The N.Z. Vintage Radio Project

The Columbus model 75, 6 valves plus magic eye tuning.

This model was the worlds first* affordable production domestic bandspread radio.

It solved the problem of tuning in short-wave stations which were notoriously finicky to get (and keep) on station in domestic receivers.  The solution was to spread out sections of the short wave spectrum to make tuning easier.  While communication receivers had used 'band-spreading' for a while, until now it had been too costly to implement and beyond the budget of a home receiver.

The model 75 prototype was developed from an existing receiver that Kem Collett, a lab engineer at RNZ, had modified at the request of William Marks (The MD) so his wife could more easily listen to the news back home in Russia.  The initial proof of concept was a success, eliciting a rare smile from WM and the question, "How soon can we put these into production?".  After Kem was called up to serve in WWII the task was passed to others in the Lab and the rest is history.  The four major contributors to the model are remembered in the Columbus booklet advertising the model 75.

Various cabinet styles were produced, most of which were carried over to the successor model (model 90).  Visually both models are similar, but the model 75 has the magic eye on the right, while its located in the center of the dial on a model 90.

There was also a battery version, the model 86.

1940 Columbus Promotional Booklet

*World first, if you don't take into account RCA's band spread radio from at least a year earlier - although the RCNZ set was certainly more refined and a 'better' all-round solution by all accounts

John W Stokes wrote of the model 75 in the NZVRS Bulletin:

Although mains-operated domestic radios incorporating one or more shortwave bands had been available since 1931, they nearly all used standard size tuning condensers on all bands, thus making tuning on the higher frequencies overly critical. It was this drawback, coupled with an inherent tendency to drift in the oscillator section which made serious shortwave listening of limited interest to the average set owner of the day. Thus it was that by the late 1930s attempts were being made by some manufacturers to provide receivers which could be more easily tuned on the shortwave bands and which were also more stable in operation. In 1938 RCA became the first manufacturer in the world to produce such a set, their model HF2/HF4. 
To obtain easier shortwave tuning it was obviously necessary to reduce the size of the tuning capacitors, but the need to cover the broadcast hand in one sweep placed a limitation on the minimum size which could be used; a suitable compromise could not be reached. So what could be done? 
One solution was to use a separate smaller capacitance condenser solely for tuning the shortwave bands, but this approach was largely impractical. Another approach was to use a split-stator tuning capacitor, as was done in Radio Ltd's model CNU of 1939, but the degree of bandspreading obtainable by this method was limited. A third approach was to artificially reduce the capacitance of the tuning condenser by inserting a small fixed capacitor in series with each section when switched to shortwave. It could easily be arranged so that a fixed capacitor of suitable size could be switched into circuit on each band, thus providing the desired degree of 'bandspreading', as it was called. This, then, was the heart of bandspread tuning. 
As used by Radio Corp, this system was modified using an additional fixed capacitor in parallel with each shortwave tuned circuit thus enabling the .use of only one shortwave coil in each stage resulting in a considerable saving in cost, not to mention the saving in space. 
It must now be recorded that New Zealand was in the forefront of this new development, for with the production by RCNZ of their model 75 early in 1940 they became not only the first N.Z. manufacturer to market a full bandspread receiver, but seemingly the first in the world after RCA. Although it was claimed in a brochure describing the model 75 that it was "....the first domestic radio of its type in the world.", the accuracy of this claim hinges on the words "of its type". Whether there were sufficient differences in the circuitry of the two makes to justify Radio Corp's claim is a matter of opinion, but certain it is that RCA were, by a considerable margin, the first in the world to produce a domestic bandspread receiver, even though it lacked the refinement of having variable tuning on the shortwave ranges of the RF and mixer stages! 
Another claim made by RCNZ was that the model 75 was put into production ...a full 17 months before any others, in October 1939," yet no receivers have. ever been sighted whose date coded serial numbers bear out this statement. Nowadays it is hard to believe that RCNZ were unaware of RCA's work in this area, though this does not alter the fact that the production of the model 75 was very much of a pioneer effort which gave the firm a clear lead lead in this part of the world. 
It was over a year after the arrival of the 75 before other firms in this country had bandspred models on the market. Even in Australia. AWA's first bandspread set did not appear until 1941, although one firm, Howard, had produced a semi-bandspread model as early as 1938. 
During its production lifespan, a matter of nearly three years, the model 75 underwent several modifications, the first of which was the replacement of the original permeability-tuned IF transformers with the (then) more conventional compression trimmer tuned types. At this time the first two valves, types 6K7G and 6K8G, were changed to Osram types KTW61 and X65 respectively. Incidentally, when Osram valves later became unavailable due to wartime restrictions, Radio Corp had to revert to using American valves. 
Final, and major, changes occurred when the three large diameter shielded aerial, RF and oscillator coils were replaced by more modern smaller types mounted in a compact metal box. The BC band coils were now litz wound, with the aerial and RF coils having non-adjustable iron cores in addition. Not only were the new coils more efficient but they also saved space. At this time also the simple continuously variable tone control was replaced by an 11-point switch together with improved circuitry providing a wider range of control. Meanwhile, the development of a new model had been going ahead and in 1942 came the model 90 to replace the the 75, though judging by observed serial numbers there was some overlap in their production. Outwardly the two chassis appeared to be identical, the only visible difference being that the magic eye tuning indicator had been repositioned directly behind the upper central part of the glass dial scale with the result that it was now slightly obscured. The altered position had been brought about by the need to provide space on the dial face for a tone indicator which took the form a moving white dot matching the band indicator at the other end of the dial. Although the tone switch still had eleven positions its associated circuitry differed from that used on the 75, being rather more complex in that as well as including variable inverse feedback from the secondary of the oversize output transformer, an iron-cored inductor formed part of a frequency selective circuit in some of the switch positions. Another externally visible change was that the 16 metre bandspread band had been replaced by a 16 to 49 metre general coverage band, a change which became standard practice in future bandspread models. 
Production of the model 90 extended from 1942 to 1949, but it must be remembered that there was a space of about three years when none were made due to wartime restrictions which prevented the manufacture of radios for civilian use. Even so, the 90 had the longest production lifespan of any model for after the war Radio Corp recommenced production by using pre-war designs. 
Following existing policy, the 75 and 90 were sold under both the Columbus and Courtenay brand names and, as usual, the cabinets differed sufficiently to make one brand distinguishable from the other. Variations in cabinet style also occurred within each brand. A distinguishing feature of the model 75 was the use of wooden knobs, each brand using a different style. By comparison the model 90 used bakelite knobs of Radio Corp's own make, again with different styles for each brand.