glass. Usually the alloy was thinly copper plated at the joint as glass adheres
readily to copper, and this was facilitated by coating the copper with borax
which allowed the glass to ‘wet’ the metal. As with Houskeeper’s technique,
the nickel alloy tube was usually turned thin at the edge (tapered or ‘feathered’) to compensate for any small difference in expansion during manufacture, and during the heating/cooling that occurred when the transmitter was
turned on and off. This second method, pioneered by Robert Le Rossignol,
was the one adopted by the M-OV in the manufacture of all its cooled anode
valves, viz, transmitters-‘CAT’s, modulators-‘CAM’s and rectifiers-‘CAR’s
and so early in the development of cooled anode technology, the engineering
of probably the most critical component had been mastered. For the craftsmen in the workshops however, his process must have been magical when
seen for the first time. For Robert tells us in his 1936 paper—in typically
modest fashion—that;
… the making of a large glass-to-metal joint usually inspires awe in the
uninitiated, but as a matter of fact, once the technique has been grasped
there is no great difficulty in making such joints …
The manufacture and operation of these valves however involved far more
than just the perfection of Robert’s anode joint. Much electronic engineering
and theory was involved; as in the design and materials of the filament; in
minimizing inter-electrode capacity, the inductance and ‘flash arcs’; attending
to the ‘magnification’ of the valve which depended upon the filament length,
its diameter, the radius of the grid and the anode/grid voltages, none of which
are remotely suitable for treatment here. The valve group also had to design
special apparatus for the workshop craftsmen to build the valves. The anode
of course is opaque and so it was impossible to see whether the electrodes
were sealed centrally in the anode tube. A ‘glass lathe’ was therefore designed
on which the electrode assembly (filament and grid) could be mounted,
centred, and then inserted into the anode tube to be sealed in. The lathe had
two heads to achieve this (one for the anode tube, the other for the electrode
assembly) in accurate axial alignment with one another. The heads, which
could be rotated at equal speeds, replaced the hands of the glass-blower who
would otherwise be unable to cope with the size and weights involved.
Mounted on the lathe was an array of blow pipes which the glass-blower
could adjust and move along the lathe bed to construct the valve. Once
constructed, the valves could only become functional by ‘pumping out’ to
obtain the necessary high vacuum. This was done on the ‘pump table’. For
the older glass valves the pump tables could accommodate just one valve, the
16 The GEC Laboratories 1930–1939
369
readily to copper, and this was facilitated by coating the copper with borax
which allowed the glass to ‘wet’ the metal. As with Houskeeper’s technique,
the nickel alloy tube was usually turned thin at the edge (tapered or ‘feathered’) to compensate for any small difference in expansion during manufacture, and during the heating/cooling that occurred when the transmitter was
turned on and off. This second method, pioneered by Robert Le Rossignol,
was the one adopted by the M-OV in the manufacture of all its cooled anode
valves, viz, transmitters-‘CAT’s, modulators-‘CAM’s and rectifiers-‘CAR’s
and so early in the development of cooled anode technology, the engineering
of probably the most critical component had been mastered. For the craftsmen in the workshops however, his process must have been magical when
seen for the first time. For Robert tells us in his 1936 paper—in typically
modest fashion—that;
… the making of a large glass-to-metal joint usually inspires awe in the
uninitiated, but as a matter of fact, once the technique has been grasped
there is no great difficulty in making such joints …
The manufacture and operation of these valves however involved far more
than just the perfection of Robert’s anode joint. Much electronic engineering
and theory was involved; as in the design and materials of the filament; in
minimizing inter-electrode capacity, the inductance and ‘flash arcs’; attending
to the ‘magnification’ of the valve which depended upon the filament length,
its diameter, the radius of the grid and the anode/grid voltages, none of which
are remotely suitable for treatment here. The valve group also had to design
special apparatus for the workshop craftsmen to build the valves. The anode
of course is opaque and so it was impossible to see whether the electrodes
were sealed centrally in the anode tube. A ‘glass lathe’ was therefore designed
on which the electrode assembly (filament and grid) could be mounted,
centred, and then inserted into the anode tube to be sealed in. The lathe had
two heads to achieve this (one for the anode tube, the other for the electrode
assembly) in accurate axial alignment with one another. The heads, which
could be rotated at equal speeds, replaced the hands of the glass-blower who
would otherwise be unable to cope with the size and weights involved.
Mounted on the lathe was an array of blow pipes which the glass-blower
could adjust and move along the lathe bed to construct the valve. Once
constructed, the valves could only become functional by ‘pumping out’ to
obtain the necessary high vacuum. This was done on the ‘pump table’. For
the older glass valves the pump tables could accommodate just one valve, the
16 The GEC Laboratories 1930–1939
369