current — below the encoder's 1 mA and the switch's 500 µA stated minimums. This follows the sensor-line standard pull-up value and is common practice for EC11-class parts at logic level, but it is formally out of the minimum-rating box: expect the first symptoms of contact aging (missed or dirty transitions) earlier than the rated life, and treat any flaky-contact bench result as possibly this, not only as dirt.

5.2 Pull-ups, series protection, snubbers#

There is deliberately no decoupling capacitor: the module carries no IC, and the snubbers already provide local charge for contact events.

6. Datasheet summary and design interpretation#

Official reference: EC11E15244G1 product page, Alps Alpine (fetched for this review). The page states output waveform details "vary according to the number of detents and other factors" — pull the full product drawing PDF before release and confirm the phase diagram against the bench capture. Check the exact orderable suffix before procurement.

7. Expected values before bench testing#

8. Manual schematic and assembly review checklist#

• Confirm J1 pin-1 orientation against Pmod spec Figure 2 (see PMOD-SPEC-NOTES.md): odd pins on one row, pin 1 at the correct corner, and 3V3/GND on 6/12 and 5/11 respectively. Buzz plug pins to nets.
• Verify SW1 pinout against the Alps drawing: A/C/B on the encoder side (C is the center/common), S1/S2 on the switch side. Confirm C and S2 both buzz to GND.
• Measure R1–R3 = 10 kΩ and R4–R6 = 100 Ω in circuit (expect near-nominal; the network topology does not parallel them significantly).
• Confirm C1/C2 sit on the RAW (encoder) side of R4/R5, and that ENC_SW has no capacitor.
• Inspect SW1 mechanical mounting legs for full solder fill — they are the torque path.
• Confirm no solder bridge between adjacent encoder pins (A–C–B pitch is tight); a C-to-A bridge reads as a permanently low phase.

9. Ordered bench-test procedure#

Stop at the first abnormal result. Record board serial, host or supply identity, equipment IDs, ambient, and operator with every capture.

A. Unpowered inspection and resistance tests#

1. Complete the checklist above under magnification.
2. Measure 3V3-to-GND resistance at the plug: expect > 100 kΩ (only the pull-up paths through open contacts; essentially open).
3. With an ohmmeter from ENC_A (pin 1) to GND, slowly rotate: expect alternating open / ~100 Ω (R4 plus contact) as the A contact closes. Repeat for B (pin 2) and for the switch on pin 3 while pressing.

B. First power and static levels#

1. Power from a current-limited bench 3.3 V supply (50 mA limit) via the plug's 3V3/GND pins, or plug into the PMOD host socket unpowered and then power the host.
2. Confirm supply current ~0 mA idle, and ~0.33 mA per held-closed contact.
3. DMM: pins 1/2/3 read 3.3 V idle; hold the shaft mid-detent / press the switch and confirm ~0 V on the corresponding pin.

C. Quadrature capture (scope / logic analyzer)#

1. Probe ENC_A and ENC_B (PMOD pins 1 and 2), GND to pin 5. Scope at ≥1 MS/s or logic analyzer at ≥1 MHz sample rate; bounce hunting later needs single-shot capture.
2. Rotate slowly clockwise through several detents. Capture and verify:

• A and B are clean 3.3 V / 0 V levels with the ~100 µs RC rising edge.
• The two channels are in quadrature: each edge of one channel falls near the middle of the other channel's half-cycle (90° nominal).
• Record which channel leads for CW. Repeat counter-clockwise and confirm the lead swaps. This closes the direction-assignment verification item from section 3.

1. Rotate through exactly one full revolution (index a knob flat or mark the shaft) and count edges on A+B combined: expect 60. Count full A cycles: expect 15. Count detents by feel: expect 30.
2. Spin fast (a few revs/second) and confirm no missed half-cycles and that the RC edges still complete (quarter-period at 3 rev/s ≈ 5.5 ms >> 100 µs).

D. Detent counting / 4x decode test#

1. Feed A/B to the FPGA 4x quadrature decoder (or count edges offline from a logic-analyzer trace with 4x decode arithmetic).
2. Rotate exactly N = 10 detents CW: expect the 4x count to advance by 2 x N = 20. Rotate 10 detents CCW: expect it to return to the start.
3. Repeat 5 times each direction; any drift means missed edges (look for chatter faster than the decoder's synchronizer/debounce) or double counts (decoder bug). Distinguish with the raw capture from step C.
4. Verify the rest position: at a detent, A and B should both be in their idle state (typically both high for EC11 15-pulse/30-detent variants — record what this unit does; it constrains the decoder's reset state).

E. Switch verification#

1. Scope ENC_SW (pin 3) single-shot, falling-edge trigger. Press the switch ~10 times; capture and measure the worst bounce burst on press and on release. Expect ≤ 2 ms per the Alps spec (specified at 5 kΩ; our 10 kΩ pull-up may lengthen apparent bounce — record actual).
2. Confirm clean 3.3 V idle via R3 and ~0 V held-down level, and that the rising edge is fast (no snubber on this net).
3. Confirm rotation does not glitch ENC_SW and pressing does not glitch A/B (common-mode mechanical coupling check).
4. Feed the measured worst-case bounce into the HDL debouncer parameters: the debounce window must exceed measured bounce with margin, while staying far below the minimum expected press duration.

F. Extended / release-only tests#

Rotational endurance toward the 15k-cycle rating, temperature-range behavior, connector retention force, and knob torque abuse are release-scope tests, not first-bench tests. The 0.33 mA contact-current deviation (5.1) should be revisited if endurance testing shows early contact degradation.

10. Troubleshooting map#

11. Bench record template#

12. Review conclusion#

The circuit is minimal and coherent: three pulled-up contact signals with series protection, and snubbers scaled (100 µs) to clean RF edges without stealing the debounce lesson from the HDL. The quadrature arithmetic for this exact part (15 pulses, 30 detents, 60 edges/rev, 2 counts/detent at 4x) is fixed by the fetched Alps data and is the core of the bench pass/fail criteria. Open items before release: the CW/CCW lead assignment must be captured and documented (the product page does not state it in text); the 0.33 mA contact current is below Alps' stated 1 mA / 500 µA minimums and is an accepted, documented deviation to watch in endurance testing; and PCB layout, DRC, fab outputs, and the HDL deliverable do not exist yet — nothing in this guide substitutes for first-article measurement.