295 Triple-Twin introduction in 1932
295 Triple-Twin introduction in 1932
Fellow Radiophiles:
The attached article from the American magazine Radio News March 1932 introduces the 295 Triple-Twin power tube that exploits the concept of "dynamic coupling" between two triodes, where a cathode follower triode with a unipotential (indirectly heated) cathode drives the grid of a grounded filamentary cathode power triode into strong forward grid bias. There were several tubes developed by the Speed (Cable Radio) company using this principle. Other models used unipotential output cathodes.
The author explains a method to reduce the non-linear distortion when the output grid goes between forward (positive) and reverse (negative) bias. The relatively low mu of the output triode makes it necessary to drive the grid positive as well as negative to deliver the promised 4.5W audio power from a 250V supply in single ended configuration. If the output triode section had a high mu and unipotential cathode, little power would be lost by restricting operation to the forward bias at the output grid.
The high input sensitivity and the two triode sections made it possible to use the input triode as a anode bend detector (C-bias detector) with the input grid-cathode bias adjusted near cut-off. A 10V carrier at the input grid was recommended for optimum detector operation. This configuration presupposes the use of volume control before detection. This has the disadvantage that lower level signals will be more distorted due to sub-optimal RF levels at the detector. In later radios, the practice was to place the volume control after the detector to allow for full signal operation at a low distortion diode detector.
Best Regards,
-Joe
Attachments:To thank the Author because you find the post helpful or well done.
More on triple-twin and class B
Fellow Radiophiles:
RM member Jacob Roschy has an extensive post showing various configurations of class B forward grid bias for power tubes, including the triple-twin. I used the Google translator to read Jacob's very comprehensive German language post.
Otto Frosinn and Alan Douglas have uploaded photos of the 295, including the shipping carton.
A further clarification is in order about "The relatively low mu of the output triode..." that I mentioned in the first post. The ideal mu for the output stage would be high enough such that little current would flow with the grid at zero volts. This means that little current is wasted by restraining the output grid drive to positive values for lowest distortion. If the output plate supply is 250V, then a mu of several hundred (>250) would be needed to insure that little current flows when the input grid is near 0V, but still positive.
The 1932 article focused on a wide range of applications for the triple-twin configuration. These include swinging the grid voltage positive and negative. This would generally cause distortion, however, the author claims that the reduced positive grid swing is counter-acted by increased gain in the same region.
Jacob Roshy expressed concern for the anode bend detector configuration, where the smoothing filter is at the cathode. A relatively large value of capacitance would be needed for ths smoothing filter, but at low voltage. If the smoothing filter corner were set at 5kHz, then the attenuation at 455kHz would only be 100=40dB, and there would be 100mV or carrier left from the recommended 10V input. This residual carrier would need further filtering at the output plate to avoid interference with the antenna circuit of the same radio and with other radios.
Robert Weaver has pointed out correctly that the input triode does not work as a cathode follower, but in an equivalent configuration to grounded cathode, because the input transformer drive is applied between the input grid and the input cathode, and not between the input grid and ground, as would be the case for a cathode follower. The source impedance for the input triode is the same as the impedance between the plate and cathode with the grid referred to the cathode, which is the same as the plate impedance of the input triode. In short, the input triode drives the grid of the output triode with plate impedance.
In a IRE paper "General theory and application of dynamic coupling in power tube design" dated July 1936, Charles F. Stromeyer shows how the non-linear cathode impedance non-linearity of the input tube configured as cathode follower, cancels the non-linear load impedance of the forward biased grid of the output tube. Stromeyer calls this non-linearity cancellation scheme "Dynamic coupling".
Late in the tube era, circa 1960, a line of high mu power triodes was produced as pulsed mode high voltage regulators in color TV sets. See how the 6JD5, which is one of these tubes, might be used with forward grid bias in a triple-twin configuration for very linear audio power amplification.
Best Regards,
-Joe
To thank the Author because you find the post helpful or well done.