My first step in electronics was to build a crystal set (circa 1955). This had a germanium diode - so no "cat's whiskers" needed. I had a long aerial (25metres or so) which gave a good signal - especially from the Welsh Region BBC medium wave transmitter in Cardiff (about 3 miles away). Headphones were needed to hear the signal.
My first valve set-up was a 2 valve TRF radio (with an adjustable feedback to get a high Q). At this time there was an electronics components shop in Cardiff (Marks in the Wyndham Arcade), so parts could be sourced. The radio needed both a low voltage (for valve heaters) and high voltage (for amplification) battery supply.
I remember that there was a small metal working business near Cardiff General railway station that would cut and rivet aluminium to make a chassis and cut holes for the valves.
By rebuilding this set with a frame aerial and smaller batteries, I managed to make it reasonably portable -- and even took it to school occasionally.
When transistors became available (I remember paying about £1 for a sub-standard one -- equivalent to £20-40 in 2012), I put a 1 transistor amplifier in the crystal set with battery power.
I experimented with a radio-powered radio. With two aerials, one tuned to the nearby transmitter and then rectified and smoothed to provide power to the transistor amplifier of the radio proper (which used the other aerial). This did work - but using a battery was simpler and more practical.
Several people gave me old radios (no longer working) for spare parts and I bought ex-government kit via Marks. I bought an R1155 (war-time communications receiver) as ex-gov stock and modified it with a home-built power supply. It covered most wave-bands and could receive distant radio amateur transmissions.
This tempted me into the radio "ham" world. I had a 19 set (ex-gov, designed for tanks) which gave access to the 40 metre amateur band with RT (ie speech not Morse).
I also had another war surplus transmitter: type 53. This had a temperature controlled oscillator and twin 807 RF output.
I taught myself enough to sit the "Radio Amateurs Exam" and passed. Although my Morse was never fast enough to use.
I built an AVO meter in a wooden box which I used for testing. I remember borrowing a precision resistor from the school physics lab to calibrate it. I still have this AVO - but a modern multimeter costs £4, is more accurate and more comprehensive.
Following the pioneering path of a school colleague (Malcolm Ryall) I decided to build my own television. My parents had bought a TV in 1953 (for the coronation) but they would not let me watch everything that I wanted to (maybe because it was on too late). At that time there only was 1 channel (BBC). My mother said "when you have your own television, you can watch when you like". So I built one. The basis was an ex-gov radar set with a VCR97 electrostatic CRT (6 inch diameter with green phosphor). Since there was only 1 channel, I used a TRF circuit with a line of EF50's as RF amplifiers and, aided by the proximity of the Wenvoe transmitter, I tuned the video signal using headphones (listening to the intensity of the noise) until it was strong enough to give a picture. The sound was from just a single valve and needed headphones. For info on a kit that does something similar, see here.
I remember reading the Radio Designer's Handbook - a thick book that was quite technical. The symbol "j" occurred in many equations - I had not met complex numbers in school, so just carried on with "j=1". This usually gave values in a usable range...
When I left home to go to university, I no longer had time to spend on electronics and, to my later regret, my kit was given away to the nearby school (where my father was head).
I do remember one bonus of my amateur interest: the practical exam in physics at the end of the first university year (Honours Moderations by name) presented each candidate with a box with 3 connections... I guessed it was a transistor and I was quite familiar with what would result...
Another bonus is that I am prepared to try to mend electrical and electronic devices. I have my trusty old soldering iron and a multimeter. Usually the fault is a blown valve filament, a short-circuited capacitor, a blown power supply diode or a mechanical failure (switch, tuning knob, meter, erratic variable resistor,..). I have managed to repair items with small scale integrated circuits (74xx series chips) but in more modern equipment can only tackle the power supply.