Hello everyone! Yesterday I’ve finished my work on simple radio receiver that I was intended to develop from scratch all by myself, just to check if I am capable of doing some shortwave electronics. I’ve picked the Direct Conversion method as it occurred to me to be the most straightforward, and after some research on HAMs present on YouTube, I knew that very decent performance is to be expected when the job is done properly. So, here’s the whole story behind this simple rig.
Principle of operation
Direct Conversion Receiver (DCR) operating principle is very very simple if you know how a component called mixer operates. For those that do not know what mixer does: basically mixer is a component (an integrated circuit most likely, but other solutions also exist) that takes two input signals : RF (Radio Frequency) and LO (Local Oscillator) and produces output that contains both: the sum and the difference of frequencies present in RF and LO. So, as long as LO is in form of pure sinusoidal wave we can expect that signal at the output port will contain two copies of RF signal shifted in frequency domain by LO’s frequency. As an example, let us consider a case when RF is a sine wave with frequency of 1MHz and LO is also a sine but with frequency equal to 800kHz. We can expect that at the output of mixer two signals will be present: RF + LO = 1.8MHz and RF-LO = 0.2MHz. Here’s a picture to illustrate this:
Red arrows indicate what’s present at the output port. Ok, so how does the Direct Conversion make use of a mixer? Well, basically it uses LO that is equal to the frequency that you want to receive, and when you have LO equal to RF this is what happens:
One of RF signal’s copies is present around 0Hz which makes it possible to directly put it into our headphones and enjoy whatever is currently going on air. Well, maybe not directly, cause some amplification will still be needed.
I tried to keep things to be as simple (cheap and obtainable) as possible and this is the final result:
Remember that part when I mentioned that LO needs to be a pure sinusoidal wave, so it contains only one frequency component? In so called real life you can forget about that case, cause you’ll never be able to generate LO that is completely free of any distortion that occur as harmonic products in the frequency spectrum. And what happens when you inject such impure LO to mixer? You start to receive many frequencies at once! You’ll be lucky if there are no actual radio stations transmitting on any of your LO’s harmonics, but even if you are really-really lucky you’ll still receive noise present at those frequencies and that can totally ruin your reception. What can be done about it? Input filtering, that’s what!
Basically the job of input filter consists of:
- filtering (duh!)
- impedance matching – to achieve the best power transfer possible, since the signals that come from antenna are very weak you don’t want to lose any dB’s there!
As you can see on the schematic above it is made of two resonant circuits: first one being C13, C8, L2 and the second one L3 C12. First one is also an impedance transformer that rises the impedance from antenna’s 50 ohm to about 7k. This is done to achieve better frequency response (narrower filter) by the second resonant circuit.
More gain, please!
During tests it occurred to me that gain provided by the mixer might be not sufficient for convenient reception of amateur band as amateur’s rigs often emit output power that is lower by many orders of magnitude as opposed to broadcast stations. Therefore some simple stage of RF amplification had to be developed.
This RF amp is as simple as it could get: Q2, R13 R12. It introduces very little noise while boosting the signal by 20dB. Neat! I’ve used BC817-40 version for Q2, just to get the maximal gain possible with this type of transistor.
Mixing – theory meets practice
Yet another basic application of SA612 (a.k.a NE612 a.k.a NE602), no fooling around. This chip is an absolute classic, and any of young RF players should definitely get to know it better, as it offers great value for the money! Of course it has it disadvantages like low IP3, but those don’t really affect this receiver’s overall performance.
Local Oscillator a.k.a Tuning Knob
The beauty of SA612 reveals itself again when you try to build up Local oscillator. This IC has a spare transistor that can be used to build up a Colpitt’s Oscillator. All you need to do is to connect all the components that will determine it’s operating frequency, and since we are constructing Direct Conversion Receiver (so LO = RF, remember?) we need to be able to tune LO from 14MHz to 14.3MHz. This tuning range is determined by components: D1, C20, C24, L5, C17, C23. Diode D1 is so called Varicap and it changes it’s capacitance as the reverse voltage changes. This ‘reverse voltage’ can be adjusted by a potentiometer connected to J3.
When building this LO a frequency counter might become very handy. I’ve used my oscilloscope that shows trigger frequency to adjust all element values so LO does effectively tune exactly from 14.0MHz to 14.35MHz when my tuning potentiometer is in outermost positions. If you use a frequency counter please connect it to LO’s buffered output present at J4. This simple JFET buffer is there just to provide isolation between LO and your meter so that connecting it wont affect LO’s frequency. If you are not planning to fine-tune LO’s components then you can leave C21, R10, R11, C22, C18, Q1 unpopulated.
DCR selectivity (ability to receive only single station at time and filter out any others) is determined by filter included in audio amplifier stage. Here I’ve used cheap and well known opamp by Microchip: MCP6002. Overall gain of this stage is set to 40dB (10000 times) but, because of the presence of capacitors C1 and C3 it rolls off gently at around 2.5kHz. Voltage dividers constructed of R2, R7, and R4, R8, set the operating DC point to 2.5V so that audio signals have plenty of space to swing around.
The output of amplifier can be directly fed into headphones (32 Ohms) by DC blocking capacitor. No need of additional amplification/buffering, cause sound level is just fine.
No magic there. Five volts LDO and some caps. I’ve used TC1185, but if you are a big fan of 7805 then it should perform just as well. Just make sure that your LDO does not produce too much noise as it can (and will) be injected into many sensitive parts of this receiver.
I’ve put this thing all together on one of my spare PCB’s for double conversion receiver, with some modifications of course. This is how it looks like:
Voila! Not many components after all, isn’t there? Very attractive for RF beginners!
YouTube movie – or didn’t happen!
Yes..yes, I know 21th century calls for moving pictures. First one shows the DCR’s ability to correctly receive CW (Morse code) siganls that are often present in the lower part of 20m band.
And some SSB for dessert:
What’s worth noticing is that I used MiniWhip antenna to record all of this. If you don’t have one, you can just use a piece of wire connected to antenna port.