QMX bench log
Tick each step as you finish it, and add your own notes under every stage. Your ticks and notes are saved automatically. Manual 1.00j and PA transformer manual 1.00d.
Before you start
This plan takes your Rev 1 PCB, built as the 80/60/40/30/20m version for a 12V supply, from parts bag to first QSO in six stages. Builders report 10 to 12 hours of work, so treat it as several sessions, not one.
Set-up and ground rules
| Decision | Choice | Why |
|---|---|---|
| PCB and manual | Rev 1, assembly manual 1.00j | Later manuals have different steps |
| Bands | 80, 60, 40, 30, 20m | Only version shipped at launch |
| Supply | 12V | Needs a 2-turn secondary on T501 (3:2 ratio) |
| T501 | Conventional 3:2, 0.60mm wire | No drill needed; fine below 12m |
| Firmware | Latest stable, 1_04_010 at time of writing | 1_04_011 to 1_04_015 are marked beta |
Tools and supplies
Keep open while you build
- Rev 1 assembly manual 1.00j and PA transformer manual 1.00d
- Rev 1 searchable schematic and searchable layout for Ctrl-F on component numbers
- Operating manual and the troubleshooting page
- Rev 1 schematics (5 pages). Component numbers start with the page number, so 1xx is page 1 (power supplies), 4xx the receiver and 5xx the transmitter
- "QMX Schematic for Dummies", a one-page overview of the whole circuit
Rules for the whole build
- Check every value and position twice before soldering; removing parts from this board is hard.
- Inspect each joint under magnification straight after soldering.
- Ground pins need much more heat than you expect.
- Never drill the board; there are hidden internal layers.
- Keep the capacitor lead off-cuts, because you use them to fit the LCD.
- Scrape enamelled wire, or hold the iron on it with plenty of solder for about 10 seconds, to burn the enamel off. Don't tin 0.60mm wire before fitting it.
One-page bench checklist
The must-do checks, in build order. Tick each before moving to the next stage.
What applies to a Rev 1 board
Only three items need action from you: the Q103/Q104 test, the USB-C check, and the optional PTT jumper. The rest are either already in your manual or don't concern Rev 1.
| Item | Applies to Rev 1? | What to do |
|---|---|---|
| 20m sensitivity mod | Already in manual 1.00j | Wind L401 as 28 turns tapped at 19. Firmware 1_00_003 and later sets the Band Config to match |
| C525 and C516 swapped | Fixed in manual 1.00j | C525 is 270pF, C516 is 180pF; check labels "271" and "181" |
| Q103/Q104 drain short on 5V SMPS board | Yes, Rev 1 specific | Test the bare board with a multimeter before building (Stage 0) |
| USB-C not recognised on USB-C-only laptops | Yes, Rev 1 only | Test early; if it fails add two 5.1k resistors on the CC pins, or use a USB-A to USB-C cable |
| PTT mod | Yes, Rev 1 to 3 | Optional. Jumper pin 2 to pin 4 of X201 after it is fitted; only needed for the grounded PTT output |
| Rev 2 diode mod (IC402 protection) | No | QRP Labs states there is no issue with Rev 1 PCBs |
| SMPS step change | Applies to everyone | Never jump the supply from a low voltage straight to 12V; power off first when changing voltage |
| Zener D109 | Applies to everyone | If a zener ever needs replacing, use a 3.6V 500mW part, not a higher wattage one |
| PA transistor doubling (PA0RDT) | Optional, later | Not for the first build; it works on QMX but space is very tight |
| RTC and 40-10m mods | No | Ignore |
The Rev 1 to Rev 2 change list also covers easier-to-solder encoder grounds, larger BS170 holes and a moved +12V pad. These explain why Rev 1 is fiddlier but don't change what you do.
Prep and bare-board checks
Do all of this before any soldering. It takes an evening and catches the two Rev 1 problems that can destroy the board.
Capacitor identification (80-20m)
| Value | Label on body | Positions |
|---|---|---|
| 30pF | 300 | C401 |
| 56pF | 56J | C521, C404 |
| 82pF | 820 | C522 |
| 100pF | 101 | C520 |
| 180pF | 181 | C516 |
| 220pF | 221 | C402 |
| 270pF | 271 | C525 |
| 390pF | 391 | C515, C524 |
| 470pF | 471 | C523, C519 |
| 820pF | 821 | C518, C514 |
| 1200pF | 122 | C517 |
Bare-board Q103/Q104 test
Solder can bridge the large drain tab of Q103 (the big AOD403) to the drain pin of Q104 (the small transistor next to it). Look at the two under magnification, then set your multimeter to continuity and touch the Q103 drain tab and Q104's drain pin. A beep or zero ohms means the short exists.
If it does, heat Q104 and slide it about 1mm away from Q103. Q104 is a small SOT-23 part, and the drain is normally pin 3, but confirm that on the Rev 1 layout PDF because I couldn't see the manual's marked photo. Repeat the test right before first power-up.
Notes: Bare-board Q103/Q104 test
Stage 0 notes
Low-profile parts
Manual 2.5 to 2.8These are the easy parts, but they set up everything after, so orientation and clearance matter more than speed.
Tips from other builders
- Sit the capacitors and diodes a fraction proud of the board, so there's room to move toroid wire around later.
- One builder suggests checking where L507 will fit before fitting C514, C515 and C517, since their board's LPF toroids then wouldn't sit as shown. Their revision may differ, so treat it as a prompt to check, not a rule.
- D512 and D513 sit under the right-hand rotary encoder. Another builder had to bend D513's lead sideways, and taped the encoder, so the encoder body can't touch it. Think about this now, and leave those two diodes leaning away from the encoder position.
Checkpoint
Stage 1 notes
Transformers and toroids
Manual 2.9 to 2.14Most build failures trace back to this stage, usually enamel left on a wire or a toroid in the wrong place. Do one component at a time, test continuity after each, and don't move on until it passes.
Wire budget
The 0.33mm wire is almost all used. By my sums from the manual's lengths, the toroids need about 353cm of the 365cm supplied, so cut to the stated length and don't waste any.
| Use | Wire | Length |
|---|---|---|
| L401 | 0.33mm | about 60cm |
| LPF toroids (six) | 0.33mm | 37, 40, 26, 27, 18, 20cm (168cm) |
| T401 | 0.33mm | 3 x 25cm |
| T507 | 0.33mm | 2 x 25cm |
| T507 single turns | 0.60mm | 2 x 25mm |
| T501 and L502 | 0.60mm | Share the rest of the 65cm. T501's length isn't specified, so keep it tight |
Wind in the direction that puts the wire ends near their holes: start from the top, down through the hole, then around and over, accumulating turns anticlockwise. Each pass through the centre counts as one turn.
- T501 output transformer, 3:2 for 12V (2.9 and PA transformer manual)
The primary is always 3 turns with a centre tap. The secondary is 2 turns for 12V. Never fit a 3-turn secondary on a 12V build, because it will likely destroy the PA transistors.
If you later get a drill, the twisted RWTST version is the recommended 12V design, but it isn't needed for 80-20m.
Notes: 1. T501 output transformer, 3:2 for 12V
- L401 tapped band-pass inductor (2.10)
Notes: 2. L401 tapped band-pass inductor
- L502 (2.11)
10 turns of 0.60mm wire on an FT37-43 (dull black). Scrape the enamel, don't tin the wire, cut to 2mm, solder 10 seconds.
- Six low-pass filter toroids (2.12)
All are yellow T30-6 cores wound with 0.33mm wire. Spread the turns evenly with a small gap; slightly tighter is a little better, but don't snap the wire.
| Band | Part | Turns | Wire |
|---|---|---|---|
| 80m | L511 | 26 | 37cm |
| 80m | L506 | 28 | 40cm |
| 60 and 40m | L512 | 17 | 26cm |
| 60 and 40m | L508 | 18 | 27cm |
| 30 and 20m | L513 | 10 | 18cm |
| 30 and 20m | L510 | 12 | 20cm |
- Count out loud. A wrong turn count is hard to find later.
- Cutting the two leads to different lengths makes them easier to insert.
- 28 turns on a T30 is tight; one builder couldn't keep all turns neat. Keep them even and don't over-tighten.
- Test continuity end to end on each toroid. Poor connections here cause low power on whole bands.
- T401 trifilar transformer (2.13)
Notes: 5. T401 trifilar transformer
- T507 (2.14)
Notes: 6. T507
Checkpoint
Stage 2 notes
Connectors, headers and SMPS boards
Manual 2.15 to 2.17The SMPS boards are, in the manual's words, the least forgiving part of the build: one short here can destroy the whole radio. Go slowly and inspect everything under magnification.
Connectors (2.15)
Notes: Connectors
Female header sockets (2.16)
Fit the 2x5, 2x4 and 2x2 female sockets. One builder fitted male pins where the female sockets belong, which is a painful mistake to undo. Keep them square.
SMPS boards (2.17)
Notes: SMPS boards
Checkpoint
Stage 3 notes
LCD, controls board and fit check
Manual 2.18 to 2.28This stage is mechanical, and the clearances are tight. Test-fit the boards before you solder anything that can't be undone.
Display board (2.18 to 2.21)
Notes: Display board
Controls board (2.22 to 2.26)
Notes: Controls board
Fit and join (2.27 and 2.28)
Notes: Fit and join
Enclosure (2.29)
The enclosure is optional. If you have it: slide the display board into the guide rails, bolt the right-hand panel to the BNC with its washer and nut, then follow the manual's photo sequence. Fit the knobs last, with a small gap to the front panel, and put the volume knob's grub screw on the shaft's flat.
My suggestion: bench-test the radio unboxed on a non-conductive surface first, so you can reach the board. Fit the case only after the sweeps pass.
Checkpoint
Stage 4 notes
First power-up, firmware and sweeps
Manual 2.31 to 2.33The radio has no firmware from the factory, so the first power-up should show zero current and a USB drive. Anything else means stop and check.
Final checks before power (2.31)
Notes: Final checks before power
First power-up
Notes: First power-up
Terminal tests
Connect PuTTY (or similar) to the QMX serial port. The built-in tools let you test without extra equipment. Work in this order, because there's no point debugging the receiver until the transmitter works:
Notes: Terminal tests
Rev 1 items to do now
Notes: Rev 1 items to do now
Stage 5 notes
Settings and operating tips for a 12V build
Keep the supply at a true 12V and the radio will look after itself. Most operating problems reported by builders came from supply voltage or audio settings.
- Supply: QRP Labs says 12V means 12V, not a 13.8V radio supply. A fully charged "12V" battery can sit several volts higher. To drop voltage, series 1N4001-type diodes each drop about 0.6V.
- Max PA voltage: firmware 1_03_000 added a Protection menu setting that caps the voltage the PA sees. One builder reports a default of 11.5V. Check the value on your radio before connecting a battery.
- Clean power: the supply should be spike-free. Power off with a long press of the left encoder, not by pulling the plug.
- SWR protection: an 'S' on the display means protection has tripped. Enter and leave the menu, or cycle the power, to clear it. It trips on a high SWR, and the threshold is set in the Protection menu.
- GPS and paddle share a jack: plugging a GPS into the paddle port puts the radio into practice mode ('G'). Connect it only for calibration or beacon use.
- Audio: the earphone output is for 32-ohm earphones. An amplified speaker is the manual's advice for loudspeakers.
- Audio clicks or thump: several builders reported band-dependent clicking tied to sidetone and AGC settings. The forum has threads with settings that fixed it. The transmitted CW stays clean.
- Firmware: updates are free. Stay on the stable releases unless a beta fixes something you need.
- SSB: use an external electret microphone on the paddle jack (mic on the ring, PTT on the tip; the QMX supplies a 2.2k pull-up). QRP Labs doesn't sell a mic, but the DL9SCO Mini Mic and a K6ARK 3D-printed one exist. Run the SSB Calibration tool in the Hardware tests menu.
Later, optional
- Doubling the BS170s (PA0RDT) is reported to work on QMX, but space is very tight and it changes the PA, so leave it until the radio works as built.
- A 3D-printed cover and rails by Tomasz Pabich are linked from the QMX page.
My notes
Troubleshooting quick reference
QRP Labs' own advice is that a carefully built QMX must work, and most faults are soldering. Fix in this order: basic operation, then transmit, then receive.
| Symptom | Likely cause | Check |
|---|---|---|
| Nothing happens on left encoder press, no current | Left encoder pins unsoldered; four pins on the bottom of the 5V SMPS board unsoldered; Q105 source joint | Reflow the encoder and SMPS pins, and check Q105 |
| Backlight on, no text | Contrast trimmer not set; 2x2 right-hand header not soldered | Turn R305, inspect the header |
| Blank screen, shows as USB drive | Normal for new radios | Install the firmware |
| Not seen as a USB drive | Short at USB connector pins | Inspect J201 for a bridge |
| "1." on top row, bottom row empty | Right encoder button shorted to its body | Look for a solder blob |
| Radio switches itself on | Left encoder (S304) button shorted to its body | Same as above |
| Low power on some bands | Enamel not burned off, a diode backwards, swapped capacitors, solder whisker | D508 to D513 orientation, C524 and C521 values, L513 joints, whiskers across D509 |
| Deaf on 40, 60 and 80m | Enamel on the L401 tap or end | Reflow the L401 joints |
| Weak 20m receive | L401 mod not matched by the Band Config | Factory reset, or edit the Band Config |
| Low power on every band, supply reading red at about 4.9V | Damaged R203 | Inspect R203 (47k) for heat damage |
| Backlit keys on your keyboard go dark | Whisker shorting USB power to ground | Inspect the USB joints |
| Radio not recognised on a USB-C-only laptop | Rev 1 missing CC resistors | Add 5.1k to the CC pins, or use a USB-A cable |
A reflow of every toroid joint is the forum's first suggestion when all filter sweeps look wrong. Experienced members link audio-sweep faults to T401 and RF sweep faults to the low-pass filters or L401.
My notes
Open questions and gaps
- Hans's 52 tips are folded in. Three points are still unchecked against the video: tip 21 (the L401 adjustment tip), tip 33 (which bolt it means) and the wording of tip 30.
- Two groups.io threads and one blog were blocked when I tried to open them. I used their search snippets, so I've flagged those tips as builder reports.
- I found no published list of changes between PCB Rev 2 and Rev 5, so I can't say what else changed.
- The manual doesn't give a wire length for the conventional T501. Measure as you go, and keep enough for L502 and T507.
- You have bought the enclosure and have your own enamelled wire in reserve. I haven't confirmed the diameter of the kit's wire or your spare, so measure both with calipers.
- The Rev 1 schematic (page 1) confirms Q103 and Q105 are the large AOD403s and Q104 is the small BSS123. Q104's drain pin number (pin 3) is still from general knowledge of the SOT-23 package, not the manual or the schematic.
My notes
Later: a bigger screen with the Tab5 panadapter
The QMX's own display can't be swapped for a better one, but an M5Stack Tab5 touchscreen can sit beside it and act as a spectrum display and control panel. Add it only after the radio works on its own.
What it is. OZ1LAV's QMX Panadapter is firmware for the Tab5, a 5-inch 720x1280 touchscreen (ESP32-P4). The Tab5 acts as a USB host and plugs into the QMX's USB-C port with one USB-A to USB-C cable, so nothing on the QMX board is modified.
How it talks to the radio. The QMX exposes two USB interfaces: 48kHz 24-bit I/Q stereo audio, which the Tab5 turns into a spectrum and waterfall, and a serial port carrying Kenwood-style CAT commands for control.
What it adds
- A 48kHz spectrum and waterfall with tap-to-tune, pinch zoom, an S-meter and a band-plan strip.
- Frequency, mode, filter, power and SWR control, band presets and 32 memory channels.
- The QMX's own menus on the big screen, over its second USB serial port.
- FT8, FT4 and WSPR on the device with no PC, with ADIF logging and uploads to QRZ, eQSL, LoTW and Cloudlog/Wavelog. JS8 is in beta.
- Decoded CW along the bottom of the waterfall (QMX firmware 1.03 and later), plus AM on firmware 1.04 and later.
- A browser version of the same screen on your WiFi, and optional POTA, RBN and DX cluster spots.
What you need. A Tab5 (about $60, per one builder's page; I haven't checked the price) and a USB-A to USB-C cable. A microSD card, WiFi and a GPS unit are optional. The panadapter is open source on GitHub and builds with ESP-IDF, but you shouldn't need to build it yourself.
Cautions
- Rev 1 and the firmware: The project says it supports the "QMX/QMX+". The builder write-ups I found are for the QMX+, so I haven't confirmed it on a Rev 1 board. Test it with your own radio, and check the project's Quick Start for the exact QMX settings it needs.
- Beta features: Tab5 audio (including binaural CW and the headphone-jack receive path) and JS8 are beta. Nothing has been decoded through the Tab5's jack yet, and audio breaks up while a browser has the panadapter open.
- The view: The QMX's oscillator sits 12kHz below the dial, so the part of the display above dial +12kHz is hatched and can't show real signals.
- Updates: If you ever see a v1.14.x or earlier version, it needs one USB-C cable update before moving to v1.15.0 or later. After that, updates are over the air. New users aren't affected.
When to add it. First finish Stage 5, get a clean RF sweep, make a few contacts using the QMX's own display and WSJT-X, and back up your configuration. After that, a Tab5 can't affect the board, so you can try it at no risk to the radio.
Sources
Pages opened in full for this plan:
- QMX kit page and firmware log
- Rev 1 assembly manual 1.00j
- PA transformer manual 1.00d
- Operating manual, firmware 1_04_004
- QMX troubleshooting
- 20m sensitivity modification
- PTT modification
- Rev 2 modification page
- SMPS step change
- Zener diode D109
- Other modifications
- PCB Rev 1 to Rev 2 changes
- AA7EE build write-up
Also read in full for the builder-lesson boxes: M7MCQ build log, and the PH2LB build log (page last updated 2023-08-17), whose text Carl supplied because the site blocks automated fetching. Both also appear in the snippet list below, from the earlier pass.
Tab5 section: the QMX Panadapter project page (v1.17.1) and the project text you supplied.
From search snippets only: