RF Signal Generator

The RF signal generator is a must to have tool when playing with radio receivers. It is used to tune a resonant circuits and adjust the gain of different RF stages. Very useful feature of the RF Signal generator is its modulation capability. If it can modulate the frequency amplitude or frequency makes it non replaceable tool for RF design works.

Some time ago I have designed an AM modulator, which could be used for such purposes. It works fine in some cases, but it has the disadvantage to be not able to work as standalone device. It requires additionally power supply module and two signal generators - for the RF carrier frequency and for the modulating signal. This makes it inconvenient to work with it outside the house. I decided to create an RF signal generator working as fully functional stand alone device. Instead to base the architecture on the modern DDS chip, I decided to use the analog approach. As a basis I have chose an existing RF signal generator published here. Similar design is described also here. The credits for this design go to their authors. I repeated mainly the first design adding additional digital frequency counter instead the not very precisely radial scale analog calibration.

I will not go deep inside the circuit explanation - you can visit the links above and read all you need there.

I will show step by step instruction how to reproduce the design with minimum efforts and errors rate.

I took the circuit without changes from the site linked before.

For the PCB I used the toner transfer method and the picture attached here. Because the circuit does not contain any chips the image can be mirrored or can be printed without mirroring. The scale is also not very important.

After the finishing of the PCB I started the soldering. Firstly I soldered the low frequency oscillator (Tr3, TR4, TR5) and the surrounding devices. It has to oscillate at some sound frequency, but it did not. I played a little with it and found that shorting the R15 resistors brings the circuit in oscillations ( that is the green wire on the picture instead 100 Ohm resistance). If your circuit makes a sound with this resistor - you can leave it there.

After that I have soldered the RF generator part. To test it functionality I used two fixed capacitors 200 pF and two chokes 200uH instead the dual variable capacitor and the inductors bank. I saw that the RF generator works fine and produces AM modulated output signal. As described in the sources there is also some frequency modulation observed at the output signal.

For the supply I used 15V transformer salvaged from some old device. It had only single secondary winding and I used a bridge rectifier module, instead the two diodes shown in the original circuit.

As a frequency counter I intended to use this DIY kit.

For the RF frequency tuning two main parts are needed:

1. Variable dual capacitor
2. Switchable inductor bank

The capacitor I ordered in aliexpress.

The inductive bank I mastered around this switch - I used a thick copper wire, which I bended in a ring (open), and I have soldered five pairs of chokes with different values (for reference see the circuit). As a holder for the bank I used a right angle metal plate bought from the local store. For the axle extension of the variable capacitor I used a 5mm brass rod and Brass Motor Shaft Coupling (5mm to 6mm).

For the housing I used a nice plastic project box.To reduce the EMI I decided to cover the internal walls of the box with a copper conducting tape. I have glued the power transformer in the corner of the box using epoxy glue. On the back wall I have mounted a power switch and a fuse holder. The power cord I mounted through a gummy grommet and fixed it strongly with hot glue. I covered the bottom of the box shell with the copper tape. On the secondary transformer winding wires I put connectors to be able easy to connect/reconnect the power when experimenting.

To fix the main board I used M3 nuts, M3 brass spacers and M3 bolts.

The tuning capacitor and the inductor bank I also fixed with bolts (M4).

To fix strongly the counter board at the proper distance I decided to use a plastic spacer frame produced by 3D printer - the file spacer.stl can be used for this purpose. In the four corners of the frame I inserted and fixed with epoxy glue 20mm M3 bolts. I glued the frame strongly. A windows for the counter LED display was cut. All needed holes for the knobs were also cut. The devices in the marked in grey area in the schematics I decided to mount directly on the front panel additionally to the frequency counter board. I fixed the potentiometer and the peripheral devices on the front panel. As output connector I used BNC type one. If you prefer to use such type of potentiometers, you can print the file pot_holder.stl and use the part to fix the potentiometer. To fix more stronger the axle extension rod I used also 3D printed bushing glued on the front panel (bushing.stl). To make the look of the counter LED display better I put a piece of blue transparent plastic sheet in the front of it and I fixed the whole counter board with nuts.In the left part of the front panel i placed an audio jack for the external low frequency source input. The switch SW2 is also placed near it. When external audio source is not connected and this switch is turned in bottom position - the RF generator produces non-modulated pure sine signal. I would suggest - to set precisely the carrier frequency - first to start the RF generator in this mode, to adjust the frequency and after that to switch on the modulation.

The output signal of the RF generator is an analog sine (AM modulated or not modulated ) signal. At high frequencies its amplitude decreases a lot. To be able to measure the frequency of the signal with the frequency counter a pre-amplifier / signal form shaper is needed. I took the circuit presented here and soldered it on small perfoboard. The voltage generated by the 7805 chip I used to supply also the counter board. This assembling was working fine at low frequencies (less than 2Mhz), but at frequencies above these was failing.I have added another amplification stage (marked in red on the second circuit attached), but it did not improve a lot the situation.

I decided to change totally the approach.

I decided to use high speed dual amplifier. First stage - inverting amplifier with gain ~ 20 (you can play with the gain reaching the best performance) and second stage - voltage comparator. The purpose of the first stage is to amplify and filter the signal, the purpose of the second stage - to produce a digital levels input signal for the counter. I have chosen to use the TI dual high speed R2R input/output amplifier LM6172 - 100MHz unity bandwidth and 3000V/us slew rate !!!. You can see the schematic in the pdf file. The "Eagle" files are also attached. On the picture the bare PCB looks larger as it is. That is because the PCB contains two separate boards for two different projects. The PCB was produced in JLCPCB only for two USD. The results with this pre-amplifier board were much better - the frequency measurements were functional upon frequencies of 5 Mhz.

On the pictures and the video you can see the fully assembled device and the signal waveforms captured by a digital oscilloscope. The achieved parameters depend on the resonant circuit parts values. In the sites describing the original designs are given tables - list pf inductor values amd corresponding frequency ranges. I put inductors with values shown in the attached circuit and here are the frequency ranges, which the RF generator covers:

1. 173 kHz - 456 kHz
2. 388 kHz - 1088 kHz
3. 862 kHz - 2600 kHz
4. 1828 kHz - 4950 kHz
5. 3818 kHz - 5380 kHz

It can be seen that there is overlap between the sub-ranges - no empty frequency band exists. Using smaller inductor values could help to reach higher frequencies. As written in the sources - the theoretical highest frequency possible could be over 12 000 kHz.

As suggestion for the people, who want to try to repeat this design - do not strictly follow this guide. May be this implementation is not the best - Because the counter board is big and the resonant circuit parts bulky - the control knobs are put close one to another. May be better solution should be to put the counter board in the middle and the turning knobs from its both sides. I will recommend to try to keep all interconnection wires as short as possible. The ground wires also. I tried to use star type connection for the ground wires, but it is difficult always to be realized. As seen on the pictures the copper conductive tape is used also as global ground and shield - the different copper areas on different housing walls are joined together and soldered on multiple places.

There was a comment from Killawhat that exactly this counter is not the best solution - he has tried also it and found some problems. May be you should give some more bucks and will use better one. It should be possible also to scale the main PCB, and use 78L15 when the potentiometer is not soldered directly on the board. This could make the mechanical design easier and allow reaching higher working frequencies because of the reduced parasitic inductances and capacitors. The main idea - use you fantasy and creativity and the pleasure of the creation will accompany you. Good luck.