QCX 5W CW Transceiver - 40m

QRP Labs QCX 40m 5W CW Transceiver

A QRP Labs QCX 40m 5W CW Transceiver kit and an QRP Labs enclosure Kit was started around MAR18 getting halfway through the assembly when having to packing things for moving to another temporary house in MAY18. Somehow, the kit was not completed since being consumed with looking for a new residence with my wife, Denice, and working (my wife and I sold our northern VA house and were looking for another house in the Colorado Springs area living in two rental houses during the 5-6 months). Then with still working at that time, Denice's health for too many months, and getting into a new place with all of the unpacking, nothing happened with getting it out and restarting the kit's assmebly. It was finaly upacked, on the bench, and most of the remaining assembly and testing done. The kit was finally completed in JUN21.

In-progress with the initial assembly. This MAR18 picture was a bit after starting the kit. did not get too much further with the assembly until packing-up everything for moving from one temporary house to another.

Tyring to figure-out what I was thinking of for controls and connectors mounting and the wiring (for the controls and connectors on the enclosure's end panels and how they will connect with the PCB). Still have yet to wind and install the toroids. Picture from DEC20.

Further along in working-out the details for the enclosure's mounting and for the wiring. The toroids have been wound along with being installed.

All done with the mounting and wiring. Unknown to me at the time, the initial "power on" testing would be quite a long time in the future. Still have to install the PIC....

Testing the QRP Labs QLG1 GPS receiver after completing the assembly. Had a quick and good lock on GPS satellites from right there on the bench.


I hate to admit it, but I was a bit “off” when initially constructing the QCX with the thoughts of putting the the QCX into a case. While the display and orientation was mechanically right, it was seen to be orientated upside down. That was noticed when first putting power to the unit and trying to get the display level right. My mistake was immediately noticed. So, the nice header fabricated to take the signals from the main board to the display board was removed along with gently deinstalling the female headers on the display and main board. The display board was orientated correctly on the front panel too. Discrete wires were soldered in-place from the main board to the display board. A quick power-on check confirmed that the display was now correct.

Showing the headers taking the signals from the main board to the display board. The problem was that it was not right with the display orientated now being upside down. The display being upside down not being noticed until that first power-up.

Note: The board assembly at he bottom of the picture is the QRP Labs 20W Dummy Load

Looks good, but not right. I also did not have any power for the display's illimination.

Discrete wires taking the data from the main board to the display board. The display board is also mounted correctly in the front panel.


Performed the alignment and an attempt with the frequency calibration – Performing menu 8.11, Cal Ref Osc, seems to persistently “hang” at “4” in the countdown – will defer messing with the calibration to later. All seemed to be good though. Received signals seem to be about 120 Hz high when compared to the TS-990S.

Needed to update the envelope shaping circuit components with those seen in the QCX Rev 4 revision and correlating with those in the QCX+ kit -- R41, R42, and C31. Desoldered and removed those components. Found that I had all but the 1.2 K ohm 1% metal film resistor (the one metal film 1% resistor kit on-hand did not have the 1.2K ohm value, or a lot of other values too). So, ordered a new quite complete 1% metal film resistor kit.


The new 1% metal film resistor kit was received. Replaced the driver components R41, R42, and C31 with the ones noted in the QCX Rev 4 revision and correlating with those in the QCX+.


Finally, was able to get “fired up”. Was seeing to only be getting 38.383 Vpp which is about 3.7-3.8 W out. Was hoping to get a bit more power out than that, an output that would be close to the 4.3 W out when working on the QCX+ kit.

Wondering why or how to get a bit more power out. The one thought was that maybe the LPF was not quite aligned or “on” with the 7.05 MHz frequency area.

There was the need to verify the QCX’s output low pass filter (LPF) to ensure that it was not attenuating too quick for the output signal. Therefore, I need to characterize the QCX LPF using the components already installed (as per the section of the QCX schematic below). So, C28, L1, C25, L2, C26, L3, C27, C29, and C30 were deinstalled. They will then be measured prior to further testing as the QCX’s LPF. L4 will not be “messed with” for now unless there is a reason to look further into it and the circuitry surrounding it.

I needed an RF test board with SMA connectors along with some discrete “pads” where I could tack-solder components. The specific thought for this “fixture” was for checking out multi-element filters such as the QCX’s LPF.

The small RF test board was made from a small piece of single-sided PCB that I had on-hand, a piece about 25 x 60 mm. A razor saw was used for separating the foil in half lengthwise, and then taking one of the two halves to further make 7 distinct “pads”. The end “pads” were then further reduced to make a “strip” for the center conductor of a SMA PCB connector to solder to. A DMM was used to verify the cuts were complete – that no two adjacent sections of foil remained electrically connected. The fixture was then tinned along with the connectors being soldered on.

Of particular interest were the toroids L1, L2, and L3. L1 and L3 were found to have 21 turns as they should have had and L2 had 24 turns as it was supposed to have. The following was the nominal (as per the QCX’s assembly instructions) and measured values for the toroids:

Nominal Measured

L1 1.4 mH 1.507 mH

L2 1.7 mH 1.860 mH

L3 1.4 mH 1.531 mH

The toroids were considered as being “close enough” for their measured values when compared to their nominal values.

The QCX’s LPF components (i.e.: C28, L1, C25, L2, C26, L3, C27, C29, and C30) were tack-soldered to the test board replicating the schematic. This was a little “backwards” for me with the LPF’s input on the right from the SpecAn’s Tracking Generator (TG) Output and with the LPF’s output on the left going to the SpecAn’s Input.

The QCX’s LPF appeared to have a 0.32 dB insertion loss. In the screenshot, the SpecAn’s frequency is set for 15.0 MHz, the span set for 20 MHz, and the TG set for -20 dBm. The attenuation at 14.1 MHz was over 40 dB. Somehow, I neglected to get screenshots of the 14.1 MHz marker and the 3dB bandwidth (BW).

The result was the LPF seemed to be working pretty good.

All of the LPF filter components were deinstalled from the test board, cleaned up, leads straightened if needed, and then reinstalled in the QCX board.

A quick test of the QCX assembly was done as a “sanity test” after getting it back together. I was still unsure “just why” not getting a bit more signal output from the QCX. I was only getting around 38.7 Vpp which is a bit lower than the ~41 Vpp when working on the QCX+ kit. Other than that, I am pretty happy with the way things are for now.

The frequency calibration was attempted once again using the QRP Labs QLG1 GPS module. Was confounded by it when starting menu item 8.11, Cal Ref Osc, seeing the sequence of “Waiting, 6, 5, 4….” And then seeming to be stuck on “4”. Found in the “groups.io” QRP Labs group that I should have had the unit set to practice mode with menu item 4.7. Also messed with menu 6.1 too a bit. Once that was done, it was switched back over to menu item 8.11, Cal Ref Osc. After playing with it a few times, I finally was able to see the entire sequence complete for the first time. I then switched over to menu item 8.12, Cal Sys Osc, which completed quickly.

In some comparison testing, a CW QSO being received on the QCX was checked against what was being heard on the Kenwood TS-990S – the QCX seemed to be “right in there” being on-frequency along with tone and received signal strength for what was being heard and seen with the TS-990S. The QCX was connected to my 80m NVIS antenna and with the TS-990S connected to the Zero-Five 10-40m Ground Plane Vertical antenna.

Just prior to putting the cover on…

…and with the cover on and ready for use!