This is the very first peek at this receiver as received. It is a regenerative receiver with tuned RF amplifier, and ganged one-hand tuning. There is an antenna tuning control (left), RF gain control (middle left), the main tuning (middle right), and the regeneration control (right). This allows maximum signal to noise ratio, and a method of reducing the input to the regenerative detector for strong signals, so that it may continue to be operated at the maximum selectivity position of regeneration. Type 58 remote pentodes are used. The first audio stage is an impedance coupled 27 triode, and a transformer coupled class A push-pull power amplifier with 45 triodes follows.
UPDATES NOV. 28/16:
The output transformer S-11 was swept using a 600 ohm audio generator on half of the primary, and a 5000 ohm secondary resistive load: FLAT (+/- 1 dB) from 10 kHz all the way down to 10 Hz !!!
The radio was then powered up: +325V plates, + 52V cathodes, +273V effective plate voltage. 750 ohm cathode bias resistor = 35mA plate current each 45 triode, resting and full power (remember class A).
The audio output stage provides 320Vp-p or 2.6 watts of clean audio 1%THD into a 5000 ohm resistive load! (I was expecting 4 watts. The output transformer S-11 has an impedance ratio of 18000 ohms p-p to 6250 ohms, however the d.c. winding resistance of the secondary is 1020 ohms, resulting in 20% power loss internally, and on top of this I suspect the RCA tube handbook allows a 5% distorted waveform to get the 4 watts. Nevertheless, this transformer design tradeoff results in a flat frequency response from 10 kHz down to 10 Hz in my sweep test with or without the d.c. primary current of the tubes!!!)
The driver transformer S-51 is almost as good. With a 9000 ohm source resistor and signal generator (simulating the 27 driver tube plate resistance), I measure a response (+/-3dB) from 40 Hz to 8 kHz. However when the radio is powered up, with an input to the 27 grid I notice a severe low frequency fall off of the overall radio: -3 dB 200 Hz, -10 dB 80 Hz. UPDATE: The cathode resistor bypass capacitor (0.5 uF/2000 ohms) is not large enough, so the effective plate resistance of the 27 driver increases above 9000 ohms, and since the unbalanced current carrying driver transformer primary impedance is falling also, severely shunts the audio frequencies below 200 Hz. Perhaps a deliberate hum reducing design?
UPDATE: I added another bypass capacitor, 8uF, across the driver cathode resistor. Overall radio response -3dB 7 kHz, -3dB 90Hz, -6dB 50Hz. The unbalanced current in the primary winding must be reducing inductance. Added an 8200 ohm resistor across the primary, penalty is 6dB gain, but improved flattening of the response: -3dB 7.5 kHz, -3dB 50 Hz, -6dB 30Hz. Very good! In fact, with the bass-boosting effect of the regeneration control set for AM operation, this will sound high-fidelity in a good 5000 ohm National speaker.
First conclusion: this receiver from 1932 was the earliest equivalent of the later 1935-1950s push-pull 6F6/6V6 shortwave superheterodyne receivers with many more tubes, such as the National NC-100/101, HRO-50/60, NC-183, and Hallicrafters SX-42, SX-28, SX-62. To have that in 1932, just a few years after the best receiver was a UX-201A detector triode and audio headphone amp, was nothing short of revolutionary.
The famous James Millen designed this receiver. This is not a toy broadcast receiver, but a serious shortwave instrument with both general coverage and amateur bandspread plug-in coil sets available. If you couldn't work Australia from North America with this set, you couldn't with any other set either.
This receiver was the last of the original National regenerative designs, following the SW-4, SW-5 and SW-45 models. Incidentally, the SW-5 was available with push-pull 27 or 45. The choice was based on low current drain of the 27s, resulting in lower ac ripple from the power supply, and the ability to connect headphones there too if one needed super gain without dangerously blasting ones ears off, vs. the 45s increased power and fidelity using flatter transformers for broadcast stations. In practice, the 45s were fine, no added ripple with a good supply and no one in their right mind would connect headphones to a push pull power amp driven by a howling regenerative receiver anyway! The smaller SW-3 omitted the push-pull stage and production continued well into the war years. In my opinion, the SW-58 is the most aesthetically pleasing design. Many ARRL handbook receiver designs in the 1930s were based on this.
*** Stay tuned for RF testing! ***