Improved Anode-Circuit Parasitic-Suppression For Modern Amplifier-Tubes
>> condensed version of the October 1988 QST Magazine article with updates. // page 6a/6b of 7
All HF-amplifiers must have electrical conductors to connect the anode-connection, or "plate-cap", on top of the amplifier-tube, the DC blocking capacitor, and the tune-capacitor [tune-C]. I collectively call these conductors and the capacitances associated with them, the anode-circuit to avoid confusing this circuit with the tank-circuit.
All conductors naturally contain inductance. In a typical amateur radio amplifier, the total inductance in the anode-circuit is roughly 0.2µH [200nH]
Since all amplifier-tubes contain output-capacitance, the combination of anode-circuit inductance, tune-C and tube's output capacitance [Cout] forms a VHF resonant circuit in the anode-circuit. See Figures 1A and 1B.
If you don't see the parallel resonant circuit, consider that Cout and the tune-C are connected in series via their common chassis connection. These series connected capacitances are connected in parallel with the anode-circuit inductance. This forms a parallel-resonant circuit.
In a typical 1.8MHz to 30MHz, 1500W amplifier, the self-resonant frequency of the anode-circuit is usually between 75MHz and 150MHz. Amplifier input circuits also contain inductance and tube (input) capacitance which forms another VHF resonant circuit. If the input bandswitch is located some distance from the cathode, a length of coaxial cable is used to connect them. This coax contains VHF resonances that&emdash;depending on its length&emdash;may be problematic.
And so, all HF/MF amplifier output sections contain two resonant circuits: 1. An HF-resonant tank-circuit between the tune-C and the load-C, and 2. A VHF-resonant circuit between the tube and the tune-C.
Amplifiers are routinely subjected to numerous, switching, keying, and voice transient-currents. These transient-currents must pass through the VHF self-resonant anode-circuit. Each transient-current causes the self-resonant circuit to ring and generate a damped-wave, VHF-voltage that is proportional to the VHF "Q" of the circuit. This process is like striking a bell with a hammer. This is the same principle behind a spark transmitter wherein a rotary spark gap provides the current transients to ring a high Q L/C RF-resonant circuit that is coupled to an antenna. Many kW of semi-coherent RF can be produced by a spark transmitter.
In an HF-amplifier, some of this shock-excited VHF voltage is fed-back to the cathode input of the amplifier-tube by the feedthrough-capacitance inside the tube. The VHF signal is then amplified by the amplifier-tube. The amplified VHF voltage returns to the VHF resonant anode-circuit where it started. Some of this now-larger signal will be fed back again, by the tube's feedthrough-capacitance, to the input, where it will be reamplified. This is called regeneration or oscillation.
Oscillation is like a snow-avalanche. It takes just the right conditions and shock-transient to start the process in motion. It starts small, but it gets out of control quickly. The amplifier becomes a VHF power-oscillator in microseconds.
There would be no problem if the VHF-power could be dissipated in a load. Unfortunately, there is no way that the large amount of VHF energy that is generated can escape because the HF tank-circuit is essentially a low-pass filter. And so, the trapped VHF energy runs amok in the amplifier. This can cause arcing in the bandswitch and/or the tuning-capacitor.
A VHF parasitic-oscillation can also cause a large current-pulse to surge through the amplifier. Since the amplifier tube is unloaded during the intermittent VHF parasitic-oscillation, the resulting grid-current is always excessive. Amplifier components are often damaged by this current-pulse. The current-pulse, which is fed by the stored energy in the HV filter-capacitors, creates a powerful magnetic-field that can pull an amplifier-tube's hot, thoriated-tungsten filament-wires off-center. This can result in a grid to filament short in the amplifier-tube. The large current-pulse can also damage the amplifier's current metering circuits, Zener bias diode, and other components.
For indirectly-heated cathode tubes such as the 8877 and its cousins, the pulse of VHF grid-current can be so large that the gold-plating may become hot enough to evaporate some of the gold into the vacuum. The hot, mobile, gold vapour-cloud may condense on the somewhat cooler inside of the ceramic anode-insulator [now a semi-insulator] or it may condense on the cathode's oxide-coating&emdash;whose electron-emitting ability is poisoned by the gold. Gold vapour may cause an arc between the anode and the grid and an arc between the cathode and the grid.
A grid to cathode arc will strip away some of the cathode coating. A [grounded] grid to anode arc can drive the HV negative circuits, which includes the tube's cathode, to many kV above chassis-ground. This may cause the stored energy in the HV filter capacitors to discharge through the often grounded filament of the tube. This can burn out the filament in an 8877.
Gold sputter damage is often fatal. If the sputtering is minor, and no arcing occurs, it will reduce the electron emission from the now poisoned cathode which results in lower power output and reduced anode-current. Testing for gold sputtering in a cold tube can done with a high-voltage breakdown tester using about 3-times the rated anode supply voltage for the tube. The leakage current in a healthy tube will be ¾5µA with positive and negative anode voltages. If loose gold is present, the positive {anode} leakage current will be much higher than the negative leakage current. The leakage current can usually be reduced by banging on the anode, with the anode up, with a screwdriver handle to cause the particles of loose gold to fall from the anode insulator down to the bottom of the tube. If the tube is inverted after this is done, the banging must be repeated with the anode up.
The September and October 1990 issues of QST Magazine contain some pictures of the insides of amplifier-tubes and other amplifier parts which were damaged by parasitics. The article is titled "Parasitics Revisited".
Warning Signs
A common precursor to a full-blown parasitic-oscillation is an intermittent arcing sound coming from inside the RF-compartment. This is the least expensive time to "pull the plug" and take care of the instability problem.
A bang or pop usually indicates that a full-blown intermittent VHF parasitic-oscillation, and its accompanying large current-pulse, has occurred. This is usually bad-news for amplifier parts. Not all amplifiers make a loud noise during a destructive intermittent VHF parasitic-oscillation. For some, yet to be understood, reason, 8874, 8877 and 3CX800A7 amplifiers may make only a slight pop when they sustain a major intermittent VHF parasitic-oscillation. If you suspect that this has happened, a test for loose gold is in order.
How A Parasitic-Suppressor Increases Stability:
A parasitic-suppressor must perform two, related jobs. The first job is to reduce the flywheel-effect of the VHF self-resonant circuit by reducing or dampening the Q of that circuit. Flywheel-effect [Q] is like the ringing-ability of a bell. For example, if a hand is placed on a bell, the vibrations will be dampened and the bell's oscillating or ringing-ability and sound amplitude will be reduced. Flywheel/ringing-effect is essential to any type of oscillation. Reducing or dampening the flywheel-effect [Q] will reduce the amplitude of the VHF-voltage that is generated which reduces the chance of an intermittent VHF parasitic-oscillation.
The second job of a suppressor is to reduce the VHF voltage-gain of the amplifier stage. The voltage-gain of an amplifier-tube is approximately proportional to the output load-resistance (RL) placed on the amplifier-tube. High-RL means high voltage-gain and low RL means low voltage-gain. ...33 If the VHF voltage-gain of an amplifier-tube can be made low enough, by decreasing the VHF RL, the amplifier will not be able to oscillate.
If an good conductor, such as copper or silver, is used to connect the anode of the amplifier-tube to the tune-C, a high VHF-Q parallel-resonant circuit will be formed by the conductor's high-Q inductance and the tube's Cout&emdash;which is a very high Q capacitor.
A high-Q parallel-resonant circuit has a high-resistance at its resonant frequency. Thus, the amplifier-tube is presented with a high VHF-RL which causes a high voltage-gain at the VHF resonant frequency of the anode-circuit. This increases the risk of a VHF parasitic-oscillation. See Figure 1,C.
A low-Q parallel-resonant circuit has a relatively low-resistance at its resonant frequency. Thus, the amplifier has a low VHF-RL which produces low VHF voltage gain&emdash;thereby reducing the risk of a VHF parasitic-oscillation.
¿Why does a parasitic-suppressor use a coil in parallel with a resistor? An IF-transformer can be broadbanded by stagger-tuning the primary and secondary r
esonant frequencies of the transformer. This lowers the Q which increases the bandwidth. A parasitic-suppressor works the same way. It has two conductors in parallel with different amounts of inductance. The coil is the higher-inductance path and the resistor is the lower-inductance path. If the magnetic fields from these two conductors do not couple, a staggered resonance effect is added to the anode-circuit. This lowers the VHF-Q. You could say that the two resonant frequencies in the anode circuit work against each other. This effect can be observed by judicious use of a dipmeter as outlined in the General Instructions for low VHF-Q parasitic-suppressor retrofit-kits.
To sum it all up: The goal of VHF parasitic-suppression is to reduce the VHF gain of the amplifier so that it will not have enough VHF gain to be able to oscillate. To accomplish this, we lower the VHF RL by decreasing the VHF Q of the anode-circuit.
Anti VHF-Parasitic Techniques:
The obvious way to decrease Q is to use low Q conductors. In other words, since Q=X‡/R‡, we need to increase R. Nichrome has about 60 times more resistance than does copper or silver. Constructing VHF suppressor-inductors with nichrome-wire instead of copper, or silver plated copper substantially reduces the VHF-Q which improves amplifier stability.
For an even lower anode-circuit VHF-Q, nichrome can be used for all of the wiring in the anode-circuit of an amplifier. The size of anode-circuit conductors should be no larger than is necessary to carry the actual current present without overheating. Wide, "heavy-duty" anode-circuit conductors &emdash;even those made from nichrome&emdash; will increase the VHF-Q, increase the VHF voltage-gain and increase the amplifier's ability to oscillate.
Nichrome can not be used for any of the conductors in the HF tank-circuit because of the presence of the large, HF-circulating-current. Tank-circuit conductors should be made of copper. At HF, silver-plating the copper provides a cosmetic improvement and corrosion resistance to humid air without the presence of sulfur compounds but no real improvement in RF efficiency.
The VHF gain of an amplifier can also be reduced by providing a RLC VHF series-resonant, low-Q, load path to chassis-ground and the cathode-terminal on the tube-socket. In other words, this device VHF- 'swamps' [loads] the input. Typical values for this device are ‰25pF in series with ‰10‡. The L is furnished by the inherent inductance in the wires, the tube, the socket and in the R and C components. This type of suppressor can not be tuned with a dipmeter, so the value of C must be determined by making an educated guess and by trial and error. This type of suppressor is especially useful for amplifier-tubes that have a large amount of feedback-C such as the 3CX1200A7.
Parasitic-oscillations and earthquakes have one thing in common: unpredictability. It takes just the right transient current plus a high-gain amplifier-tube to start an intermittent VHF parasitic-oscillation. Some amplifiers may oscillate only once in 5 years. Other amplifiers may never oscillate unless a higher-gain tube is installed in the amplifier. A new amplifier that uses traditional, high-VHF-Q "parasitic-suppressors" may oscillate the first day it is used&emdash;or it may operate for years without trouble.
Nickel-Chromium Alloys
There are at least 10, different alloys that contain nickel and chromium that are called "nichrome." Besides nickel [Ni] and chromium [Cr], some types of nichrome may also contain silicon [Si], cobalt [Co]. iron [Fe], manganese [Mn], aluminum [Al] or copper [Cu]. Each type of nichrome alloy is designed for a specific job. All nichrome alloys are semi-magnetic and they will respond very slightly to a magnet whether or not they contain iron.
The basic nichrome alloy is Nichrome-80 which is also referred to as "pure nichrome" by one manufacturer. It contains 79% Ni, 20% Cr, and 1% Si. The 1/4 inch [6.3mm] ribbon in some of the suppressor retrofit-kits is made from Nichrome-80. The ribbon is normally used for making flexible links in the anode leads. The ribbon has a much higher VHF-Q than the wire, so it should not be used for anode-circuit leads unless the wire can not handle the current burden.
The RF resistance of nichrome can be enhanced by adding Fe to the Ni and Cr. The #18 [1mm] wire in the suppressor retrofit-kit is made from 59% Ni, 24% Fe, 16% Cr and 1% Si. This material is referred to as Nichrome-60 by Harris Alloys, Inc.
For building low VHF-Q circuits, Nichrome-60 is ideal. The RF-resistance of Nichrome-60 accelerates as frequency increases due to the added ferromagnetic loss caused by the iron in the alloy. The loss mechanism appears to be similar to what happens in a VHF-UHF attenuator-rated ferrite-bead.
Nichrome-60 is probably the best conductor material available for constructing effective VHF parasitic-suppressors for use in MF/HF-amplifiers. In a nutshell, it is the worst conductor material at VHF&emdash;that's solderable&emdash;that I've found.
Looking Back
proverb: Most of what's "new" is what we forgot.
After the final edit of the Sept/Oct 1990 2-part QST article "Parasitics Revisited" had been 'locked up', I was talking with Dave Newkirk, WJ1Z. Dave edited these articles. He told me that he got curious and he wanted to see if there was any information on building an improved VHF parasitic suppressor in some of the old ARRL publications. In the 1926-edition of The Radio Amateur's Handbook, on page 72, there's something worth reading. Here's what it says regarding building VHF suppressors with resistance-wire [nichrome], instead of copper wire:
"The combination of both resistance and inductance is very effective in limiting parasitic oscillations to a negligible value of current."
/ Rich, AG6K