The 8122 Enigma K9AXN Service Note 42a
The 8122 is one of the most misunderstood power tube in use today. It is a close spaced compact high power tetrode that has generated an incredible amount of diverse opinions --- a love hate relationship. The following information should help to choose the correct tube, operate within the design limits, and maintain it's health. There is every reason to expect an 8122 to last 30 years with appropriate care.
The focus of this information will be on the Hallicrafters SR-2000 and other transceivers that use the 8122 in class AB1 mode i.e. the NCX-1000. Nearly all desktop transceivers that implement the 8122 using class AB1 mode for SSB also use high impedance voltage amplifiers as drivers for the finals.
What is class AB1 and why use it:
AB1 mode is confined to a range of control grid voltages up to zero volts which is an exceptionally linear range of operation. It requires strict controls to prohibit any control grid current resulting from driving the grid above zero volts. The drive circuits are voltage amplifiers requiring very little space, power, and exercise tight control over drive limits. Most transceivers use class AB1 mode for SSB operations.
In most cases the AB1 design limits the available power out from the 8122 to a lower value than AB2 operations because AB2 mode allows the tubes to be driven above zero volts exploiting the extended capabilities of the tubes.
Note, I said in most cases. The following will help to understand why that is not always true.
THE PERSONALITY OF THE 8122:
The best description of the eccentricities of the 8122 is a conversation between a group of Hallicrafters engineers late in 1965. Quote (We have to purchase a crate of 8122's to get a handful of tubes that we can use). If you have purchased 8122's over the last few years you to have experienced the same problem with tubes for any class AB1 transceiver and any beam type of tetrode.
The Hallicrafters SR-2000 final amplifier design point was to deliver 950 to 1050 SSB watts to the antenna and the output circuits were sized accordingly with the usual tolerances of the times. See later limits depending on the 8122 selected.
The Hallicrafters people had the choice of using the 8122V1 factory Matched pair or the less expensive hand selected 10% method. The 10% method was chosen.
The SR-2000 is designed to run the 8122's at 200ma of idle current. The 8122 data sheets and most tube data sheets formally defined plate current @ screen, control grid, and plate voltages.
The 8122 has a bias range from -46 to -22 volts that will result in 200ma of plate current with 1700 volts plate and 300 volts screen. The low and high values will will result in between 140 and 300 ma at the above voltages. The (-46/-22 volt) 2/1 range is common if not universal in beam tetrodes. That partially explains the dilemma at Hal where they needed a crate of 8122's to render a hand full of usable tubes. The factory matched 8122v1 tubes have the same bias range variations, and even though they are matched throughout their conduction range, chances are good they may not be right for your radio.
No 8122 vendor has explained the reason for the wide bias range, whether it be manufacturing tolerances or a deliberate decision to enable class AB1 amplifiers to deliver higher outputs than expected. You will see as we go, that wide bias ranges will be inherent in other tubes i.e. the 6146/6146B, 6HF5 used in the SR-400, the 6KD6 used in the SR-400A, and many other tube types.
Beam tetrodes are designed with precisely aligned control and screen grids as follows. If you look toward the anode from the cathode, you will see only the control grid because the screen grid is in the shadow of the control grid and if you look toward the cathode from the plate you will see only the screen grid because the control grid is in the shadow of the screen.
If either grid is skewed out of the shadow of the other two things are effected. The plate to control grid capacity increases and the idle bias parameter increases. This happens in any beam power tube.
One thing is certain, The tube data sheets that define the inter-electrode capacity between the anode and control grid in tetrodes speaks to the average tube. The range from the lowest to highest grid to anode capacity is great enough to require design changes for bias and stability especially in multi band transmitters.
When Burle purchased the 8122 production, they color coded the 8122's to enable designers to more accurately design the circuits that use them. This enables them to acquire what is a closely matched pair with effective inter-electrode personalities that matched their design parameters. There are four colors; Red, Blue, Green, and Black representing the idle bias levels. Within each color group there is an approximate range of just over 6 volts for idle bias in the SR-2000. Each color range will deliver greater output from Red to Black respectively when used in class AB1 service. In class AB2 all would deliver about the same output.
Great, lets go out and buy the Black Burle tubes! STOP, they won't work in the SR-2000 without redesign of the bias, neutralizing, and other circuits and when finished the design limits of the radio will have been exceeded.
The SR-2000 using the red tubes will deliver approximately 800 to 950 watts to the antenna; The Blue 950 to 1050 watts, Green 1050 to 1150 and the black have not been used. The Black tube exceeds the output design limits of the SR-2000 and some reach full conduction very close to or before zero bias where linearity becomes a problem: Too late for AALC to intervene.
WHY DO 8122's BEHAVE AS THEY DO:
The basic reason is element geometry. The higher the bias range the more like a triode they behave. They need more positive feedback (Neutralizing) to stabilize them and the control grid must be operated at a significantly more negative bias level to idle them. This increases the useful peak envelope current
The SR-2000 is designed to use the Blue Burle tube or any 8122 that has the same bias range. Many have tried to use the Black which comes to no good end. The Blue are plug In's and the border line Blue/Green will work without mods in the SR-2000. The high bias Green tubes require modification to the bias circuits in the power supply and radio, as well as mods to the neutralizing bridge. If you attempt to neutralize the radio using high end Green tubes without mods, you will find the neutralizing tab dangerously close to the plate. Also with the tab so close to the plate there is an enormous amount of feed back sent below the chassis disturbing stability. Without mods you will also find that the radio is no longer neutral on all bands. The mods will correct these problems.
Last note, a bit off subject --- the 6HF5, 6KD6, 6146B and many other tubes. Reviewing the experiences recorded by people who have had trouble with the these and various other tubes, I can't help but believe the 8122 and these tubes draw a parallel. The problems experienced are inherent in all beam tetrodes.
Tube data sheets are somewhat like speed limit signs. Range of operation specs are estimates at best for beam tetrodes.
There has been a plethora of mysterious highly volatile happenings when switching from the old original 6146 to the 6146B. With Hallicrafters transmitters using the original 6146, you would set the bias to a voltage depending the line input voltage achieving a consistent idle plate current. Using class AB1 mode resulted in a fairly consistent output. Then came the 6146B and a litany of legendary strangeness.
I believe tube designers facing competition for sales had to do something. How can you get more output from a 6146 running class AB1 mode? Lets make it look more like a triode, like the high bias 8122's. Offset the tube elements, broaden the operating range increasing the AB1 mode plate current and install a higher dissipation plate: a great way to increase the output of the tube by 30% in AB1 mode.
You install a pair of 6146B's in a transmitter that was designed for the original 6146 and unknowingly set the bias as defined in the manual to a specific bias voltage based on line input. You shortcut and neutralize them on one band. Squirrely things begin to happen.
If you check the idle current at the prescribed bias it will probably be much higher.
Second, you neutralized the radio on the 10 meter band assuming the other bands will also follow --- not likely.
On a multi band transmitter, the required neutralizing feed back is treated differently for each band. The neutralizing bridge is also designed for a specific geometry of a specific tube. This results in stability on all all bands, not some. That is the primary reason that Hallicrafters and others recommended that you buy their tubes for their radios --- not just profit.
With bridge redesign and mods, the SR-2000 will be stable on all bands even while using any of the Blue or Green Burle tubes as well as other vendor tubes with similar characteristics.
We are holding this spot to discuss the 6HF5 and 6KD6 tubes used in the SR-400 and SR-400A. There seems to be endless comments regarding these tubes that elude to instability, inadequate output, more out than expected, etc. These radios and the tubes used, have the same issues as the 8122 used in class AB1 radios; varying geometry. A new matched pair does not necessarily qualify them. Most will seem perfectly matched, and probably are but that tells you nothing about the geometry of the tube. The idle bias does. If you have a good pair of 6HF5's that test perfectly in your tube tester and need -75 volts of bias to idle in the SR-400 and a pair that needs -100 volts to idle, there is a difference. The -75 volt tube behaves more like a tetrode and the -100 volt tubes begin to mimic the triode. The -100 volt tube will deliver more output than the -75 volt tube but will also be less stable, given that both tubes are good. If you test a pair that bias at -50 volts, they likely will be stable but deliver less output.
The point here is all of these tube groups will provide the same output if a suitable class of service is used. Example: Using AB2, they will be much the same. Using class AB1 mode, the tube needing the higher idle bias will produce a higher output. It will also need more neutralizing feedback for stability.
Conclusion: If you are purchasing new balanced 8122 or 6HF5 tubes for your class AB1 transmitter, request tubes that meet your idle bias parameters. Numbers for other tubes will be published as testing is complete The 8122 bias range for install and no changes needed is -23 to -30 volts and -30 to -36 volts with changes to bias and the neutralizing bridge. The 6HF5's -75 to -85. After -85 volts stability suffers. The higher the bias needed the higher the output obtainable on all of these tubes in AB1 mode.
The following url http://k9axn.com/_mgxroot/page_10873.html Neutralizing concepts describes in depth easy to understand neutralization concepts.
If any part of this SA is not clear please send a note and I'll try to make it so. Thanks for your interest and good luck with your 8122 selections.
Kindest regards Jim K9AXN
Copyright James Liles K9AXN 12/18/2013