11-12-2016, 08:28 PM
(This post was last modified: 11-12-2016, 08:37 PM by Mike Watterson.)
The details of 1j42a construction:
http://www.radiomuseum.org/forum/142_photos_curves.html
In general it's similar to the 1j24b, but the two g1 plates are not connected together internally. Also the spacing, anode and
Joe Sousa writes [my comments]
Joe's article is slightly misleading, the g1 is always plates, not rods. It's g2 rods that have become two slots. I don't know why the two anode plates are now wedges, perhaps due to low HT and close spacing.
Regular Rod Tube (not 1j29b, 1p22b, 1p24b)
Note that the 1j37b is essentially the 1j18b with the link between the two g1 plates cut. ALL the rod pentodes use plates either side of filament for g1. The 1j37b lacks some screen rods to reduce capacitance.
The 1j42a, also top view
The 1j37b has flat anode plates and two pairs of rods for the g2.
You can see that g2 has become two slots instead of two pairs of rods. This increases the effect of any differential voltage on g1a and g1b!
All the elements are much closer to allow low HT, 6V to 8V is best!
The two anodes are triangles instead of plates.
Thus the polarity of g1a and g2 doesn't matter! ANY mis-match will deflect the two beams away from the slot, reducing Anode current and gain. However the g2 triode gain will increase. Normally g1a and g1b "pinch" the space charge and thus reduce the Anode current, g2 and g3 are Electrostatic focus. ALL Rod pentodes. But on 1j37b and 1j42a the two g1 plates are not internally connected, thus just like X or Y on an electrostatic CRT / Scope tube, any differential voltage will deflect the beam!
Thus it's a sensitive full wave rectifier to the differential voltage on g1a and g1b. The common signal on both is amplified in the normal way to the anode, or indeed g2, if there is a differential voltage!
A full wave rectifier is either a detector or a frequency doubler purely based on having low pass or high pass filter on the anode!
The differential signal and "common" signal is also thus multiplied (mixed). Not separate signals on g1a and g1b, except fortuitously. For MAXIMUM mixer action one signal (Frequency A) must be equally applied to g1a and g1b, the other signal (Frequency B) must be applied in antiphase to g1a and g1b (or simply applied to only g1a OR g1b!).
The 1j37a works the same, but is designed for 45V (works at 22V) and takes MUCH more filament. The 1j42a and 1j24b take similar filament currents, 11mA to 15mA.
Russian Electrometer using 1j42a as input (g1a and g1b tied!) and two 1j24b as amps. The only known application circuit for 1j42a doesn't even use its strange behaviour.
For completeness, top view of the 1j24b and 1j29b, showing that the 1j29b is really two rod pentodes in parallel.
The 1j24b is later enhanced 1j18b, the 1j37b based on 1j18b.
The 1j42a was the last. Special low voltage design.
The 1j29b is doubling the 1j18b design to get twice gain and 0.5W or more. The 1p22b is short life for proximity fuse radar, the 1p24b is enchanced version that can be 1W long life or 800W peak repetitive pulses for proximity fuse (by definition, a short life!)
http://www.radiomuseum.org/forum/142_photos_curves.html
In general it's similar to the 1j24b, but the two g1 plates are not connected together internally. Also the spacing, anode and
Joe Sousa writes [my comments]
Quote:This rod pentode was designed for operation with 6V at the anode and screen grid, while drawing only 15mA from 1,2V at the filament. Other Russian rod pentodes were designed for at least 45V. As with any other tube that is designed for operation at low voltages, the internal mu of the 1j42a is low µg1g2=2.5 at 6V in triode connection and the total intrinsic gain in pentode connection is µg1A=16 for Anode voltages between 3V and 12V. However the Anode I/V curve has a flat spot at 6V, where the total intrinsic voltage gain from g1 to Anode should be some 10x higher, which is to say, more than 100. This flat spot is located where one would find the tetrode kink. The transconductance is nominally 400µS with g1=0V and g2=A=6V, g3=0V. Despite the low voltage operation, the data sheet still specifies the input resistance as 60kΩ at 60MHz, but the equivalent noise resistance is specified as a relatively large 90kΩ at 30MHz. This noise resistance is substantially higher than the 6kΩ equivalent noise resistance of the 1j24b.
...
The screen grid is now the largest element, consisting of two embossed plates with a central 240µ slit that accelerates the electron beam to the anode. [it's two pairs of two rods in all the other tubes]
The suppressor grid g3 is the only element that is still made of 4 round rods. The anode is a brass rod with a triangular cross-section. The point of the triangle is aimed to receive the bidirectional beam with two g3 rods on either side. [On all the other types, the anode is a plate. He's describing one half of the device. There are two anodes, two g2 slits and two pairs of g3 rods]
Joe's article is slightly misleading, the g1 is always plates, not rods. It's g2 rods that have become two slots. I don't know why the two anode plates are now wedges, perhaps due to low HT and close spacing.
Regular Rod Tube (not 1j29b, 1p22b, 1p24b)
Note that the 1j37b is essentially the 1j18b with the link between the two g1 plates cut. ALL the rod pentodes use plates either side of filament for g1. The 1j37b lacks some screen rods to reduce capacitance.
The 1j42a, also top view
The 1j37b has flat anode plates and two pairs of rods for the g2.
You can see that g2 has become two slots instead of two pairs of rods. This increases the effect of any differential voltage on g1a and g1b!
All the elements are much closer to allow low HT, 6V to 8V is best!
The two anodes are triangles instead of plates.
Thus the polarity of g1a and g2 doesn't matter! ANY mis-match will deflect the two beams away from the slot, reducing Anode current and gain. However the g2 triode gain will increase. Normally g1a and g1b "pinch" the space charge and thus reduce the Anode current, g2 and g3 are Electrostatic focus. ALL Rod pentodes. But on 1j37b and 1j42a the two g1 plates are not internally connected, thus just like X or Y on an electrostatic CRT / Scope tube, any differential voltage will deflect the beam!
Thus it's a sensitive full wave rectifier to the differential voltage on g1a and g1b. The common signal on both is amplified in the normal way to the anode, or indeed g2, if there is a differential voltage!
A full wave rectifier is either a detector or a frequency doubler purely based on having low pass or high pass filter on the anode!
The differential signal and "common" signal is also thus multiplied (mixed). Not separate signals on g1a and g1b, except fortuitously. For MAXIMUM mixer action one signal (Frequency A) must be equally applied to g1a and g1b, the other signal (Frequency B) must be applied in antiphase to g1a and g1b (or simply applied to only g1a OR g1b!).
The 1j37a works the same, but is designed for 45V (works at 22V) and takes MUCH more filament. The 1j42a and 1j24b take similar filament currents, 11mA to 15mA.
Russian Electrometer using 1j42a as input (g1a and g1b tied!) and two 1j24b as amps. The only known application circuit for 1j42a doesn't even use its strange behaviour.
For completeness, top view of the 1j24b and 1j29b, showing that the 1j29b is really two rod pentodes in parallel.
The 1j24b is later enhanced 1j18b, the 1j37b based on 1j18b.
The 1j42a was the last. Special low voltage design.
The 1j29b is doubling the 1j18b design to get twice gain and 0.5W or more. The 1p22b is short life for proximity fuse radar, the 1p24b is enchanced version that can be 1W long life or 800W peak repetitive pulses for proximity fuse (by definition, a short life!)
