Skip to the content
4 layers52.3 × 93.2 mmPantilt controller, the printed circuit board in 3D

Pantilt controller

AMOVFC/PanTilt

Three projects live in this folder. They share NewLibrary.kicadsym, TMC2209.pretty and the same philosophy: nothing on the board needs precision placement, every active part is a module in a socket.

Size
52.3 × 93.2 mm
Copper layers
4
Thickness
1.6 mm
Parts
84
Nets
97
Pads
321
Tracks, vias
707, 25
Pours
1
Routed
Yes
KiCad
KiCad 10.0 (file 20260206)
Created
17 Aug 2026
Updated
9 Oct 2026

Overview

Pantilt controller is an open-source printed circuit board by AMOVFC, from the AMOVFC/PanTilt repository on GitHub. It is a 4-layer board of 52.3 × 93.2 mm with 84 parts on 97 nets, routed with 707 tracks and 25 vias.

It is built around the 2.54-1*8P母 5.0H, CC0805KRX7R9BB104 and KM107M050F12RR0VH2FP0, with the RCA03100RFLF, ZX-PM2.54-1-22PY and KH-2.54FH-1X12P-H8.5 among its other parts. Its parts include 30 connectors, 20 capacitors, 20 resistors, 5 diodes and 4 ICs.

Main components

12

README

legacy/pantiltslide/README.md

Carrier boards — everything plugs in

Three projects live in this folder. They share New_Library.kicad_sym, TMC2209.pretty and the same philosophy: nothing on the board needs precision placement, every active part is a module in a socket.

project5 V / 3.3 VdecouplingI2C pull-upsfor
pantiltslide_fullexternal — regulated 5 V into J37nonenonethe board as you routed it
pantiltslide_full_pwronboard LM2596S-5 buck → AMS1117-3.3the regulators' own8one 24 V brick runs the rig
pantiltslide_full_turnkeyexternal, same as the carrier board17 caps8hand to PCBWay, plug modules in

Both variants are derived copies: same sockets, same connectors, same net names, same footprint positions, same routed tracks. Regenerate either with tools/gen_variants.py.


pantiltslide_full_turnkey

Everything the carrier board leaves to chance, made explicit, with no regulation added. The intent is a board an assembly house can finish and you can populate by plugging modules in.

What was already fine

Nearly everything. Of the 28 parts on the carrier board, 27 are already sockets or connectors an assembly house can place: both 1×22 DevKit sockets, all four TMC2209 stepstick sockets, both 1×12 mux sockets, every JST XH/EH connector, both screw terminals, the fuse clip.

The one gap was J23, the OLED — a male pin header. An SSD1306 module comes with male pins, so it could not plug into it; you would have needed a female-to-female jumper. It is a PinSocket_1x04 here. Same pads in the same places, so the copper already routed to it is untouched.

What stays on a crimped cable, and why

Motors, limit switches, encoders, buttons and the two AS5600 heads keep their JST XH/EH connectors. Those are not modules that plug into a board — they are parts bolted to a moving rig, and the JST choice is deliberate: polarised so a harness cannot go in backwards, latching so it cannot shake loose. Swapping them for 2.54 mm sockets would let you use bare DuPont jumpers, at the cost of both those properties. Crimping the harnesses is the one job this board does not remove. Say the word if you would rather trade that for DuPont.

Decoupling

17 caps. Nothing on this board is an IC, so these are rail caps — but what each one is for is still specific, and every cap carries a Decouples field naming the pin it belongs beside (C3 → U2 VM). The rails here are drawn as global labels, which tells a layout pass nothing about association, so that field is the only thing carrying the intent. It is hidden and on F.Fab — metadata, not silkscreen.

capswhy
+24VC1 470 µF at J38, C2–C5 100 µF one per driver, C6–C9 100 nF one per driverthe stepsticks carry their own VM bulk, but what kills drivers is the inductance of the supply leads between the PSU and the board
VCC_5VC10 100 µF, C11 100 nF at J37the 5 V arrives on wires from an external brick
3.3vC12 22 µF + C13 100 nF at J1, then 100 nF each at J20, J21, J22, J23the DevKit's LDO feeds every peripheral through header pins

Inrush. ~470 µF of bulk on a screw terminal will spark when you connect a live 24 V supply. Normal for a stepper board, worth knowing before it startles you.

I2C pull-ups

R1–R8, same scheme as the pwr variant — see below. Only the designators differ (R1–R8 here, R3–R10 there, where R1/R2 are LED ballast).

Status

carrierturnkey
footprints2853
tracks469469 — none touched
ERC errors40
DRC1 clearancethe same 1, nothing new
schematic parity00
unrouted1 (TILT_M1A)51 — that one plus 25 new parts

This board is not fab-ready yet, and cannot be made so automatically. The outline is unchanged and the 25 new parts are parked outside it, because unlike the power section they do not belong in one tidy block — a decoupling cap that is not beside the pin it serves is decorative. Placing them is a human pass: read each cap's Decouples field, put it against that pin, grow the outline if you need to, re-route. integrated/tools/check_decoupling.py is the pattern for verifying the result (it measures cap-to-owner distance and flags anything over 10 mm).

What it adds

LCSC single-unit prices, checked 2026-08-28:

eachqty
470 µF 50 V$0.121$0.12
100 µF 50 V$0.064$0.24
100 µF 16 V$0.031$0.03
100 nF 0805$0.01510$0.15
22 µF 0805$0.0151$0.02
0805 resistors$0.0038$0.02
total~$0.58

Under a dollar in parts. What it actually costs is 25 placements on the assembly quote and a layout pass.


pantiltslide_full_pwr

The carrier board has no regulation at all. On the bench that means two supplies: 24 V into J38 for the motors and a separate regulated 5 V into J37 for the DevKit. This variant runs the whole rig from one 24 V brick.

text
J38 24V ─ F1 ─┬─ D3 TVS ─┬─ C1 470µF ─┬─ C2 ─┬─ R1/D4 ─┬─ U5 LM2596S-5
              │                                         │
              └──────────────────── U1..U4 VM ──────────┘
                                                         │
                        L1 100µH ── D2 SS34 ──► VCC_5V ──┴─► J1.21 (DevKit 5V)
                        D1 SS34 catch                        │
                        C3 220µF                             └─► U6 AMS1117-3.3 ─► 3.3v
refpartjob
U5LM2596S-5, TO-263-524 V → 5 V, ~1 A available after the drivers
L1, D1, C3100 µH / SS34 / 220 µFthe buck's switching loop — keep it tight
D2SS34ORs the buck into VCC_5V
U6AMS1117-3.3, SOT-2235 V → 3.3 V for the peripheral rail
C1, C2470 µF 50 V, 100 nF24 V bulk, right at U5's VIN
D3SMBJ33Aclamps 24 V spikes; reversed supply crowbars it and blows F1
D4/R1, D5/R2LED + 10 k / 1 k24 V present, 3.3 V present
JP10 Ω, DNPfallback link, ESP_3V3 ↔ 3.3v

D2, and why the buck sits behind a diode

VCC_5V has three possible sources: the buck, J37, and the DevKit's own USB port. D2 means the buck can push the rail up but nothing can push current back into its output, so plugging USB in while 24 V is live is safe — which is exactly what you do when reflashing on the rig. The cost is one Schottky drop: the rail sits at ~4.7 V on buck power, still comfortably above the DevKit LDO's dropout.

The one net that changed: ESP_3V3

On the carrier board the DevKit's 3V3 output pins (J1.1, J1.2) feed every peripheral. Here U6 drives that rail instead, so those two pins had to come off it: two regulators must never share a net. They are now ESP_3V3, and the DevKit's LDO only powers the ESP32. JP1 is the escape hatch — fit the 0 Ω, leave U6 unpopulated, and the board is back to the old topology.

This is the one thing that costs re-routing. J1.1/J1.2 were the source of the whole 3.3v tree, so the two segments that fed it out of those pins are deleted here — otherwise they would short ESP_3V3 to 3.3v. The rest of the tree is untouched and needs a feed from U6. DRC shows it as one dangling track plus the ratsnest.

Status

carrierpwr
outline76 × 113 mm123 × 113 mm (47 mm column added on the right)
footprints2853
tracks469467
ERC errors40
DRC1 clearancethe same 1, plus 1 dangling track
schematic parity00
unrouted153

The 25 new footprints are shelf-packed into the new column with 3 mm gaps: no courtyard overlaps, no shorts, but not hand-optimised. When you rearrange:

  • Keep C1/C2 hard against U5's VIN pin, and U5/L1/D1 in as small a loop as the parts allow — that loop is the board's main EMI source.
  • U5's tab (pin 3) is its heatsink. At 24 V → 5 V it burns ~1 W; give it a copper pour, not a track. LM2596T-5.0 in TO-220-5 is a drop-in if you would rather bolt a heatsink to it — same symbol, different footprint.
  • There are still no copper zones on this board. A ground pour is worth adding before the switching regulator goes in.
  • The eight pull-ups want to end up near what they pull: R7–R10 beside J20, not in the power column where the packer left them.

What it costs

LCSC single-unit prices, checked 2026-08-27:

refparteach
L1SRN8040-101M 100 µH$0.73
U5LM2596S-5.0$0.53
C1470 µF 50 V$0.12
D3SMBJ33A$0.03
C3220 µF 25 V$0.03
D1, D2SS34 ×2$0.06
U6AMS1117-3.3$0.04
rest4 caps, 3 resistors, 2 LEDs, 8 pull-ups~$0.13
total~$1.68

L1 is the one to watch: SRN8040-101M had 253 in stock. Any shielded 100 µH with ≥1.5 A saturation works, and the LM2596 datasheet allows 33–100 µH, so there is room to substitute.


I2C pull-ups (both variants)

The carrier board has no pull-ups anywhere — every I2C device is a plug-in module, so the bus works only if the modules happen to carry their own. They usually do, and "usually" is the problem: a module that doesn't leaves you soldering resistors onto header pins on a finished board. Fitting them costs about 3 cents.

Four buses actually move traffic:

busnettovalue
AGPIO_11 / GPIO_12TCA9548A breakout J20, on-board4.7 k
BGPIO_13 / GPIO_14SSD1306 OLED J234.7 k
mux ch 0MUX_CH0_SDA/SCLAS5600 pan J21, at the joint2.2 k
mux ch 1MUX_CH1_SDA/SCLAS5600 tilt J22, at the joint2.2 k

Mux channels 2–7 only reach the spare header J24, so they get nothing — an unused channel with pull-ups is just idle current.

Why 2.2 k on the mux channels. The AS5600 heads sit at the end of the longest cables on the rig, and the TCA9548A's pass gates put that cable capacitance on its own isolated segment. That is the one place where rise time gets tight, and it is the segment whose pull-up you cannot move closer to the load.

These are deliberately belt-and-braces: in parallel with a module's own 10 k they land near 3.2 k / 1.8 k, still well inside the ~3 mA an I2C device has to sink. If a module turns out to carry a strong pull-up of its own, lift the board resistor rather than the other way round.


ERC

Both variants: 0 errors. The carrier board's 4 power_pin_not_driven errors are gone — PWR_FLAGs (and, on the pwr board, the real regulator outputs) drive the rails now.

Both gain the same 8 warnings: Pins of type Bidirectional and Power output are connected. Those are the TMC2209 MS1/MS2 address straps, typed bidirectional in New_Library:tmc2209 and tied to GND/3.3v to set the UART addresses. They only appear because the rails finally have a real driver. Fixing it means retyping those two pins as inputs in the shared symbol library, which would also touch the carrier board — left alone deliberately.

The remaining warnings (145 off-grid endpoints, 20 isolated labels, 3 duplicate net names) are inherited from pantiltslide_full unchanged.

Regenerating

bash
python pantiltslide/tools/gen_variants.py

Reads pantiltslide_full.*, writes both variants. UUIDs are derived, so re-running produces the same files and an empty diff. Note that pantiltslide_full.kicad_sch is itself generated (see tools/README.md) — if you rebuild it, re-run this too.

The script asserts on everything it depends on: the 3.3v label at J1, the J23 header footprint, and the two tracks it deletes. If the base board moves under it, it stops rather than producing a quietly wrong board.

Before ordering

  • Turnkey: place the 25 new parts against the pins their Decouples field names, grow the outline, re-route.
  • Pwr: route the power section, and add the 3.3v feed from U6.
  • F1 is a 2 A placeholder — size it to the motor supply.
  • On the pwr board, D3 is a unidirectional SMBJ33A. A bidirectional SMBJ33CA will clamp spikes but will not crowbar a reversed supply.
  • Neither board has copper zones yet.

Parts

84
QtyValueFootprintReferences
9100nFCapacitor_SMD:C_0805_2012MetricC11, C13, C14, C15, C16, C17, C18, C19, C20
710kResistor_SMD:R_0805_2012MetricR25, R26, R27, R28, R29, R30, R31
52.2kResistor_SMD:R_0805_2012MetricR5, R6, R7, R8, R32
4100nF/50VCapacitor_SMD:C_0805_2012MetricC6, C7, C8, C9
4100uF/50VCapacitor_THT:CP_Radial_D8.0mm_P3.50mmC2, C3, C4, C5
44.7kResistor_SMD:R_0805_2012MetricR1, R2, R3, R4
4tmc2209TMC2209:TMC2209_SilentStepStick_Socket_18PU1, U2, U3, U4
2100RResistor_SMD:R_0603_1608MetricR10, R11
2Conn_01x22_SocketConnector_PinSocket_2.54mm:PinSocket_1x22_P2.54mm_VerticalJ1, J2
2TCA9548A_MuxConnector_PinSocket_2.54mm:PinSocket_1x12_P2.54mm_VerticalJ20, J24
11.5Asww1235 KiCAD-Libraries master SW-Littelfuse:NANO-0154-OMNI-BLOK-Fuse-HolderF2
1100uF/16VCapacitor_THT:CP_Radial_D6.3mm_P2.50mmC10
11kResistor_SMD:R_0805_2012MetricR33
11k RResistor_SMD:R_0603_1608MetricR12
122uFCapacitor_SMD:C_0805_2012MetricC12
124V_InTerminalBlock:TerminalBlock_MaiXu_MX126-5.0-02P_1x02_P5.00mmJ38
1470uF/50VCapacitor_THT:CP_Radial_D10.0mm_P5.00mmC1
14APanTilt:Fuseholder_Blade_Mini_XFCN_XF-508P-B-BF1
15V_InTerminalBlock:TerminalBlock_MaiXu_MX126-5.0-02P_1x02_P5.00mmJ37
1AS5600_PanConnector_JST:JST_XH_B4B-XH-A_1x04_P2.50mm_VerticalJ21
1AS5600_TiltConnector_JST:JST_XH_B4B-XH-A_1x04_P2.50mm_VerticalJ22
1Aux_MaxConnector_JST:JST_XH_B2B-XH-A_1x02_P2.50mm_VerticalJ41
1Btn_ClearKeyframeConnector_JST:JST_XH_B2B-XH-A_1x02_P2.50mm_VerticalJ34
1Btn_PlayPauseConnector_JST:JST_XH_B2B-XH-A_1x02_P2.50mm_VerticalJ35
1Btn_ResetConnector_JST:JST_XH_B2B-XH-A_1x02_P2.50mm_VerticalJ36
1Btn_SetKeyframeConnector_JST:JST_XH_B2B-XH-A_1x02_P2.50mm_VerticalJ33
1Enc_AuxConnector_JST:JST_XH_B3B-XH-A_1x03_P2.50mm_VerticalJ32
1Enc_PanConnector_JST:JST_XH_B3B-XH-A_1x03_P2.50mm_VerticalJ30
1Enc_SlideConnector_JST:JST_XH_B3B-XH-A_1x03_P2.50mm_VerticalJ29
1Enc_TiltConnector_JST:JST_XH_B3B-XH-A_1x03_P2.50mm_VerticalJ31
1LED_24VLED_SMD:LED_0805_2012MetricLED1
1LED_3V3LED_SMD:LED_0805_2012MetricLED3
1LED_5VLED_SMD:LED_0805_2012MetricLED2
1Limit_MaxConnector_JST:JST_XH_B2B-XH-A_1x02_P2.50mm_VerticalJ27
1Limit_MinConnector_JST:JST_XH_B2B-XH-A_1x02_P2.50mm_VerticalJ26
1MUX_2Connector_PinSocket_2.54mm:PinSocket_1x04_P2.54mm_VerticalJ3
1MUX_3Connector_PinSocket_2.54mm:PinSocket_1x04_P2.54mm_VerticalJ4
1MUX_4Connector_PinSocket_2.54mm:PinSocket_1x04_P2.54mm_VerticalJ5
1MUX_5Connector_PinSocket_2.54mm:PinSocket_1x04_P2.54mm_VerticalJ6
1MUX_6Connector_PinSocket_2.54mm:PinSocket_1x04_P2.54mm_VerticalJ7
1MUX_7Connector_PinSocket_2.54mm:PinSocket_1x04_P2.54mm_VerticalJ8
1Motor_AuxConnector_JST:JST_XH_B4B-XH-A_1x04_P2.50mm_VerticalJ28
1Motor_PanConnector_JST:JST_XH_B4B-XH-A_1x04_P2.50mm_VerticalJ18
1Motor_SlideConnector_JST:JST_XH_B4B-XH-A_1x04_P2.50mm_VerticalJ17
1Motor_TiltConnector_JST:JST_XH_B4B-XH-A_1x04_P2.50mm_VerticalJ19
1OLEDConnector_PinSocket_2.54mm:PinSocket_1x04_P2.54mm_VerticalJ23
1SMBJ33ADiode_SMD:D_SMBD1
1SMBJ6.5ADiode_SMD:D_SMBD2

The 81 most common of 84 parts.

More from AMOVFC/PanTilt

6
  • 4 layersPantiltslide integrated, the printed circuit board in 3D

    Pantiltslide integrated

    AMOVFC/PanTilt

    Three projects live in this folder. They share NewLibrary.kicadsym, TMC2209.pretty and the same philosophy: nothing on the board needs precision placement, every active part is a module in a socket.

    86 parts
  • 4 layers52.3 × 93.2 mmPantiltslide full mini extra, the printed circuit board in 3D

    Pantiltslide full mini extra

    AMOVFC/PanTilt

    Three projects live in this folder. They share NewLibrary.kicadsym, TMC2209.pretty and the same philosophy: nothing on the board needs precision placement, every active part is a module in a socket.

    73 parts
  • 4 layers52.3 × 93.2 mmPantiltslide full turnkey mini, the printed circuit board in 3D

    Pantiltslide full turnkey mini

    AMOVFC/PanTilt

    Three projects live in this folder. They share NewLibrary.kicadsym, TMC2209.pretty and the same philosophy: nothing on the board needs precision placement, every active part is a module in a socket.

    66 parts
  • 4 layers89.5 × 113 mmPantiltslide full turnkey, the printed circuit board in 3D

    Pantiltslide full turnkey

    AMOVFC/PanTilt

    Three projects live in this folder. They share NewLibrary.kicadsym, TMC2209.pretty and the same philosophy: nothing on the board needs precision placement, every active part is a module in a socket.

    57 parts
  • 4 layers123 × 113 mmPantiltslide full pwr, the printed circuit board in 3D

    Pantiltslide full pwr

    AMOVFC/PanTilt

    Three projects live in this folder. They share NewLibrary.kicadsym, TMC2209.pretty and the same philosophy: nothing on the board needs precision placement, every active part is a module in a socket.

    53 parts
  • 4 layers76 × 113 mmPantiltslide full, the printed circuit board in 3D

    Pantiltslide full

    AMOVFC/PanTilt

    Three projects live in this folder. They share NewLibrary.kicadsym, TMC2209.pretty and the same philosophy: nothing on the board needs precision placement, every active part is a module in a socket.

    28 parts

Similar boards

  • 4 layers100 × 100 mmLabathome pcb, the printed circuit board in 3D

    Labathome pcb

    klaus-liebler/labathome

    Der kleine Bruder von lab@home. Ermöglicht das Experimentieren mit PT1...PTn-Gliedern, die intern als RC-Kombination (entkoppelt über Operationsverstärker) ausgeführt sind.

    292 parts5MIT
  • 2 layers98 × 98 mmPixel Tree Controller, the printed circuit board in 3D

    Pixel Tree Controller

    Ben-BJD/Pixel-Tree-Controller

    Safely power 800 LEDs with one 22A ESP32 WLED controller

    28 parts
  • 2 layers100 × 100 mmMach robot, the printed circuit board in 3D

    Mach robot

    vinhpham12/Quadruped_Robot

    178 parts
  • 2 layers59 × 29 mmTable display, the printed circuit board in 3D

    Table display

    yeyintlwin/kicad-board-design

    Design a two-layer KiCad PCB entirely from Python scripts - no GUI step, no commercial autorouter

    53 partsMIT
  • 2 layers80 × 55 mmIm01, the printed circuit board in 3D

    Im01

    amuzdgp3-prog/im-01-pulse-device

    Отдельный модуль к системе обслуживания аппаратов 2.apixspb.ru. Устройство слушает импульсные выходы монетоприёмника и купюроприёмника, копит показания в FRAM и дважды в сутки передаёт их на сервер по LTE Cat.1.

    108 parts
  • 2 layers100 × 65 mmAstroplant AL, the printed circuit board in 3D

    Astroplant AL

    AstroPlant/astroplant-hardware

    Raw data files for hardware PCB's

    102 parts3
  • 4 layers275 × 240 mmSupervisor, the printed circuit board in 3D

    Supervisor

    BennetLeff/temper

    a breville control freak ++

    440 parts1
  • 4 layers87.7 × 80.2 mmLedcube, the printed circuit board in 3D

    Ledcube

    Flawake/Ledcube

    287 parts

Questions

How big is the Pantilt controller PCB?

It measures 52.3 × 93.2 mm and has 4 copper layers, 1.6 mm thick.

What parts does the Pantilt controller use?

84 parts, built around the 2.54-1*8P母 5.0H, CC0805KRX7R9BB104, KM107M050F12RR0VH2FP0 and RCA03100RFLF. The full list of its parts is on this page, with links to each part's symbol and footprint where PCBKit's library has them.

Where can I download the Pantilt controller KiCad files?

From the AMOVFC/PanTilt repository on GitHub, at the commit this page was made from: the board is hardware/pantilt-controller.kicad_pcb and its schematic hardware/pantilt-controller.kicad_sch. PCBKit opens the board in the browser, with nothing to install.

Can I use the Pantilt controller design in my own project?

Its repository states no licence, so its author keeps all rights: ask AMOVFC before reusing the design.