QCX+ sidetone key clicks and circuit noise SOLVED

Al Holt

I finished building this QCX+ (40m) yesterday and initial testing seemed like everything was OK. It was doing a pretty good job receiving over-the-air signals and audio tone was good. I was also getting a pretty good RF output and I put the unit into practice mode to try out an iambic key I have.

But, I noticed I had excess sidetone key clicks when sending characters; some very loud DC-like clicks. I also noticed while wearing headphones some very annoying circuit noise, sort of a sizzling noise along with some swooshing, cracks and pops. It sounded like a component going bad. This was all with a 50 Ohm dummy load attached and a known good power supply.

To jump to the chase, it turned out being a bad capacitor, C21, one of the two 1.0uF ceramic caps. I replaced it with a 0.1uF (along with C22, the other 1.0uF cap). I had these in stock, some leftovers from modifying my original QCX a couple of years ago.

It took some time to isolate C21 being the problem. I decided to cut the PC trace jumper, JP11, between the I/Q section and CW filter section just to rule out the front-end and detector. With a 'scope I was seeing noise at the output of the CW filter, JP12; I had no noise on the other side of C21, but it took me a while to finally decide all that noise was coming from just the capacitor! I decided to go ahead and also replace C22 as a precaution thinking those caps were probably from the same batch.

I redid the alignment with very good results. Sidetone key clicks are gone along with the circuit noise.

Whew and 73!




nice detective work.  if you didn't already throw away that capacitor, I would test it out of curiosity.  if you have anything to measure capacitance, see what it reads.  I tend to doubt handling the voltage would be an issue, if it lands around 1 uF.  If you don't want to know, that's okay. 

73 Curt



Thanks Al. I also finished my QCX+ (40m) the other day and appear to have the exact same issue as you. All seemed to work very well, good reception etc, but as I am learning Morse Code I put the radio into practice mode and for good measure also connected my QRP labs 20W dummy load. Through headphones I was surprised at the background noise but key tone sounded great... and then it suddenly didn't! The background noise suddenly stopped and the tone was now nasty clicks.

I discovered by turning the volume to max and keying that it would suddenly start working again!

The problem is not consistent. I can turn on the QCX+ and get good tone (with some background noise) while other times I can power on and immediately just get loud nasty clicks when I key. I know the keying works because the TX decoder displays the correct characters.

I don't have a scope so replacing C21 and C22 is a easy first thing to try.

Aside: This may be a coincidence, but the problem seemed to start after I lowered the supply voltage to 7V. From the experience of others, I changed R48 from 270R to 470R. This resistor controls the LCD brightness. Hans mentions he chose this value so the display would still be good in daylight at 7V and I was curious if it was still good with 470R (it is!)

Alan G4ZFQ

The problem is not consistent.

That makes me think bad soldered joint.

73 Alan G4ZFQ

Al Holt

Hello all and thanks for your comments!

I think the way sidetone is generated is one reason for the big thump IF C21 isn't doing it's job blocking DC. To begin, the CW filtering section is all directly coupled between stages and we have around 2.5 Volts at most pins. 

As I understand it, sidetone is a PWM square wave at a ~48 kHz coming from the microprocessor and injected into the CW filter section to be filtered to a pleasant tone. My guess is there's a fairly 'large' DC shift when sidetone is injected and IF C21 is leaky will not block enough of the shift and so the key click is generated. 

My troubleshooting focused more on eliminating the circuit (background) noise, which replacing C21 took care of. With the sidetone problem, I'm mainly fitting my view of its "theory of operation" as the reason it fixed the key clicks as well. 

Unfortunately, I pitched the original caps after removing them from the PC board and don't have them to test. During original construction I did not mount the caps flush to the board, and upon removal I cut their leads off very close to the body to save the traces during desoldering and removal. Part of my "save the board, not the component" philosophy :-)  

Ian, I also replaced R48, display brightness resistor, during construction. I chose 560R, and like yours, results are good!  

So far so good, I hope to get it on the air today!