If you’ve been keeping up on the MiscDotGeek YouTube channel or the QRP Labs Groups.io forum, then you’re likely aware that the QCX Mini has had some teething issues. We’ve already covered the C38 problem, and that’s been fully resolved. Instead we’re going to look at the issue that some builders are having: The inability to align the receiver. First we’ll put it in perspective, then we’ll dive into the troubleshooting.
Before you read further
The purpose of this article is to document a troubleshooting procedure. Troubleshooting is an important skill that many lack. Sometimes, the things you think are troubleshooting, well… aren’t. This article serves to highlight a proper troubleshooting procedure for a very specific problem. At some point if there is interest I’ll do a video/article about how to troubleshoot in general. To me, it is a fascinating subject.
If you have an unbuilt QCX Mini, do not be deterred! Build on! And if you DO have a problem, be sure to read all the way to the end.
And One More Thing: This is in no way designed to replace Hans Summer’s official QCX Troubleshooting video, which is the basis for everything I learned about this subject. Furthermore, looking at the DC voltages is often enough to determine which IC is bad. In my case it was not and so it is the subsequent troubleshooting that I am focusing on.
All that being said, there are many QCX Mini’s that never experienced a problem. Let’s put this into…
Perspective
The QCX Mini’s first run has had some trouble. Take into consideration though that some issues are to be expected in the first run of almost any product. Additionally, 997 kits were sold, and a huge amount of people are building right now. It is relatively uncommon for people to say “It worked!” Most will download the directions, build it, and say nothing. This is fine, of course- it’s the expected action. The vast majority of people who post however are going to be those that are having problems. After all, that’s what community driven technical support groups are for.
Furthermore, don’t let this first batch dissuade you from purchasing a QCX Mini from the second batch (which I will definitely be doing). Hans has said that no matter the percentage of people having problems, it’s too many and he’s working to prevent such issues in subsequent batches of QCX Mini’s. He’s taking a two-pronged approach: Beat up his suppliers who have caused so much grief, and stringent QA testing of each and every board that gets shipped in all further batches. In fact, he FIRED the fab house that built the first set of boards!
Troubleshooting QCX Mini Alignment Problems
Now we’ll dive into troubleshooting alignment problems with the QCX mini. To troubleshoot, you first need to understand and be able to reproduce the symptoms. Then you need to understand how the QCX receiver works so that you can take a systematic approach to resolving the problem. The documentation is the best place to go for a description, but I’ll summarize it.
Symptoms
After QCX Mini assembly is completed, the instructions call for going to menu 8.8, 8.9, and 8.10 to minimize the signal strength at each step by turning the trim potentiometers R27, R17, and R24. These work to cancel out the opposite sideband. Or in some cases, they don’t.
The primary symptom is that sideband cancellation is impossible. A secondary symptom is that when nulling out the signals (or trying, anyway) the volume potentiometer goes from about ‘5’ to ’13’ in well under 1/4 of a turn. It’s supposed to take most of the travel to do this.
Other symptoms might be that the audio isn’t getting to the output, so nothing is heard in headphones. That is a different problem that could be related to op-amps, and so the same type of troubleshooting techniques will apply.
Understanding Phasing Cancellation
The symptoms tell us that there is no phase cancellation happening. Essentially, the receiver works in a way that eliminates half of the signal (the sideband we don’t want), and then amplifies what’s left into audio that you hear. For this reason, an untuned receiver overloads all of the audio amplification chain, because there is too much signal (both sidebands) going through the audio chain. So we have to troubleshoot to find out why the phasing sideband cancellation is not working.
Phasing sideband cancellation is something I’ve struggled to understand myself, but tracing the fault in my own QCX has made it far clearer. Let’s take a quick look at it. First, go grab a copy of the schematic and follow along:
http://qrp-labs.com/images/qcx/trouble/circuit.png
Follow the blue line in the schematic. The desired signal comes in from the antenna and is routed through to T1. Imagine that the whole signal is across both windings, and the center tap cuts it in half. Each half is a phase. These two phases are fed to the FST3253 which is what’s being used as the Tayloe Detector. The SI5351 chip also feeds in two phases of the desired frequency. The output of this circuit is the resulting audio, just as you’d have in any Direct Conversion receiver, but four times. Each output is the same audio, 90 degrees out of phase the one before it. 0 and 180 degrees are combined in the I channel for a dual sideband signal. The 90 and 270 degree phases are combined into a dual sideband signal and sent through the Q channel.
The next thing that happens is where the magic is. The two Dual SideBand (DSB) signals are maneuvered around in such a way that the unwanted sideband is cancelled out, thus killing off half the signal.
If none of this is working, then we need to troubleshoot to find out why. First we’re going to take an analytical approach to solving the problem, and then we’ll look at the “shotgun” approach. (Sorry, no firearms are involved!)
Troubleshooting the QCX Mini Op-Amps
In the QCX circuit, there are two op-amps that take care of the sideband cancellation: IC6 and IC7. In the schematic, they’re shown as IC6A, IC6B, IC7A and IC7B. This is because each IC contains two op-amps. Pin 8 is the power supply, pin 4 is ground. Pins 3 and 2 are the input for the first op-amp, and pin 1 is its output. Pins 5 and 6 are input for the second op amp, pin 7 is its output. IC7 takes care of the I channel, and IC6 the Q channel.
The surface mount devices are what are called SOIC8 packages. You can see them in the picture below. You’ll note that they have 8 pins and that one side of the chip has a bevel that the other side doesn’t have. You can find the chip layout on page 15 of the QCX Mini assembly manual so you can see which is which, but the following may help as well.
If you follow the schematic, you can see how the IC’s are configured:
The channels are independent until their signals are combined by R27 at the far right. IC6B feeds IC6A which outputs to R27, and the same for IC7B and IC7A. The difference is that IC6A and B are adjustable which is why in the alignment, you want less signal. That signal you’re nulling is the unwanted sideband. When R17, R24, and R27 are ineffective, that means that sideband cancellation isn’t happening, which means there is a problem somewhere in this circuit.
I showed how I measured the op-amp inputs and outputs in the video below. Essentially, the inputs should have a lower signal voltage than the output. We’re talking about audio signals, not DC voltage. My own op-amps showed fine DC voltage, but one of the inputs on IC7 was abnormally low when measuring the Peak-to-Peak AC voltages.
The best way to check the AC voltages is with an oscilloscope. A simple 200khz scope, linked to the right, is inexpensive and works fine for audio. Check the video for a demonstration: