Schematic of the 4 tube radio, experimentally built in a way of the radio-pioneers (1910-1945). The radio acts as a cascade type receiver when the antenna coil has a high self-inductance, with lower self-inductances of the antenna coil you can put in a mixing process with the oscillator. Perhaps both phenomena are present and with (very) strong radio signals the oscillator cannot influence the process. This radio functionates only for stronger radio stations. In evening/dark (conditions) also stations from far away come through. This radio has all of the "problems" of the radio pioneers: "screaming" sounds, double tuning (antenna and oscillator) but handled with feeling and care will work acceptable for radio amateurs. By changing the value of L1 (antenna) and L2 (oscillator) you can try to reach other frequency bands. My experience was that I could not get the oscillator higher as 4 MHz and the oscillator only wanted to work with a ferrite rod as core. But I dit not experiment it all out, so wish you luck. I presume the concept is only successful for the stronger radio stations on middle wave and shortwave, though in the evening/night also weaker stations are received well (dit not test it all definitely). Measured and concluded: oscillator frequency = received frequency, received and amplificated HF energy "modulates" the oscillator frequency in tube 4. Tube 4 detects/rectifies the HF-AM signal. Signal strength of received frequency is important/critical for (un) distorted reception. Schematic works better with a capacitor of 270 P parallel to the antenna coil (better tuned for 1,4-1,8 MHz) and a capacitor of 50 N over the output (softer sound/less noise).