The following was compiled by Ron WA7GIL based on his experience with fixing 60 QCX-series transceivers. Thanks Ron!
Compilation of QCX Issues Found and Fixed in 60 QCX radios of the three models [QCX, QCX+, QCX-mini]
In this compilation I will attempt to point out the most common issues I've found in fixing 60 QCX radios. An attempt will be made to list some of the more common issues first with the less common ones following later. I will show what I found in each case and also provide an idea of which QCX models were most affected by each issue if that is relevant. It is assumed the reader will have a copy of the schematic diagram and appropriate board layout as components will be referred to by component number … R36 … C21 … IC7 … etc. This is by no means an exhaustive study and your results may (will) vary but I hope to give the reader some good clues as to where to look for a few of the more common issues found in the QCX radios. The vast majority of kit builders will never need to even read this document as they successfully built and aligned their kits the first time. But some will hopefully find it useful.
Low or No RF Power Output
IQ Balance, Lo and Hi Phase null not adjusting
BPF Adjustment Issues
Display Showing One Row of Blocks
Si5351a Error message
Various Audio Issues
Intermittent issue with output power suddenly shifting from 4 watts down to 2 watts and back up to 4 watts.
Backwave from Transmitter for a few seconds after Keyup
Loud 700 Hz Oscillation in Receive Mode:
High Current on Receive
Sudden Frequency Shift of a Few Hundred Hz:
Buttons Don't Work
SUMMARY
Conclusion
Low or No RF Power Output
This was the most common issue/complaint found in 60 units. It occurred in approximately the same percentage in all three models of the QCX. Assuming this is your first power up and you haven't seen sparks or smelled any of the magic smoke, then various items can be checked easily by the builder:
- If you have low RF power output, follow the procedural video from Hans to peak it up to normal [see https://youtu.be/eN7wER05T-c]. This is also important in order to get good readings when doing the receiver Bandpass Filter alignment so it should be done first.
- If there is no power output then, with power off, use your ohm meter and check for continuity between the center pin of the BNC connector and the L3 end of C29. If this is no continuity between these two points, then go through each toroid in the lowpass filter until you find the one or ones that are not making contact with the board and re-solder them. As Hans has pointed out, improper soldering of the toroids, ie improper removal of the enamel wire coating, is the number one construction issue with these kits.
- While you have the ohm meter out, also make sure there isn't a short between the center pin and the shield of the BNC connector. If this is shorted, look for solder shorts in the lowpass filter. On a QCX Mini, look particularly in the area of the BNC connector itself. It is easy to short the center pin to ground on the board if too much solder is applied on this unit
Issues I've found when the complaint was low or no output power:
- Open circuit on output end of L1
- Q6 failed open. This can be easily checked with a DVM. You should see 12 volts on the emitter all the time and 11+ voles on the Collector when the key is pressed.
- Q6 failed shorted. This can be checked by looking at the collector of Q6. If there is 12 volts there on receive and transmit, this device is shorted. Less likely is the slight possibility that it is being held in the on state by Q4 or its associated circuitry, but I've yet to see this case.
- Q1, 2 and 3 shorted. This was a result of transmitting into an open circuit or very high SWR for an extended period of time and other issues resulted as well.
- L4 continuity issue due to poor enamel removal.
- Very low (almost no) power in one case was due to using a 39 pF capacitor instead of a 390 pF capacitor in the Lowpass filter.
- Bad IC1, IC3, Q1, Q2, Q3 and Q6 caused by suspected lightning strike
- One extra turn on L1,2,3&4. This is due to a common misunderstanding of how to count turns.
- One turn less than required on all toroids. Same reason as number 8.
- Bad C25 and 26 on early QCX Classic 20 meter models caused power drift. Corrected on later revisions.
- No CLK2 output from IC1. Usually caused by a solder short between pins 6 and 7 of IC1. No longer an issue with Rev2 Mini which is tested prior to shipping.
- Bad ground connection (pin 7) on IC3. User connection on models up through QCX+.
IQ Balance, Lo and Hi Phase null not adjusting
This was the second most common issue reported. It occurs across all the models for various reasons.
Users who report getting a good BPF adjustment but the phase adjustments can't be made because the volume can't be set to a level above “1” or the volume cannot be set lower than level “13”, usually have an issue with the volume pot being poorly soldered on the high side or wiper (low volume level) or the ground side (high volume level).
On QCX Classic models I found this volume pot issue was usually due to poor solder connection to ground.
On QCX+ models the issue on two of them was a damaged potentiometer. Bending the leads caused a disconnect of the bradded connection between the lead and the carbon trace on the pot. This was fixed by squeezing the brad with a thin pair of needle nose pliers.
On QCX Mini models the most common issue was improper soldering of the 2X4 header on top of the control board. The pins on this header do not protrude all the way through the control board. They need a lot of heat and good solder flow down into the plated-through holes with good adhesion to the pin inside the hole. Other issues on the control board can be caused by this and may be mentioned later in this article. One Mini had a bad volume pot and it had to be replaced. It was noisy and it is not repairable.
- On one unit, this issue was caused by a leaky C12 which was making no phase change possible at IC6A.
- An intermittent C1 on an early Classic model was causing issues with this as well as wth BPF peaking.
- IC4 was bad on a few units with this issue. This was usually caused by putting 112 volts on the 5 volt bus and caused other issues as well with failures of most or all of the 5 volt ICs.
- Open circuit trace between R17, C10 and Pin 5 of IC6. This was a classic model and was repaired with a short 3-way jumper.
- On the Rev 1 version of the QCX Mini, several different models had this issue and other issues due to bad op amps. This was a known problem and corrected with new parts and board tests with the release of Rev 2 of the Mini.
- C5 was installed on a 20 meter Mini. The 20 meter and higher units do not use this part. Bandpass filter could not be aligned and test signals were also so weak that the phase and balance adjustments weren't possible either.
- IC10 had the wrong bias voltage. This was caused by a cut lead on one of the bias resistors (R39 & 40).
BPF Adjustment Issues
- On a 40 meter unit, found the 27 MHz and 20 MHz were swapped in position. Because of this, when the radio was set to 7.2 MHz it was actually tuned to 5.02 MHz. That was too far out of the tuning range of the BPF to get it properly tuned. 5.2/7.02 = 20/27. This mathematical relationship is what caused me to take a closer look at the crystals.
- An internal intermittent connection in C1 (mentioned earlier) caused erratic tuning and lack of ability to get the BPF tuning in range or stable. Replaced C1. This was an older Classic model before the ceramic capacitors were provided for C1.
- Found a secondary tuning peak of level 5-7 as well as a primary tuning peak of 8-9 on a couple of units. Both peaks were in the tuning range of C1 (not max or min). Set to higher peak.
- Poorly soldered ground on C1 caused intermittent de-tuning when the capacitor was tapped. Re-soldered. Ground connections on the board are the most commonly mis-soldered ones.
- Found an open connection at one end of L2 in the lowpass filter. Remember that the test signal also passes through the lowpass filter so both receive and transmit as well as test signals will all be affected by issues with this filter circuit.
Builder Tips: Aways use extra heat and time on ground connections. These take more heat because of the heat flow into the circuit board ground plane. Always do the power output test and adjust/repair the lowpass filter circuit before moving to the receiver alignment steps. Note: this is also the order of events in the excellent tuneup video provided by Hans on the website [see https://youtu.be/eN7wER05T-c].
Display Showing One Row of Blocks
- 5 volt bus may be shorted by a bad component, like IC1,2,3 or 4. If high current is being drawn on receive check for low resistance fault on both the input and output of the 5 volt regulator. Also check voltage on the output of the regulator. If it is low and/or if there is a low resistance from the output of the regulator to ground, it follows to isolate the 5 volt ICs one at a time till the offending unit is found. For IC1, simply lift one end of either D1 or D2. For IC2 (the microcontroller) just remove it from the socket. For IC3 heat up pin 14 and lift that pin clear of the board connection. Same with IC4, except lift pin 16. If no change in current is seen with one step, reinstall that component or connection and move to the next one. If a change is seen mark that component for replacement and continue to test. In many cases, the 5 volt short could have been caused by the useer accidentally shorting the 12 volt bus to the 5 volt bus while testing voltages. In that case, all 4 ICs could have been damaged.
- IC2 may not be getting power. Check for 5 volts on pins 7 and 20 of IC2. If either is missing check for open condition in L6 and L5 respectively. This is easily done by checking for 5 volts on both sides of each inductor.
- IC2 may be bad. Not highly likely but it does happen, especially when accidentally powered at 12 volts.
- A few cases of this were found to be caused by poor solder joints in the area of IC2 and the LCD display. In the case of the QCX+ and Mini versions, pay close attention to solder connections at the headers and pins that join the display boards to the main boards.
- One case in a Rev1 Mini involved an intermittent condition of the single row of blocks. In that case the 5 volt regulator was found to be putting out voltage spikes on turn-up. Regulator was replaced with the better 78M05 and the problem did not recur.
- Bad LCD modules were found in 3 cases, one on QCX+ and 2 on QCX mini. If you have access to a spare LCD module or display board try a swap out to verify this issue before removing the display, just to make sure. This involves a lot of de-soldering on the QCX+ and Mini, so it's good to know ahead of time if this is the issue.
Si5351a Error message
- No communication between IC1 and IC2. In one case, pin 28 (SCL) of IC2 was bent over and not in the socket.
- Found many cases of a bad IC1 with this message. Most of those were due to accidentally putting 12 volts on the 5 volt bus.
- Solder short between pins 6&7 if IC1. Hans mentioned this in one of his articles. This was found in a couple of cases.., from the factory … and it no longer occurs with Hans' factory testing of the boards before shipment [see https://youtu.be/xb_ZmPx7Py4]. The fix involved using solder wick and liquid flux to remove the offending solder blob. Note: sometimes these are not even visible under a USB microscope, but you can find them by measuring resistance between adjacent components in the circuit which are connected to these pins. For this particular case, you can measure between the cathode of D1 and pin 4 or 10 of IC3. In one extreme case I found a short between pins 4 and 5 (SDA and SCL) and an open connection on pin 8 (ground) of IC1. This is definitely not common and was only found on one in 60 units and it was before the aforementioned testing was begun.
- In one case I found a shorted TCXO which caused D1 or D2 to open up with high current. I removed the TCXO and temporarily replaced it with a crystal and IC1 came alive with no error message. I inspected the TCXO module and found nothing obvious. I then connected it independently of the QCX and it worked correctly. Upon closer inspection it was found that the builder had mounted the TCXO board all the way down against the surface of the main board and used excess solder. That and a whisker of wire allowed the DC input to the TCXO module to be shorted to ground.
Builder Tips: If you use the built-in DVM function of the QCX, be sure your test lead is connected to the DVM input (pin 1) and not the 12 volt pin (pin 3). This has caused a few problems for some when the test lead itself then put 12 volts on the 5 volt bus as soon as it was touched to any point in the 5 volt system. Always be careful when measuring voltages that your test leads (even on external DVMs) do not touch adjacent components. On the positive probe of one of my DVMs I sharpened the tip and insulated all but that very tip with heat shrink tubing to avoid adjacent component shorts. Even with that I have to be careful when measuring values in the more crowded Mini.
Don't push the TCXO board all the way down to the main board. It is usually OK to do this as the bottom of the little module has no copper on it, but it really makes it difficult to remove when replacing IC1 and it could hide an issue like the one mentioned above.
Various Audio Issues
This is a more broad topic and encompasses a lot of the earlier issues since most of the receiver is operating at audio frequencies. Some repetitions of earlier issues may be recognized.
- Shortly after power up on one unit a rumbling sound came into the audio, which got louder with time. I found that C22 was leaky. This was found by watching the bias voltages on IC10 and noticing that they started out somewhat normal but gradually rose to much higher levels.
- IC8 was defective in one Classic QCX. This was the only bad IC I ever encountered in a Classic model, of which were more than half the total QCX models I've worked on. It was causing a loud oscillation in the receive audio that affected alignment and operation of the rig.
- Harsh sidetone while transmitting with a QCX+. Found bias on IC10 was wrong. Fixed that by repairing bad solder joint on R39. Still had harsh sidetone. Discovered D5 was installed backwards and receiver was not muting properly.
- On one unit, the receive audio did not immediately return on key up transition from Transmit back to Receive. It would eventually come back. Q7 was staying triggered but the signal from the microcontroller was off. Removed C52 and operation was normal. I thought then that maybe Q7 was bad but replacement device didn't fix it, so left unit with no C52 in place. This was not an ideal fix but I was not able to find the true cause.
- Rev. 1 Mini... no audio on initial power up. Found defective IC10. Replaced IC10 and unit worked normally. This was discussed earlier. Several issues of no audio or oscillations in the audio on Rev. 1 Minis were solved by replacing various defective op amps. In one case there were two bad op amps in the same unit. As discussed earlier, this issue was resolved in Rev. 2 Minis with better parts and assembly house and factory board testing before shipment.
- Noisy audio in a Mini. Found bad solder joint(s) on the 2X4 header on the top of the controls board. These need to be soldered with plenty of heat and good solder flow as discussed earlier. Several Minis were found to have this and related issues with controls due to the soldering of this one connector.
- Noise in receiver while transmitting … harsh sidetone … loud racket … Various descriptions of the same issue. Noticed especially on the Mini, it appears that RF can get into the audio circuits via the headphone jack, especially when using antennas very close to the rig, like the popular end fed half wave types used in field operations. One fix is to move away from the antenna, add chokes in the coax line or go to a more balanced antenna like a dipole. Another solution is to add a 10nF capacitor across the audio output jack on the bottom of the board. In one case I found that IC10 seemed to be a little more prone to RF than most. I replaced it with the higher current 4562 type component and resolved the issue in that particular unit for that particular antenna arrangement.
- Scratchy audio. Found bad volume potentiometer in Mini. This one is not repairable and was replaced to solve the issue.
Builder Tips:
No need to immediately suspect op amps on anything but the Revision 1 version of the QCX Mini. Even then it's always best to do all the bias voltage checks and check resistor values around the op amps as well as continuity between circuit points and any shorts before thinking about replacing op amps.
Consider adding a 10 nF capacitor across the audio output jack so that use of a compromise antenna in field conditions won't cause sidetone issues.
Make sure all header pins are properly soldered, especially those that do not protrude from the top of the controls board in the QCX Mini.
Intermittent issue with output power suddenly shifting from 4 watts down to 2 watts and back up to 4 watts.
- All the usual things were checked like solder joints, components in the output circuit and nothing was found. It actually appeared like it was shifting in and out of Class E operation mode. I checked C30 and it was good so added 2 turns to L4 and the problem was resolved.
- Same case as number 1, but this time C30 was bad. Replaced C30 and the problem was resolved.
Backwave from Transmitter for a few seconds after Keyup
- A low level, on frequency, signal was heard on a local receiver for a few seconds after transmitting. This is not to be confused with the local oscillator signal that is 700 Hz away from the on frequency signal all the time during receive. This problem was resolved by performing a factory reset.
Loud 700 Hz Oscillation in Receive Mode:
- In one case I found missing capacitors at C14,16,18,23 and 33. These were all the 1 nF capacitors in the circuit and simply indicative of missing one of the construction steps. Other issues were present besides the loud oscillation but all were resolved by installing these components.
- Bad IC8 in Three Rev.1 Minis. No longer an issue. See previous discussions.
Builder Tips:
Make sure you have done all the construction steps with your kit. Maybe write down the step numbers as you complete them. In the days of written manuals, like Heathkit (for example) there was a check box to mark as each step was completed. That's not possible on an on-line manual but it would be easy enough to just write down the step number if that is an issue.
High Current on Receive:
- In addition to the 5 volt component shorts mentioned above, there could also be shorts on the 12 volt bus. These are usually found in the area of Q1,2,3 and/or 6 and usually after a high SWR long transmission issue, like WSPR on a bad antenna. Check and replace bad components.
- Check the input to the voltage regulator as well as the electrolytic capacitors in the area if output components aren't causing the issue. Even the 5 volt regulator cam be bad, but less likely.
Sudden Frequency Shift of a Few Hundred Hz:
- 1. In one case out of 60 units I found a bad 27 MHz crystal causing this issue
Buttons Don't Work:
- On two Minis I found an open circuit board trace between IC2 pin 26 and the “buttons” pin of the 2X4 header, on the bottom of the main board. These were replaced with a short jumper to resolve the issue.
- A few cases of the same issue involved poor soldering of the header pins to the top of the control board as discussed earlier.
- Right button didn't work in one case. That was found to be R45 not being installed by the builder.
SUMMARY
Note: The quantities of issues in this summary may not correspond exactly with the totals discussed above, as similar issues may have only been discussed once. These are totals out of 60 units to give the reader an idea of the frequency of occurrence of each issue.
- Open board traces on 6 units of all models. Half of these caused by the builder.
- Volume pot issues on 6 units of all models. 3 bad pots. 3 poor soldering issues
- Bad IC4 in 5 units of all types. Usually caused by operational issues
- Bad IC2 in one unit.
- Bad IC1 in 7 units of all models. Most caused by putting 12v on the 5v line.
- Bad IC3 in a few cases. Only one discussed herein. Usually operational or user testing issues.
- Major soldering issues in at least 11 units. Several minor soldering issues not discussed.
- Bad operational Amplifiers in 10 units (1 in a QCX Classic and 9 in Rev.1 Minis)
- Most of the major problems in damaged transmitters were caused by running WSPR into high SWR conditions.
- Most low power output issues caused by not properly adjusting lowpass filter toroids.
- Most no power output issues caused by soldering, usually by inadequate removal of wire enamel coating. Other no power output issues caused by running WSPR into high SWR conditions.
Conclusion:
This is by no means an exhaustive discussion of issues/solutions for the QCX line of QRP radio kits. It was only based of repairs I made on 60 units total from the three different variants of the radio. It is hoped that the reader will find something close enough to the issue they are investigating in order to complete their own repairs. If not, then hopefully the various trains of thought will give the reader some clues as to where to look next.
Best 73's to all and good luck with your fault-finding and repairs (when needed) for these amazing QRP radios that usually don't require anything but building, aligning, and using.
Ron Taylor WA7GIL