QMX – From Kit to Field!

by Matt (W6CSN)

The Radio

The QMX by QRP Labs is a five band, multi-mode radio introduced by Hans Summers G0UPL at FDIM in 2023. The QMX is the next logical step in the radio development journey at QRP Labs, bringing together the innovative approach to FSK modes like FT8 from the QDX and the CW performance of the QCX series. This is all done in same enclosure as the QCX-Mini which is not much larger than a deck of playing cards.

Size of the QMX compared to standard deck of playing cards.
The QMX is about the size of a deck of cards.

The keen observer will notice that the QMX also sports a built-in microphone which, along with associated circuitry, supports future introduction of voice modes by way of firmware update. Other notable features include SWR metering with protection as well as solid-state “PIN” diode T/R switching, plus the option for “high band” coverage of 20 through 10 meters in addition to the original 80m-20m version.

Most hams have, at one point or other, forgotten to throw the antenna switch, adjust the tuner, or even connect an antenna before transmitting. The SWR meter and protection really sets the QMX apart from earlier QRP Labs radios. There are plenty of sad tales on the QRP-Labs forums from QCX or QDX users that “smoked” the BS170 mosfet finals in a moment of operating into a badly mismatched load. It’s remarkably easy to do, ask me how I know!

The Build

Based on experience with the QCX-Mini and having put together several QDX’s, I admit to being hesitant to starting assembly of this radio. The level of integration in the QMX as a multi-band, multi-mode unit is probably the highest yet to come out of QRP Labs. The components that were merely small in earlier radios are absolutely tiny in the QMX, specifically the LPF toroidal cores. The main board, internal switching power supplies, display, and controls board are all sandwiched together in a tight fit for the custom enclosure.

photo showing a partially completed circuit board
Most of the electronic components installed on the main board.

The build started with a slow and methodical approach of doing a little bit each day and working in the morning when my mind was fresh and there was good light on my workbench, a.k.a. the dining room table.

The first “disaster” happened when performing the factory recommended modification to Revision 2 boards shipped in 2023. The mod calls for a protection diode to be installed across a SMD mosfet. Several leaded 1N4148 diodes were available in my “junque” box so I attempted to carefully fit one of these in the right position on the board. While soldering the diode in place I managed to lift C508, a microscopic .1uF SMD capacitor, clean off the board!

photo of circuit board with modifications
The suggested modification adding a protection diode.

Given the tools at my disposal, there was no way I was going to be able to get that capacitor back into place. Therefore, I grabbed a standard through hole .1uF cap and painstakingly got it connected in the right place, verified by lots of continuity testing.

Once finished with all the electronic components on the main board, I was feeling pretty cocky and also the completion of the project was more clearly in view. The cautious and methodical approach gave way to a faster build pace, which directly lead to the second “disaster.” While installing the headers that connect the main board to the display board, I failed to CAREFULLY READ THE ASSEMBLY MANUAL and soldered the male pin headers where I should have installed the female sockets!

This misstep might have ended the project right then and there if I had not at some point in the last year bought a proper desoldering tool. This is not an expensive automatic vacuum pump powered solder re-work stations, but rather a heating tip and manually actuated solder sucker built into one tool. Twenty-two unsoldered connections later, we were back on track.

The remainder of the assembly was pretty much drama-free. I did manage to break a couple of the leads on the rotary encoders (easily fixed by replacing the lead with a component cutoff) and it is kind of tricky to get the controls board properly through the hole in the display board and get the whole thing buttoned up in the enclosure. I’m still not convinced I got the fit 100% right.

photo of the complete QMX not installed in the custom enclosure
After firmware installation the radio comes to life.

With the build complete, and the radio connected to a dummy load, I applied a reduced voltage DC supply of 9v, and good news! No smoke!

I pressed the power switch and the QMX showed up on my computers as a USB flash drive. The firmware installs easily per the instructions in the manual after which the radio power cycles and the display comes to life.

First off, I performed audio and rf filter sweeps on all bands using the in-built terminal application and recorded screenshots to use as a baseline if I ever decide to try to tweak any of those response curves. In all honesty there is not a lot of physical adjustment to make other than compressing or spreading the turns on the toroids. And given they are pretty hard to reach, I likely will not make adjustments unless field testing proves that something is seriously misaligned.

photo of the finished QMX connected to a power meter
Checking the power output of the QMX.

Next, I tested the radio at the full 12v (still on the dummy load) and measured the following CW transmit power results:

20M – 3.6W
17M – 4.4W
15M – 5.9W
12M – 2.9W
10M – 3.6W

All are very acceptable power levels for CW and FT8. For some reason this particular QMX is an absolute blowtorch on 15 meters!

Field Testing

As of this writing, the QMX has been on two field test outings at my local POTA spot in the Presidio of San Francisco (K-7889).

photo of QMX, battery, and key in a small zippered pouch
The QMX “field kit” is ready to go.

The first on-air test of this “high band” QMX was with an MFJ-1979 17 foot telescoping whip antenna clamped to a short steel fence pipe next to San Francisco bay. This installation gives pretty close to a 1:1 SWR on 20 meters with the whip fully extended. By collapsing sections of the whip, a good match can found for the higher bands.

photo of QMX, battery, key, and log book
Making QSOs with the QMX.

The test/activation started on 15 meters and the first two-way contact on the QMX was with W8NGA from Ohio on CW. Eleven more QSOs on 21 Mhz, including DX from Japan, proved that the radio worked well. Up to 12m for a several QSOs and then 28Mhz for a handful more. Working DX to Chile with CE3FZL showed that the QMX as built was sufficiently sensitive to be taken seriously as a QRP rig.

map showing first QMX QSOs including DX to Japan and Chile.
The QMX is quite capable of working DX.

Next, I tuned the antenna for 17 meters but there seems to be some sidetone induced audio clicks that are so strong when using 18 Mhz that it was actually hard to send accurately. I gave up on 17 meters without a single contact before wrapping up the test on 20 meters.

a map showing multiple QSOs
North America QSOs tools.adventureradio.de/analyzer

At home, I next consulted the QRP-Labs forum on groups.io for the audio clicking issue and made some adjustments to the AGC and sidetone settings. Others have reported a similar problem and it only seems to effect the audio channel on the QMX, the transmitted CW signal is clean.

photo of QMX in operation on the center console of a vehicle
The QMX fits easily in the car when outdoor conditions are “hostile.”

The next testing session came a few days later, but conditions were substantially different. Locally, the wind coming through the Golden Gate was strong enough to preclude the use of the MFJ 17 foot whip, the antenna is just too bendy in high winds for my comfort. In addition, solar activity made propagation a bit of a crap shoot. I operated from my vehicle with the Gabil GRA-7350T antenna on a rooftop mag-mount. The good news is that the AGC settings changes substantially improved the audio clicking issue, the bad news was that I was only able to muster a single QSO on 17 meters.

As yet, I have not tested the Digital mode features of the QMX.

Final Thoughts

With its quadrature sampling front end, custom SDR based receiver architecture, plus multi-band and multi-mode capabilities, the QMX is an extremely good value in my opinion. At less than US$200 for a factory assembled unit, this is price competitive with other field radios with less capability that use 20 year old transceiver architectures.

The other side of the coin is that currently you’ll need to either assemble the QMX yourself or accept a lengthy lead time for a factory assembled unit. In fact, the very reason I built the QMX kit was that I wanted to possibly have the radio available for an overseas trip before the QMX on order from last fall arrives. It’s a good performing multi-band radio that if it somehow got damaged, lost, or stolen while traveling then it would not be a huge financial loss. I’d much rather sacrifice a $200 QMX to the perils of world travel than a $1600 KX2.

Another reality is that the QMX is a bit bleeding edge. The firmware, which is a substantial portion of the radio, is under constant development and building and operating the transceiver to its full potential requires some effort. It’s not an “appliance operators” radio and you definitely need to read the manual and be willing to consult the forums when you need support.


With some affiliate links to support QRPer.com

16 thoughts on “QMX – From Kit to Field!”

  1. Matt (W6CNS), great article and good info on the QMX! After numerous hunter contacts with the QMX from home, I used it (low band REV 1 kit with V17 firmware installed) for a POTA activation at the Little Indian River Wildlife Management Area in WV (K-7003). The antenna was a semi-home brew 3 band trap EFHW up 28 ft. For those that worked me on February 21, 2024, at K-7003, you heard the QMX. It worked surprisingly well on CW with the Elecraft T1 tuner and a small buck-boost converter from Amazon to keep the input voltage at about 11.9V. The battery was a Bioenno 3Ah.
    72, all!

  2. Smart looking field kit you’ve curated! I keep debating on the QMX, partially due to the band selections. I realize the solar cycle peak is coming soon, but I would sacrifice 10m for 30m coverage. The assembled rig lead time is also disconcerting, but then again, if I don’t order one, it will never arrive! Thanks for taking the time for the review.

    1. Now you’ve got me thinking, I suspect you MIGHT be able to configure the high band QMX to cover 30 meters instead of 10m without hardware modifications. I think the 20m/17m LPF position may also cover 30m. The other factor is whether the receive bandpass filter is forgiving enough to give sufficient sensitivity on 10 Mhz.

      It might be worth some experimenting with the band configuration, if it doesn’t work well, you can just do a factory reset to the firmware. It would also be worth a discussion in the QRP Labs forums and some careful bench testing before transmitting into an antenna.

      I’m thinking to move the 10m configuration into the 11m slot (yes, the high bander covers 27 mhz out of the box) and put 30 meters in the old 10m position. That would give you the right order when cycling through the bands and you wouldn’t need to sacrifice 10m.

      1. Investigating this idea a little more, and looking at the Band Configuration screen in the QMX terminal application, I see the default config has 17m/15m sharing a low pass filter rather than 20m/17m as I had assumed.

        Also, the 20m receive BPF curve on my QMX is attenuating 10mhz somewhat. But getting these filters to peak at exactly the right spot is tricky.

        I think the key here is to look at the actual filter sweeps and select the filters that provide the best response. And of course, I would not put an alternate band configuration on-air without first verifying that harmonic suppression meets FCC requirements.

  3. Excellent write up on the QMX and a real nice compact go kit for sure!

    But I don’t understand why QRP Labs doesn’t offer a QMX with the most useful bands, considering limited band availability. As it stands, to get the most useful bands, you need two QMX. You end up with a daytime rig and a nighttime rig. The most useful bands for portable field QRP rigs are 40, 30, 20, 17 and 15 meters. 80 meters 5 watt QRP with typical field antennas? No thanks, that’s not my idea of a good time. 60 meters? Hard pass. 12 and 10? Ok, but not as useful as 17/15.

    By all means keep making the HI/LOW band format, but why not the 40, 30, 20, 17 and 15 format for folks who want to just carry one radio? I would have already built one if it was available.

    IMHO, of course????

    73 Kevin K3OX

    1. I agree that 40m – 15m would be the “Goldilocks” band selection for field portable QRP rigs.

      That’s the KH1 coverage, no?

  4. Matt,

    Such a great post. I have built several QCX mini’s and decided to order the factory assembled QMX high band. Your writeup confirmed I made the right decision. Like you, I much prefer these rigs to my KX2 when traveling, and for digital, the QDX is a great rig too. Thanks for your comment on the desoldering pump, I just ordered one for my bench.


    1. I got the desoldering tool to replace the BS170 finals in the QDX that I killed while running FT8 into the wrong antenna one too many times. The tool really saved my bacon on this QMX build however.

  5. Does anyone have a comparison between the QMX transceivers and the (tr)uSDX which has somewhat similar capabilities?

    I ask because I’m drawn by the low-band high-band variants. … and yes, I agree with Kevin K3OX about why not the bands he likes? Maybe we can get to the point where kit builders build the filter board with the bands they really want and switches in the software allow easy configuration.

  6. Matt, Thank you for your report. I found it very insightful.

    My factory-built LoBander is awesome. I’ve had a blast using it with the FT8CN software. My kit uses a K6ARK EFλ/2, a 12v 3aH TalentCell w/DC lead, Palm Pico, USB C/c to a Samsung S20 Ultra running FT8CN… I love this little transceiver!

    72 de W7UDT (dit dit)

  7. Hi Matt,

    The problems you had building the QMX were the exact same problems I had. Messed up the board trying to add the 1N4148 and damaged one of the rotary switches. I ordered the version 3 kit and will use the old version 2 for parts if needed. Will definitely take my time building the new one. I do have a QMX-mini for 20 meters and it works great.

    73 de W9QP

  8. Great article!

    I currently own my 1st QCX-mini on 20M. I hope to become a POTA person this summer.

    I’m waiting for QMX Rev 3 and updated firmware before I make my commitment to purchase.

    All good ideas about changing the band selection to 40M, 30M, 20M, 17M and 15M as well so I would only need to carry a single radio for day & night use. I wonder if this is a consideration by Hans for a future release?

    I’d even love a 20M, 17M, 15M, 10M and 6M bands version as I LOVE the 10M and 6M bands.

    1. Garnet, I hope to hear you on the POTA airwaves soon. The QCX-Mini is a great little radio, and very POTA-worthy, especially with a well matched antenna system.

      If you’re thinking of getting a factory assembled QMX then I’d recommend ordering as soon as you are able as the build queue is now about 6 months.

      And updating the firmware is extremely easy, I wouldn’t wait on the hardware for a specific firmware feature release.

      72, matt

  9. The QMX firmware is coming on in leaps and bounds. It’s already an astounding little radio even before SSB is implemented. I use it more than any other of my QRP radios, including the TX-500!

    At only $115 for the kit version, including the metal case, it’s astounding value for money. The build instructions are about the best I’ve seen and guide you through the relatively easy process in great detail.

    There is talk of a mid-band kit being developed eventually. I have a V1 low-band kit but would happily dump 80m and 60m in favour of 17m and 15m. I really don’t understand why it wasn’t originally developed as a 40m through to 15m rig.

    In response to Bob’s (N4REE) question about comparison with the (tr)uSDX. The simple answer is there’s no comparison. The QMX is a much more refined radio and better equipped in every department. It doesn’t have SSB capability yet but when that’s implemented, the hardware is capable of much better results than the (tr)uSDX. Don’t get me wrong, I like my little (tr)uSDX. Itcs great fun and is a remarkable bit of engineering, but it’s a novelty toy compared to the QMX.

    Dah-dah-diddly Diddly-dah-dah

    1. Thanks Ronan. I appreciate your views of the QMX vs (tr)uSDX. I’ve had a good number of QSOs and a lot of fun with the (tr)uSDX, and more than enough enjoyment building it to be worth the price.

      Thanks & 72

  10. Bill Porter, N6MTB has modified a High Band QMX to a “Mid Band” version (60-15 in his case). He’s got a nice PDF posted with component values and feedback on the QRP-Labs groups.io.

    I’m planning to try a 40/30/20/15m QDX build soon, and if I’m lucky, I’ll get 17M too.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.