Foxmox

FoxMox hardware

This directory contains the complete hardware design and its source references for Larry Benko W0QE's Low Power FoxMOX Controller V2.6 (December 2004), transcribed by Chris Keller K0SWE.

Open foxmox.kicad_pro from this directory. Library references use ${KIPRJMOD}/Libraries/..., so moving the project and Libraries/ together preserves their relative paths.

The target radio in this build is the Alinco DJ-S11, connected through its 2.5 mm microphone jack.

How the controller keys the radio

The DJ-S11 microphone jack has ground, a bias rail, and audio, but no separate PTT contact. The microphone tip serves as both the audio input and PTT sense. The radio biases that contact to power an electret microphone and watches its DC level: pulling it toward ground through a resistor keys the transmitter; releasing it returns the radio to receive. Transmit audio is AC-coupled onto the same contact.

This resistive keying arrangement is common on Alinco, Icom, and Yaesu handheld microphone interfaces. It does not depend on a VOX setting in the radio; VOX accessories for these radios implement their own resistive pull-down. Kenwood accessory wiring is a notable exception because it commonly provides a separate PTT contact.

Audio path — PIC RA0, pin 17

RA0 ── R1 20k ── R2 5k pot ──(wiper)── C4 1 µF ── P1 tip
                      └── bottom ── GND

RA0 generates the square-wave tone. R1 and R2 attenuate it, while C4 blocks DC so the tone does not disturb the keying level.

W0QE annotated the original drawing:

Adjust for 20 to 40 mV peak to peak. Or set R2 for ~300 ohms from wiper to ground.

That setting is internally consistent: 3.0 V × 300 / 25000 ≈ 36 mV. Avoid excessive drive; the source waveform is already square and harmonic-rich.

PTT path — PIC RA1, pin 18

RA1 ── R3 20k ── Q1 base (2N2222A, emitter → GND)
                 Q1 collector ── R4 2.7k ── P1 tip

RA1 high saturates Q1 and pulls the microphone tip toward ground through R4. R4, 2.7 kΩ, is the radio's keying resistor. RA1 low turns Q1 off and allows the microphone contact to return to its bias voltage.

Both paths meet at the microphone tip. In the PCB netlist, Net-(C4-Pad2) contains C4.2, P1.T, and R4.2.

Power source and current budget

The controller has no battery of its own. It is parasitically powered by the radio's microphone-bias rail:

P1 ring ── S3 (SPDT power switch) ── VDD ── U1.14, U1.4 (MCLR), C3, D1, TP1
P1 sleeve ── GND

W0QE recorded:

Measured current draw @ 3.0 V from Ring to Sleeve = 0.5 mA idle, 0.9 mA transmitting

This sub-milliamp supply is a binding design constraint. It explains the PIC16F84, resistively attenuated square-wave audio, and absence of more power-hungry circuitry. Before substituting another MCU, verify that its complete idle and transmit current does not collapse the microphone-bias rail or false-key the radio.

D1 is a 6.2 V zener used for overvoltage protection. It is not regulating the normal approximately 3 V supply.

R5 is the boot/service-mode strap

R5, 10 kΩ, is intentionally connected between RA2 (PIC pin 1) and RA3 (PIC pin 2). RA4/T0CKI (PIC pin 3) is unconnected. Do not move R5 to RA4.

The original scan is easy to misread:

1. The R5 10k text is printed below its resistor body, placing the label near the RA4 row even though the component itself connects RA2 to RA3. Component label placement in this drawing is not a reliable indication of connectivity.

2. The X on the RA4 row is a printed no-connect mark. The pin stub ends at that mark; it is not a wire crossed out by hand. Pixel measurements support that reading: its darkness matches printed material rather than the lighter pencil handwriting elsewhere on the scan.

   Scan feature	Mean gray
   Printed wire	73
   RA4 X mark	81
   Printed text	89
   Pencil handwriting at top margin	140

The reconstructed schematic has an explicit RA4 no-connect at (97.79, 105.41); with U1 placed at (80.01, 110.49), that location resolves to PIC pin 3. The PCB independently names the pad unconnected-(U1-TOCKI{slash}RA4-Pad3). The reconstructed schematic and PCB therefore agree.

The firmware also proves R5's purpose: at reset RA3 is driven high while RA2 is an input, so R5 pulls RA2 high for normal startup. Holding RA2 low during startup enters the clock-calibration service mode; the firmware waits for RA2 to be released before continuing. Moving R5 to RA4 would break that boot test.

The service-mode procedure is documented in ../Firmware/README.md.

Bench checks with an Alinco DJ-S11

Before relying on a rebuilt controller in the field, check it with the actual radio and cable:

1. Bias supply: measure approximately 3 V from ring to sleeve and confirm it can source at least 1 mA without excessive sag.
2. PTT keying: confirm that the 2.7 kΩ pull-down keys reliably with both a fresh radio battery and a nearly discharged one. The microphone-bias voltage can move with battery condition.
3. Audio level: begin with R2 adjusted to approximately 300 Ω from wiper to ground, then verify 20–40 mV peak-to-peak or check deviation on a service monitor or trusted receiver.

Verifying the transcription

The scanned schematic is lossy and its geometry is easy to misread. Use these sources in descending order of confidence:

1. Treat W0QE FoxMox scanned.pdf as the source of truth for the original design.

2. Use the resolved nets in foxmox.kicad_pcb when checking the reconstructed connectivity. Footprint pad records carry their net names and avoid visual wire-tracing mistakes.

3. Cross-check hardware-dependent behavior against ../Firmware/foxmox-v2.6.asm.

4. For close inspection of the scan, render it at high resolution:

   pdftoppm -r 300 -png "Hardware/W0QE FoxMox scanned.pdf" out

Hardware/.history/ and *.kicad_prl are intentionally ignored by the outer repository. They are local KiCad state, not versioned design artifacts.