Fellow Ham's QCX+ 5W CW Transceiver - 40m

Repair, Testing, Troubleshooting, and Assembly

QRP Labs QCX+ 5W CW Transceiver Kit, 40m -- Repair, Testing, Alignment, and Assembly

Background

Bob, KB0**^[1], contacted me in mid-March, 2021 with an interesting project – could I repair his QCX+ kit? Bob and I had talked some since having the original QCX kit along with having a pretty fair background as an Electronic Tech and Engineer^[2]. I stopped by Bob’s place later that day to see what the challenge would be. Bob explained the kit that I was looking at was really the second one. He further explained the first one was partially destroyed trying to remove some incorrectly installed components. Silently, I thought this second one was probably not looking to be much better. I asked him what parts were left over, or from the first kit. He mentioned that he threw out the first kit’s main board, but still had some of the parts. At least there was something to start with in hopefully resurrecting the kit.

After getting it all home and taking a more detailed look at the kit such as what had already been done along with what parts were available, the following initial assessments were made:

The Main Board and the (assembled) Display Board were seen to have quite a few issues such as bad soldering, blobs of solder in many places, quite a few printed circuit board (PCB) pads seeming to be “burned off”, missing components, and wrong components installed.
There were quite a few components in the box of parts. At this point, I was not sure of having enough of everything or what may be needed for any repairs, but it appeared that there was a sufficient amount “just in case” something was needed.
A mostly unassembled Display Board along with what seemed to be the right parts being in the parts box to assemble a Display Board.

After getting Bob’s kit and seeing the challenge ahead, I really thought of just ordering a new kit (for USD$55), assembling it, and then “gifting it” to Bob. I like Bob – he is just a great person to know. The decision was to hopefully resolve the issues being seen.

The kit’s repair process started around the third week of March, 2021 and successfully completed on 18MAY21. The duration between the start and the end was due to getting a few parts in, ordering a desoldering gun (Hakko 301 as described below), two of our four single adult children visiting my wife and I over two separate weekends, and a bit of travel that my wife and I did.

^[1] The name and call of “Bob” and “KB0**” are fictional and being used to protect the actual person’s identity.

^[2] Many years ago, 45 is more approximate (between 1976 to 1978), I was a Technician at the GE Medical Circuit Board Plant in New Berlin, WI working full-time on the midnight shift (from 11pm to 7am Sunday night to Friday morning) while concurrently enrolled at the Milwaukee School of Engineering in their Electrical Engineering Technology program (taking 12-16 hours per quarter). It was tough going to school and working. I gained a great deal of experience resolving new boards that had failed for some unknown reason(s) from a day’s production runs. Several other Techs and myself fixed those failed boards getting them back into their production lots. After a while, I typically wound-up getting the new ones what I silently termed “the b!+ch boards” at first due to their complexity and destined for assembly into the (then) new Computerized Tomography (CT) units. The problems encountered ranged with bad solder from the wave solder process (seemed to be a recurring issue when coming in on Monday evening getting boards from that day’s first run or two after the wave solder being fired-up from being off all weekend), components not right or missing, along with a just a ton of weird and convoluted issues. The end result was getting just about any failed board fixed.

Prior to GE and returning to college, I was a pretty good ECM Tech fixing airborne ECM receivers to the component level until getting out the USAF and going back to school (I was drafted in SEP72 while in college enlisting in the USAF several days later).

I also “touched up”, repaired, redid circuit board traces, and/or “rewired” my prototype digital, analog, and/or RF PCBs or module designs as an Engineer in Receiver Systems Development at E-Systems, Inc., Garland Division in Garland, TX from 1983 to 1989. The prototypes were design proofs prior to finalizing the electrical and mechanical designs for release to manufacturing.

Repair - Display Board

The first thing tackled was assessing the assembled Display Board finding the following:

Soldering – blobs of solder
Burnt-off pads and traces leading to pads. The ones that appeared to be missing and burnt-off were the pads going to the two (2) 2x3 female headers (JP2 and JP3).
Incorrectly installed parts (i.e.: switches tilted (S2, S3), 1x16 male header installed backwards, and 2xnn female headers installed on wrong side of the board (as mentioned above))
- Uninstalled part (R46)
- This was determined to be the Display Board from the second kit that was ordered (replacing the first one)

The second display board was in pretty good condition although noticing the following:

Incorrectly installed parts (i.e.: 2xnn female headers installed on wrong side of the board). What appears to have happened was the single female 2x5 and the two 2x3 headers being installed on the wrong side of the board (they were soldered-in facing the front when they should have faced the rear). My guess is they were noticed as not having been installed correctly, then attempted for unsoldering along with removing the headers’ housings. Since having a solution for the female headers, the remaining housing was removed and then the connection points desoldered.
I figured that assembling the Display Board from the mostly bare display board PCB, display, and parts was the better option than trying to repair burnt-off pads located just slightly under the soldered-in display with the assembled Display Board as received. So, I assembled the Display Board using the parts in the box along with a few from my parts (i.e.: the female headers).

The original Display Board (left, showing the 1x16 header installed backwards) and the newly assembled Display Board (right). The newly assembled Display Board (right) also used cardboard spacers added under each of the nylon nuts (between the display and the PCB) to get the board-to-board spacing to around 3.0mm.

Repair - Main Board

Wow, where does one start with listing the variety of issues? As with the Display Board (above), the Main Board’s PCB had some similar issues along with some unique ones:

Soldering – blobs of solder all over the place
Burnt-off pads and traces leading to pads.
Wrong parts installed. This was mostly wrong value resistors having been installed. Most were located after doing a mostly complete visual inspection comparing the value found in a location on the board to that what it should be per the schematic and the parts listing. Admittedly, I did manage to miss a few, ones that would be found later when troubleshooting a stage that was not appearing to working correctly and noted later when located/discovered.
Incorrectly installed parts
Uninstalled parts
Toroids – T1 and Low Pass Filter

After having a good inspection of the component and solder sides of the mostly assembled PCB, I decided to completely desolder the board for two reasons: too much solder just about everywhere, and I needed to ensure there were no other burnt-off pads, traces leading to pads, and especially pads that may have a problem hiding under the globs of solder. Plus, I wanted to ensure that good solder was being used (since unsure of the current solder’s pedigree).

I was definitely not going to use “solder wick” or a manual trigger-type “solder sucker” to remove almost all of the solder on the main board. Doing either of those methods would have ensured a sure death of the board destroying more pads using a lot of heat and longer heating times trying to get solder removed. So, there was a delay for about a week while ordering and then receiving a new Hakko FR-301 Desoldering Tool (wow, have been wanting one of these for many years never justifying the price). Getting the Hakko FR-301 made the desoldering task near effortless getting the board cleaned-up very quickly (about 15-20 minutes from start to finish) and no damage – definitely the tool to use! It has been used more and more ever since receiving it!

Burnt-off PCB pads/traces. There were numerous areas with burnt-off pads found on the Main Board. Most of the burnt-off pads were found on the solder side although a few were found on the component side. One of the many areas worked was in the JP11 area under R30, R27.1, and R27.2 as seen in the before and after photos.

RF BNC Connector. The RF BNC connector was seen not sitting flush to PCB. Desoldered, repositioned, and resoldered the connector. This would have been an issue when trying to install into the enclosure later.

Toroids. The toroids in general were a "mixed bag" of issues.

L4 was seen to have 17 turns when it should have had 16 turns (as seen in the near right photo). There was also a small kink about 6 turns in (deemed ok at the time). Very unsure about the leads’ enamel removal.

Made decision to:

deinstall and rewind L4
deinstall all toroids to verify each one’s turns count and especially for ensuring each coil’s leads had the enamel adequately removed.

L4 after rewinding for 16 taught and unkinked turns. Measured-in at 1.089 mH (as shown in the far right photo). Good enough!

The other toroids also had issues such as wrong turns (with one needing to be rewound as not enough turns), one wound wrong, and each toroid's leads needing the enamal removed for correct soldering. Each of the toroids (i.e.: L1, L2, and L3) after being checked out or redone, and the enamal removed from thier leads, was found to be "close" to the assembly manual's nominal value.

Even T1 was desoldered and the leads redone for enamal removal. I had wound T1 for Bob some weeks prior, but Bob had installed it.

Testing and Troubleshooting

First “power on” commencing the testing phase was on 20APR21. The QRP Labs Dummy Load was connected along with an earphone. All seemed to be good when initially powering up such as with the receive current draw. The first thing noticed though was nothing being heard in the earphones – not even a background hiss regardless of the volume setting. Something, or somethings, were amiss! More discovery was ahead!

JP7 and JP8 signals when starting the troubleshooting (outputs from IC5A (IC5.1) and IC5B (IC5.7) respectively) showing the I (channel 1, yellow) and Q (channel 2, blue) channels.

Found several issues in the I and Q channels including a missing capacitor and wrong value resistor. Should have caught both of these during the initial visual.

Troubleshooting the audio stages at JP11, IC8B (IC8.7), IC8A (IC8.1), IC9B (IC9.7), IC9A (IC9.1), IC10B (IC10.7) and IC10A (IC10.1).

Was having issues with accomplishing the “Adjustment of I-Q balance” procedure with the expected results and attempting the “Adjustment of 90-degrees audio phase shift” section as covered in the assembly instructions’ paragraph “3.74 Adjustment and alignment” was a total disaster.

The initial alignment was being done as covered in the assembly instructions’ paragraph “3.74 Adjustment and alignment”. Just was not seeing at JP11 what was thought to be seen along with not getting the expected results of the “Adjustment of I-Q balance” procedure. After a bit of troubleshooting, found that C12 had not been installed (in the Q signal’s IC6A stage). The expected signal was now there at JP12, and for being processed during the self-test what was believed to be using the “AUDIO2” signal.

A further issue was then discovered with the IC9A audio stage finding R35 being a 1 K ohm resistor when it should have been a 750 K ohm resistor. Now, the right signal was believed to being seen at JP12.

Note: Somehow, my visual inspection when starting the project should have picked-up on the wrong or missing parts on the board. I should have caught C12 missing along with R35 being the wrong value. Somehow, these two parts were missed from being caught. Had I caught them during the visual, then that would have prevented me from having to troubleshoot and figuring-out “why” things were not right at that time.

Was having an issue with not seeing a signal after the volume control – there was a good audio signal at JP11, though the filtering stages (IC8B (IC8.7), IC8A (IC8.1), and IC9B (IC9.7)), out of IC9A (IC9.1), through C21, to JP12, and to the Display Board’s mounted Volume control (R1.1). The problem was nothing going forward to the next audio stage (the input of IC10B) seeing nothing at C22. What was weird was seeing a miniscule signal increase and then decrease slightly changes in the volume control’s position. With some troubleshooting of the signal’s path and with the Display Board disconnected from the Main Board, I discovered a direct short from the JP5.3 pad (electrically connected to the Volume control’s, R1, wiper (R1.2) to ground using an DMM. A quick breakdown of the test configuration ensued. With a bit of looking with some magnification, a portion of trace was now being seen going from the JP5.3 pad (which mechanically and electrically connects with the Display Board’s JP3.4 with a trace running to the Volume Control’s, R1, wiper (R1.2)) to the Main Board’s ground plane. That little bit of trace should not be there! So, the JP5.3 pad was desoldered, and with an Exacto knife, that little bit of trace was removed followed by resoldering the connection. A quick check of things was done followed by reassembling the test setup. Wow, not the receive chain was finally working right! A nice audio signal was finally there with the volume control working as it should.

With things appearing correct now all the way through the I and Q and the audio stages, the alignment process was run through once again and finally completed. The “Peak BPF” was checked followed by performing the “Adjustment of I-Q balance” procedure and the “Adjustment of 90-degrees audio phase shift” procedures as covered in the assembly instructions’ paragraph “3.74 Adjustment and alignment”.

The last steps in the alignment were also successfully completed too since having a good signal all the way through the I and Q and the audio stages for the “AUDIO1” signal being processed during the “Adjustment of 90-degrees audio phase shift” section as covered in the assembly instructions’ paragraph “3.74 Adjustment and alignment”.

Using my QRP Labs QLG1 GPS Receiver, Bob's QCX+ was “calibrated”. It did not seem to be “off” from 7.020 MHz very much, but performing the calibration with the QCX+ unit’s menu 8.11 “Cal ref osc” and 8.12 “Cal sys osc” calibration procedures got it “spot on”.

The 412 mVpp seen from the o'scope catpure was the initial RF output. This paltry signal output works out to be about 3.5 mW which is nowhere near the 4-5 W expected. There seemed to be something amiss with the transmit side. There was plenty of drive out of IC3A (IC3.3).

Upon closer inspection around the “finals” and supporting circuitry, R42 was found to be a 120 K ohm resistor (as shown above) instead of being 1.2 K ohm. Somehow, I missed the value difference when doing a visual compare of the schematic and parts list when first starting the project. The 120 K ohm resistor was deinstalled and replaced with the correct 1.2 K ohm resistor from the extra parts.

Amazing what changing out R42 from the wrongly installed 120 KW resistor with the correct 1.2 K ohm resistor. The output shot up from 412 mVpp to 41.2 Vpp after getting R42 exchanged. That 41.2 Vpp output voltage correlates with being right around 4.25 W.

Test setup showing around 4.3 W on the little Welz power meter (that I use when outside…have had the unit for probably 30+ years and still works great for quick checks) and at 41 Vpp on the o’scope.

Assembly

Several issues were found fitting the assembled Display and Main boards into the enclosure. The first was the board-mounted 1/8-inch (3.5mm) phone jacks were each cocked slightly where they would line-up to enter the rear panels holes (as shown below). I also had to modify a nylon M3 nut by cutting off one side so that it fit next to the Volume control, R1, when the Display Board was fit to the enclosure’s front panel (not shown).

The pieces and parts ready for assembly after the checking out the QCX+ on the bench.

The QCX+’s Main Board showing the 1/8-inch (3.5mm) board-mounted jacks not lining-up for inserting through the enclosure’s rear panel holes (rear panel view). The jacks were then seen to be visibly not perpendicular or “cocked” to the right side of the Main Board (left side of the picture from this rear panel view).

The QCX+’s Main Board showing the 1/8 (3.5mm) board-mounted jacks after being desoldered, positioned correctly (now perpendicular to the Main Board), resoldered, and now aligning with being inserted in the enclosure’s rear panel holes.

Completion

All buttoned-up! Everything seemd to check-out. Bob's QCX+ 5W CWd Transceiver is ready for some use and fun on 40m!

QCX+_Kit_Repair_210524_HI.pdf

The QRP Labs QCX+ 5W CW Transceiver - 40 m Kit -- Repair , Testing, Troublshooting, and Assembly is a PDF document with a more detailed account along with more pictures of the issues encountered with the kit and thier resolutions. The document can be viewed online or downloaded.

Page updated

Google Sites

Report abuse