Experimenting with and repairing radios requires signal sources. I have used three different signal sources so far:
- My MFJ259B antenna analyser is a useful tuneable signal source covering 1.8-170MHz. Its output have various spurious and harmonics. The biggest issue I have is that it drifts in frequency. But as a starting point, it has been very useful.
- A Direct Digital Synthesiser based on the Analog Devices AD9851. Mine came from a Chinese supplier via ebay for around £30. This is a 10 bit DDS clocked at 180MHz, so able to give a sinewave output from near DC up to around 70MHz. It also generates an alias frequency at (180-F)MHz, which I've been able to use to generate a signal in the 2 metre band. The DDS is stable and has very fine tuning steps and its output is essentially harmonic free; but being a 10 bit DAC the noise floor is relatively high.
- An Si570 based oscillator using the PA0KLT module (http://sdr-kits.net/PA0KLT_Description.html). This has very fine tuning steps and covers approx 4-280MHz at around 0dBm. It's a square wave output, so has very high harmonic content but essentially it's otherwise spurious free. This is a form of digital synthesiser and is to all intents and purposes as stable as a crystal, but you can set it to any frequency in that rangeMine cost about £35.
We are somewhat spoiled with things like this nowaways - stable sources, synthesised, and low cost!
Ideally I'd like a bench oscillator with the tuning range of the Si570, an ability to do frequency sweeps (for passband checking), and low harmonic level. I'm contemplating making a set of diode switched low pass filters, possibly with a gain block to give ~10dBm out, as an add-on. That would give a very useful bench signal source. I'm also contemplating a 12 bit DAC DDS (AD9854) for which there's a module available for around £50 which will be able to do automatic, programmed sweeps.
A bit of a "proof of principle": could I design from scratch a frequency counter? 30 years ago the appropriate design route was to use an Intersil IC7216 (as I recall) counter timer chip: but somewhere along the way these have been completely displaced by microcontroller based designs. The PIC series of microcontrollers are able to implement a true "single chip" frequency counter that will work to beyond 40MHz; but all of my design experience is with the Atmel AVR family, whose timer counters won't operate at that speed... so what could I do?
A dummy load with a difference: mine also includes a 30dB coupler, to sample the incoming power.
Everyone should have one! A switched attenuator is incredibly useful when making RF measurements, and they are very easy to make. I have used 7 sections switched as 1dB-2dB-2dB-5dB-10dB-20dB-20dB allowing up to 60dB attenuation.
This was made from the excellent Radio-Kits kit after I came across it, entirely by chance, at a rally. It measures forward and reverse power over the HF bands and displays its findings on an LCD display. There are various options for how the data is displayed. A microcontroller does all the processing, and includes a "peak hold" function to display PEP values.