![[Structure of RF tubes.]](PentagridConverters_files/Remote-Cutoff-Grid.png)
This document describes how pentagrid converter tubes operate, touching briefly upon the triode/hexode and other converter tubes. It does not describe how vacuum tubes operate in general. The reader is assumed to understand the basic functioning of vacuum tubes, and the specific functions of the screen grid and suppressor grid in pentode tubes used at radio frequencies. The reader is also assumed to understand the general theory of operation of superheterodyne radio receivers.
For a presentation of the general operation of vacuum tubes, see the author's documents "Tube Theory", and the followup "Multigrid Vacuum Tube Theory", available on the VRPS web site.
The pentagrid converter is somewhat more properly known as the heptode converter, for reasons described below. The two names stem from Greek "penta-", meaning five for the number of grids, and "hepta-" meaning seven due to the number of electrodes, viz., the cathode, plate, and five grids. This tube was designed to reduce the number of tubes required to make superheterodyne radio receivers. Here, in one tube, we find combined the functions of electron coupled oscillator, mixer, amplifier, and gain control. This tube is a truly amazing device, combining complexity in simplicity, and one stands in admiration of those who conceived of it, and perfected it in its operation.
The converter grids are, perhaps, less well understood elements than the screen grid and suppressor grid used in tetrode and pentode tubes, both as to their functions and their relationships to one another. The grids are named simply by number, proceeding from the cathode to the plate in order, G1, G2, G3, G4, and G5. A few of them have alternate names, as described below.
Before delving into the specifics of how pentagrid frequency converter tubes work, a word about how Radio Frequency (RF) amplifier tubes generally operate is in order. Specifically, a bit about how gain control is achieved is necessary. Superheterodyne receivers have a feature (inappropriately) called Automatic Volume Control (AVC). AVC is intended to address two annoyances present with radio receivers. One, called "fading", occurs when one is listening to a distant station.
As an RF signal makes it way from a distant transmitting station, it encounters variations in atmospheric conditions. These conditions may cause the signal strength at the receiver to vary over time, sometimes strengthening, sometimes weakening. When this happens, the listener finds himself turning the volume control first down, then back up, every few minutes or in extreme cases every few seconds.
Another annoyance, called "blasting", occurs when one is tuning a receiver looking for a distant station. One turns the volume control up relatively high so as not to miss finding the station sought, and tunes across another, much stronger station. This results in a startlingly loud burst of sound.
The actual problem is that of varying signal strength, not that the volume control needs to be automated. A way was needed to vary the receiver's sensitivity, not the volume control setting. Tube designers came up with RF tubes having variable gain. This was done by altering the structure of the control grid. Up to this time control grids were made as helices with the turns uniformly wound along their length. These grids are called "sharp cutoff" grids. Non uniform spacing would result in distortion, which was considered to be undesirable. In variable gain tubes, the control grid is deliberately wound with non uniform spacing. See figure 1.
Tubes with this altered grid structure, like the 6SK7, are called "remote cutoff" tubes, sometimes "supercontrol". Remote cutoff tubes are not suitable for AF amplifiers, since the variable spacing of the turns results in distortion. For low power RF amplifiers this is not an issue for two reasons. One is that only a small portion of the plate transfer characteristic gets used. Another is that the output normally gets heavily filtered by tuned circuits, which remove the distortion.
The gain (gm) of a tube is defined as the slope of the plate transfer characteristic curve, the steeper the curve the more gain. With a uniformly spaced grid structure there is a long nearly straight portion of the plate curve where the slope (gain) hardly varies at all, no matter the DC bias on the grid. See figure 2.
![[Plate Transfer Characteristics of RF tubes.]](PentagridConverters_files/Remote-Cutoff-Characteristic.png)
With a remote cutoff grid, however, we see that the slope of the plate curve definitely depends upon the DC bias on the control grid. With more negative grid voltages, the slope of the plate curve definitely decreases, i.e. the gain of the tube is reduced. The reason for this is that as the negative DC bias on the control grid is increased the outer portions of the grid actually block all electrons from passing by, that is, they are "cut off". The more widely spaced center windings exert less control over the electron stream (less gain), so do not cut off. In this manner, the DC bias on the control grid can be used to control the gain of the tube.
The detector produces a negative voltage proportional to the received signal strength, which can be used as this bias. This control bias is derived from the detected audio, and all audio frequencies must be filtered from it. The normal way this is done is with a resistor (typically about 2 meg Ω) and a capacitor (typically 0.05 μF). It is important that all the audio signal be removed, or there is risk of audio frequency oscillations in the RF tubes. It is also important that the AVC circuitry not be too heavily filtered, or it will not be able to follow rapidly changing signal strengths.
As we shall see, all pentagrid converter tubes have one grid with sharp cutoff characteristic, and one grid with remote cutoff characteristic.
To understand the operation of the heptode converter, one may consider it to be actually two tubes, which overlap and share elements. Your attention is directed to the cathode and the first two grids, G1 and G2 in the 6A7 tube. See figure 3. These three elements may be considered to form a triode, which is used as the local oscillator. G1 and G2 are sometimes termed the "oscillator grid" and the "oscillator plate", corresponding to their functions, respectively.
![[Schematic of the 6A7 Pentagrid Converter Tube]](PentagridConverters_files/6A7-schematic.png)
G2 is actually used as a plate, and has a structure which is rather unlike the other grids', being a very "open" structure. While it is connected to a source of B plus, it collects few of the electrons as they proceed to the plate element of the tube. Indeed, an oscillator needs a power gain of only just over unity.
This oscillator is of the sort one sometimes calls "electron coupled". That is, the output of the oscillator is not connected externally to other tubes for further processing. However, as there is only one electron stream inside the tube, what affects the stream affects the output of the tube taken from the plate. In effect, the cathode and first two grids act as an electron gun which shoots a variable density electron stream into, and provides the space charge for, another tube which is a variable gain tetrode.
The tetrode is formed by the cathode, G4 (serving as control grid), G3/G5 (serving as screen), and the plate. This tetrode serves as the mixer, and as such it intentionally has a non linear transfer characteristic. The output of the tube as a whole comes from the plate.
The local oscillator signal is injected into the tetrode portion of the tube, as described above, where it mixes with the received signal, which is applied to G4. There the IF signal is developed, and amplified to a degree which is controlled by the DC bias on G4. Since the received signal is connected to G4, G4 is also called the signal grid.
If one carefully studies the tube manuals, three different types of heptode converters/mixers can be found. There are two types of pentagrid converters with slightly differing grid connections, and the pentagrid mixer.
In one type of tube, exemplified by the 6A7 described above, we find G3 and G5 are internally connected within the tube, and G4 is used for the radio frequency signal input to the tube. In this tube, G1 is wound with even spacing for sharp cutoff characteristic. By contrast, G4 is wound for remote cutoff characteristic.
Tubes like the 6A7 give very adequate performance at medium frequencies, up to about two megahertz, after which their conversion gain begins to decrease. This is due to reduced efficiency of the local oscillator portion of the tube. So, this tube has limited application at short wave frequencies, and is practically useless at FM radio frequencies used today.
In another type of tube, exemplified by the 6SA7, we find that G2 and G4 are internally connected, and no one electrode serves as the sole plate connection for the oscillator portion of the tube. These two grids also shield the radio frequency signal input grid, G3 from the plate, and act as the screen grid in a pentode. The radio frequency signal grid G3 is wound with variable pitch, as mentioned above, for AVC function while G1 is wound with fixed pitch, as before. The final grid G5 acts as a conventional pentode suppressor grid. See figure 4.
![[Schematic of the 6SA7 Pentagrid Converter Tube]](PentagridConverters_files/6SA7-schematic.png)
In these tubes, the signal grid G3 has very little effect upon the space charge surrounding the cathode. This reduces the "pulling" effect, in which the tuned circuit attached to the signal grid affects the tuning of the local oscillator. This effect may be noticed at the high frequency end of the reception band, where in relative terms the tuning in the antenna circuit is not much different from that in the local oscillator circuit.
In the third type of tube, properly called the "pentagrid mixer" and exemplified by the 6L7, the grid connections are as in the 6SA7. However, the roles of grids G1 and G3 are reversed. G1 is the RF signal grid, wound for a remote cutoff characteristic, and AVC bias voltage input. G3, sometimes called the modulator grid, is a separate control grid, wound for sharp cutoff, and is used to accept the local oscillator signal. The local oscillator signal must be provided by a separate tube, usually a triode due to the lower noise present in triodes. See figure 5.
![[Schematic of the 6L7 Pentagrid Mixer Tube]](PentagridConverters_files/6L7-schematic.png)
G3 exerts a strong control over the current flowing to the plate. G2 and G4 are internally connected, and act to accelerate the electron stream toward the plate. G5 is used as a suppressor, similarly to that in a normal RF pentode tube. In this form, the tube provides extreme isolation between the oscillator and antenna circuits, and may be used at frequencies in the hundreds or thousands of megahertz. The pentagrid converters with grid connections similar to the 6SA7 may also be used with separate local oscillator injections (into G1, however) with improved high frequency performance, though not as effectively as the pentagrid mixer.
Note that, while the pentode form of pentagrid converter (e.g. 6SA7) and the pentagrid mixer (e.g. 6L7) have the grids internally connected in the same manner, they are functionally rather different, as the grids are reversed as to which has the remote cutoff characteristic, and the pentagrid mixer requires a separate oscillator tube.
The two types of pentagrid converter and the pentagrid mixer do not exhaust the list of multigrid tubes designed for frequency conversion in superheterodyne receivers. One other fairly common one is the triode/hexode converter, exemplified by the 6K8 tube, though triode/pentode (6AT8), triode/heptode (7J7), octode (7A8, though not a combination tube), and other combination tubes also exist. They all feature shared internal elements, at least the cathode, and sometimes the triode grid is internally connected to one of the other unit's grids. The functioning of these tubes is all very similar to that of the pentagrid mixer. Here, we'll examine the triode/hexode 6K8 tube. See figure 6.
![[Schematic of the 6K8 Triode/Hexode Tube]](PentagridConverters_files/6K8-schematic.png)
Functionally, the triode unit is independent of the hexode mixer, and serves as the local oscillator. Note that the control grid of the triode is internally connected to G1 of the hexode mixer, which is wound for sharp cutoff as in the pentagrid mixer. This serves to couple the local oscillator signal into the hexode mixer unit. The signal grid is the hexode's G3, which is wound for remote cutoff characteristic. Having a separate triode unit for the oscillator has three benefits. First, a triode oscillator is less noisy than a multigrid tube, though any multigrid mixer is going to be noisy, and this one has two grids at positive potential as well as a plate, so is subject to partition noise current. Second, and more important, there is increased isolation between the oscillator tuning circuitry and the antenna or signal tuning circuitry. Especially at high frequencies this is important to reduce pulling effects. Of course, there is the obvious benefit of not requiring a separate tube for the oscillator, which is the prime motivator for preferring this type of tube to the pentagrid mixer.
In summary, there are pentagrid converters with G3 and G5 internally connected. Grid G1 is wound for sharp cutoff and is used as an oscillator grid, while grid G4 is the signal grid, wound for remote cutoff/gain control. These tubes are useful up to moderately high frequencies, but not much beyond two megahertz.
There are pentagrid converters with G2 and G4 internally connected. Grid G1 is wound for sharp cutoff and is used as an oscillator grid, while grid G3 is the signal grid, wound for remote cutoff/gain control. These tubes have improved high frequency performance by comparison with the other type of pentagrid converter.
There are pentagrid mixers with G2 and G4 internally connected. Grid G1 is wound for remote cutoff/gain control and is used as the signal grid, while grid G3 is wound for sharp cutoff, for connection to a separate local oscillator tube's output. These tubes have improved high frequency performance by comparison with both types of pentagrid converter.
There are other combination frequency converter tubes featuring two units in the same envelope. These comprise a triode oscillator and a multigrid mixer unit with characteristics similar to that of a pentagrid mixer with separate oscillator, but do not require a separate oscillator tube.
All of these multigrid frequency converter tubes are noisy since there are multiple elements at positive potential, and the electron stream divides randomly between them. For this reason, these tubes are often preceded by an RF amplifier. This has two benefits: it almost completely eliminates image frequency reception, and improves the signal to noise ratio, since there is more signal available. The RF amplifier in this case is not used to increase receiver sensitivity. That is better done by having additional IF amplifier stages. The RF amplifier does little to enhance adjacent channel rejection. Additional IF stages do so, and are less prone to oscillation than multiple additional RF amplifier stages.