Fullscreen HTML5Med Res (??MB)HTML4

Vintage Radio The Reinartz 2 TRF receiver from Electronics Australia Just over 100 years ago, John Reinartz was the consummate radio designer. He was brilliant and his circuits put many amateur radio enthusiasts on the air. His innovation opened up shortwave as we know it to general use. By Philip Fitzherbert & Ian Batty R einartz’s publication, “The Reflection of Short Waves”, put forward theories that contradicted the academic teachings of the day. Those theories are now proven scientific fact. He was the first person to plan and take part in trans-Atlantic two-way communication at 100 metres (3MHz). He is also credited with contact from the US to England, and from the US to Australia, both for the first time, using 20 metres (15MHz). This was an incredibly short wavelength for the time. In the very early 1920s, Reinartz published a circuit for a two-valve receiver, a TRF circuit with adjustable feedback, which was published in the American Radio Relay League’s QST magazine, in the June 1921 edition. It was later updated for the March 1922 edition. Reinartz went on to head up the US Navy Radio and Radar Laboratory during World War 2. He held many patents, but never profited from any of them. Reinartz was honoured by many 94 Silicon Chip organisations in his lifetime; he was a real pioneer of early radio. Fast forward to July 1984. Interest in early radio circuits was fairly strong and Electronics Australia published an article by David Whitby on a receiver based on Reinartz’s. This used parts available in 1984, as opposed to its 1920s forebears, although some of them (like the valves) were already obsolete at the time, being available only as ‘new old stock’ (NOS). The kit was manufactured by Technicraft of Katoomba, NSW, and sold by several outlets. The unit was designed to look and feel like a set from the 1920s. It is built on a timber baseboard with a circuit board screwed to it (not a printed circuit board). It has sockets for the valves and the plug-in spiderweb coils. It carries connection points for headphones, aerial, Earth and other necessary voltages. The baseboard is pre-drilled, and instructions for assembly are Australia's electronics magazine provided. The baseboard was unfinished as supplied, but normal staining, polyurethane gloss coating and much sanding gives a very pleasant appearance on which to build the receiver. Circuit details The full circuit is shown in Fig.1; it is based on Reinartz’s design. The set uses the 200pF section of a double-­ gang tuning capacitor to tune the main winding of the plug-in coil. It uses what was called a ‘leaky grid’ regenerative detector. The antenna tuned circuit feeds to the grid of the first VT50 valve (V1). From the anode of the VT50, the second (90pF) section of the tuning capacitor forms the basis of the regeneration. This arrangement gives some regeneration circuit ‘tracking’, so you don’t have to continually fiddle with the reaction as you tune to different frequencies. The reaction control is fine-tuned by a Philips “Beehive” trimmer capacitor siliconchip.com.au Fig.1: the Reinartz-derived radio receiver circuit from Electronics Australia, July 1984, page 54. Capacitances are in microfarads (μF) unless otherwise stated, similarly resistance is in ohms (W). in series with the reaction coil, which is thus able to control the RF current through the coil. Adjustment of the capacitor provides precise control of the amount of positive feedback (regeneration). The physical layout here is interesting. The trimmer is fitted horizontally on the circuit board. Its top is connected to a threaded shaft, which exits via the front panel. This is a very clever use of modern parts, which I’m sure would have appealed to John Reinartz! This set’s circuit follows the later March 1922 outline. It added a second valve to give better matching to a set of headphones. This is described further below. L3 is an RF choke that prevents loading of the regeneration system by the following stage(s). It also operates in conjunction with bypass capacitor C5 to prevent RF currents from passing to the output stage. An RF choke in this position was always a feature of the Reinartz sets. siliconchip.com.au The VT50 medium-μ triode is an ex-RAF (UK) disposal item, designed in the 1920s. First manufactured in the USA in 1924, it is identical to the HL2K. The second valve, again a VT50, gets its drive via C6 into the grid of the valve, which is used as an audio amplifier for driving the headphones. The original circuit allows a choice of values for demodulator load R2 and output grid return R3. As built, these were 100kW and 1MW, respectively. How good is it? For any two-valve set to give 1mW of output with just 5mV of signal in is pretty impressive. While 1mW doesn’t sound like much, it’s loud for headphones. For ordinary listening, the level would be in the hundreds of microwatts range. The set tuned from the middle of the broadcast band at 1.05MHz, up past the 180m Ham band at 1.75MHz, and up to 2.7MHz. Australia's electronics magazine Tuning was affected by the antenna – no surprise, as it connects directly via C1 onto the stator of the main tuning capacitor, C2a. The best sensitivity was gained at 2.5MHz. This gave 1mW of output with full regeneration for only 5mV input with 400Hz modulation. The RF bandwidth, for a 3dB drop-off, was about ±500Hz (really!). With no regeneration, it spread out to ±177kHz, and needed around 140mV of RF input to give the 1mW audio output. This implies that regeneration increases the demodulator’s stage gain by around 30 times. It’s the equivalent of another HL2K running at full gain. Take your pick: you can have great sensitivity with a signal that sounds like it’s coming through a drainpipe, or the channel just next to your station (possibly more than one!). With moderate regeneration, the audio response was -3dB down at 2.5kHz. I should mention that the October 2025  95 Photos 1 & 2: the top view of the set with the reaction control in the centre highlighted is shown here, while the wiring hidden under the base is shown in the photo at lower right – it couldn’t be much simpler! low-­frequency cutoff is 22Hz (really, again). With only one coupling capacitor, it shouldn’t be surprising that the demodulator’s low-frequency response is so dramatic. Time to tune in to Beethoven’s Ode To Joy and experience the bass fiddles in their full glory. You will need a good set of earphones, though. At its best performance, 5mV input for 1mW output at 2.5MHz, the RF stage is delivering some 1.2V of audio to the output grid at maximum sensitivity. That implies that the RF stage gain is 240 (1.2V ÷ 0.005V). Not bad for a triode with an intrinsic gain (μ) of just 27 times. But the audio stage, for 1.2V input, Fig.2: the output voltage of an ideal triode is μ × Vg, ie, μ times the input signal voltage (Vg). However, the intrinsic resistance of a real triode (Rp) forms a voltage divider with the load resistance (Rl), reducing the magnitude of the voltage applied to the load. 96 Silicon Chip only delivered about 1.4V to the headphones, a gain of just 1.2. That’s only just more than unity. We can do better. The model for a triode is a voltage generator with an output of μ × Vin. But the generator has an internal resistance, Rp, which is in series with the load resistance, Rl – see Fig.2. The stage gain – and thus the output voltage – will depend on the relative values of Rp and Rl, using the triode voltage gain formula: The μ of the HL2K/VT50 is 27 at a specified anode current of 3mA and anode voltage of 100V. With the load as a pair of 2kW headphones, Av is theoretically about 2. The difference between the measured and calculated gain is easily explained; the HL2K’s Rp is quoted at 18kW, but that’s only for an anode current of 3mA. This circuit’s lower anode current of about 1mA increases Rp, so its increased series resistance means even less voltage across the headphones’ 2kW impedance. A quick back-of-theenvelope indicates an effective Rp of around 40kW. So, why bother with the second valve at all? It’s a question dating back to Lee de Forest’s low-gain Audions, with a μ value less than 5. Still, the Audion’s low output impedance allowed it to drive a transformer. Using a primary-­ secondary step-up in the transformer allowed the stage to develop substantial power gain, with the Audion itself needing virtually no driving power into its grid. The first valve needs a load impedance of 50~100kW to give useful voltage gain. Shunting that with headphones would drastically cut the stage gain, so the second valve’s main function – as originally designed – is to present virtually no loading to the first valve, thereby allowing the demodulator to develop its full potential gain. But a stage gain barely more than unity? Raising the load impedance would increase the second stage’s gain, so I switched my Marconi TF8793A wattmeter up to a 20kW load impedance. With that, I got a voltage gain of around 6.6, increasing the set’s Fig.3: a redrawn version of the circuit from Fig.1. Resistors R2 & R3 can be a range of values as shown in Fig.1. Australia's electronics magazine siliconchip.com.au V intage Radio Collection now covering March 1988 – December 2024 Updated with over 35 years of content Includes every Vintage Radio article published in Silicon Chip from March 1988 to December 2024. In total it contains nearly 500 articles to read. Supplied as quality PDFs on a 32GB custom USB All articles are supplied at 300DPI, providing a more detailed image over the print magazine. Physical and digital versions available Buying the USB gives you access to the downloadable copies at no extra charge. Or if you prefer, you can just buy the download version of the Collection. $70 PDF Download SC4721 siliconchip.com.au/Shop/3/4721 $80 USB + Download SC6139 siliconchip.com.au/Shop/3/6139 sensitivity to under 1mV of RF input for 1mW out. This higher impedance could be provided by a suitable 3:1 audio transformer, making the 2kW headphones appear as a 20kW load to the valve. This would also give better fidelity at full volume. The HL2K’s low anode current had it clipping at 1mW output with the 2kW load, while the 20kW load allowed a visually perfect sinewave of around 4.5V peak-to-peak to develop the 1mW output. Considering that a 1mV RF signal, modulated at the standard 30%, contains only about 0.3mV of audio, it looks like the overall gain from antenna terminal to output is around 15,000 times (4.5V ÷ 0.0003V). Beat that! We noticed one peculiarity in the design: neither of Reinartz’s two original circuits (like the March 1922 QST circuit) include a resistor from grid to ground for the demodulator, or any other form of biasing. This omission had frustrated Lee de Forest’s application of the Audion at audio frequencies, and was remedied by Lowenstein’s 1917 “Grid Bias” patent. Considering Reinartz’s formidable engineering skills, this omission cannot be a mistake. We sense a mystery lurking in this simple design. Perhaps, dear reader, you can help us out. Postage starts at $12 within Australia for the USB. See our website for overseas & express post rates. A final note; the HL2K/VT50 uses the “British” B4 base (Fig.4). This has pin 1 offset to provide indexing, unlike the American UX4, which has two large and two small pins (the follow-on B5 adds a fifth pin in the centre). The B4’s numbering is unusual, with pin 1 opposite pin 2, then pin 3 opposite pin 4. While it’s not obvious from the circuit diagram, this method places the anode and grid connections opposite each other, with the filament connections (at RF/audio ground) between to provide some shielding. The UX4 places anode and grid adjacent to each other, with an increased possibility of undesirable output-­ SC input coupling and instability. Fig.4: the B4 valve base has pins 1 & 2 offset from the centre, so it can only be inserted one way, even though all four pins are the same size. siliconchip.com.au October 2025  97