[Dept. of Computer Science]

French Tombstone

Our Department and University.
See my other radios too! (Here is the next one.)

My oldest radio.
This radio was chased for my by Piet Hoen; it was offered to him but he wasn't interested and I was. Despite various modifications the set has gone through, the appearance of the radio still very much shows the original nineteen thirties design. From time to time I play the radio to enjoy the antique sound or demonstrate to my visitors what radio was like in 1935. I like to have it in my collection.

Data for this radio
Type Unknown
Bands GO (LW),PO (MG).
Cabinet Wood, 40x52x29 cm.
Tubes 2A7 (Mixer), 58 (IF stage), A257 (=57, detector),
47 (output), V580 (=80, rectifier), one dial light.
Controls Volume/off, Bandswitch, Tuning.
Produced In France, circa 1935/6.

Cosmetic condition of the set

A previous owner has probably refinished the cabinet at some time, and it was not done extremely well. Some spots are shiny and smooth, while others have a more dull and rough appearance. The look of the set could be improved by a complete sanding and refinishing operation and a fresh grille cloth. I currently prefer to leave the radio as is, because the technical condition of the radio does not merit such an operation.

Technical Details of the Design

A radio of the thirties differs in some repsects from a more modern radio of, say, 1950. First, the tubes used are of a 2.5V filament type and the output tube is directly heated; understandibly, these early tubes do not match the quality characteristics of later tubes.
The antenna circuit is a band filter.
The radio does not have Automatic Gain Control (AGC).
The Intermediate Frequency is circa 100 kHz, which is very low compared to the values of 452 to 472 kHz that were used later.
The detector is a penthode, while in later designs separate diodes were used for detection.
The volume control is not in the signal path, but is a variable resistor in the kathode circuit of the IF amplifier (tube 58). Turning the knob varies the bias between 5 and 30 Volts, and with high bias there is almost no sound. Needless to say that the volume control does not affect the volume of the Phono input.
The speaker was originally a field coil type, with the field coil being used as a choke in the fpower supply; these speakers were connected to the radio chassis through three wires (with a plug in this case) and often had the output transformer mounted on the loudspeaker. In this radio the loudspeaker was replaced by an electromagnetic type, with separate output transformer and choke coil mounted under the chassis.

The Technical Condition

The set was a puzzler for me and I took down the schematic diagram from the wiring. The drawing may contain some errors, but shows a few peculiarities.

What really puzzles me is the detector: I find it strange that the secondary of the second IF transformer is not grounded, but connected to the PU input. This arrangement necessitates shorting the PU input when listening radio stations (I hardwired this connection under the chassis). Even more strange is that the detector tube is biased (the 50k resistor in the kathode lead drops circa 6V), so I expected it to operate linearly and not detect at all! Connecting the IF transformer to the kathode does not improve the situation as I had expected.

The sensitivity of the radio is in no way comparable to a radio only few years younger, such as this French radio. I tried several things to improve this but without succes, and now I am not sure weather this is what can be expected of a radio with this design, or there is something wrong with the radio.

Finally I fear there is something wrong with the output tube: at a strong signal the voltage on g1 may suddenly rise to +160V, which effectively kills the sound. Switching the radio off and on makes the radio operate as usual.

Alltogether the condition of the radio is not exactly optimal, but it works and I like to play it from time to time.

Gerard Tel, gerard@cs.uu.nl.