Neutralization is often overlooked in vintage transmitters. The legendary Collins 32V did not have it at all. The Johnson Valiant did. It was retrofitted to the Heathkit DX100 and Apache in the field and late production units. It can be added to the Johnson Viking II; I will show how as part of the restoration.

This article is a compendium of some correspondence that I have exchanged over the last couple years regarding this problem. I have referenced work by W8JI and others on successful repairs in this article. Full credit is given to the sources of the information shown here. I am just trying to get all the true and relevant facts assembled in one place, and eliminate the misinformation on these subjects (and there is PLENTY of it out there on the internet).

Lack of neutralization shows up in a number of ways. Transmitters that are on the verge of instability show radical swings in grid current while adjusting the plate tuning for plate current dip. Because there is only one display meter switched between plate and grid functions is the reason owners do not pick up on problem. You load the plate to resonance. Then, if you happen to recheck the grid drive, you find it is way off. Tuning the grid or driver and readjusting the drive control (if one is available) seems to correct the problem. Currents seem to drift as you operate for a time. If you have a highly reactive load, the plate current or grid current can increase to alarming levels, or even pin the meter and blow fuses or damage tank circuit components. Spurious out of band signals can occur when the RF amp is oscillating. If it is on the ragged edge, the problem can manifest itself only in transient conditions such as modulation or keying.

Modern HF Linears often neglect this important detail. Even the Collins 30L1 resorted to a unique scheme that was not unconditionally stable; quirky behaviour resulted. The Heathkit SB200 did not have it at all. The Heathkit Warrior had a very nice neutralization circuit (which will be shown when I complete that article on this page). The SB400 and SB100 families all had one or more neutralization circuits. A complete theoretical discussion of the problem of neutralization is available at: http://www.w8ji.com/neutralizing__amplifier.htm

Some tubes are more likely to self oscillate. The 6L6 often is used without the benefit of neutralization and such circuits rarely work well above 40 meters. One handbook design for two parallel 6L6s employed neutralization and worked well even on 15 meters. The 807 internal screening was much better. In a neutralized circuit, it was OK on 10 meters. The Harvey Wells Bandmaster used the 807 up to 2 meters, but the final was a multiplier circuit. When the input and output frequencies were different, the tendency to oscillate was less likely. That is why the Johnson Adventurer could work on 10 meters without neutralization. (It doubled to 10 meters from a 20 meter signal on the 807 grids.) The 6146 worked well as a straight through amplifier on 6 meters and could be used with reduced power on 2 meters with careful circuit design and lead dress. Neutralization can go a long way, but it cannot fully fix a crappy tube. While we are on the subject of crappy tubes, I need to state unequivocally that the 6146B is NOT one of them, as some have asserted. For the most part, the only advantage of using a B version is that you happen to have one on hand, or you want a little extra margin of plate dissipation. Loading the rig up to a higher output in AM mode will likely be of no advantage, because the rigs of this time barely made 100% modulation with a standard 6146. The current increase in the modulation transformer also had liabilities including core saturation and reduced low frequency response. I DO recommend 6146Bs in the modulator, with the modifications shown elsewhere on this page, to obtain class AB2 and full modulation, but neutralization is not an issue in the audio circuit.

The Valiant (which IS neutralized as part of the design), Viking II, DX-100, Apache and others will work just fine with a 6146B; but they will work better if neutralization is added. The DX-100 and Apache were retrofitted by the manufacturer to stabilize grid current on 15 and 10 meters. Neutralization does not drift if properly done, in an otherwise stable circuit. The tube characteristics are set in manufacturing by the gain and interelectrode capacitance, and the bridge circuit balances out those characteristics and are stable until the tube has aged for many hours till the end of its useful life. Certain Motorola commercial radios operating far above 30 MHz and certain Collins HF sideband rigs had problems with the 6146B as a replacement. You might find certain Yaesu compact HF transceivers do not have sufficient range of adjustment in the neutralizing circuit to accept any other than the exact replacement. Note that this series of Yaesu rigs sometimes used sweep tubes, and did not play nice with tubes from certain manufacturers; this owes to the fact that the power tube and its circuit was designed for 15 KHz switching, not RF service, and this was an "off label" use. These sweep tube miseries are in no way an indictment of the 6146B. Collins HF transceivers and transmitters often use an inverse feedback circuit wrapped around the driver and the finals; all goes well if everything is in resonance, since the phase shifts are predictable. Phase shift is the lynchpin of neutralization. If you insert any new phase shift, the bridge neutralization scheme falls apart, especially when off resonance by small but significant amounts - what part of "INVERSE" did the designer not understand? In this case, the design was too clever by half, grasping for slightly better intermod performance and sacrificing basic stability.

Parasitic oscillations are totally different from neutralization issues. Neutralization addresses self oscillation ON THE FREQUENCY OF OPERATION. Parasitic oscillations occur on frequencies OTHER THAN THE OPERATING FREQUENCY.

The RF amplifier can also have parasitic oscillations at low frequency from poor choice of RF chokes in the grid circuit. Having too high an inductance can cause spurious outputs at spacings from the main carrier on the order of KHz. This can happen with modulation or keying transients. A spectrum analyzer or well shielded receiver can detect this condition while transmitting into a dummy load. Any new transmitting system should be inspected this way before connecting it to an antenna. A signal report asks how the other ham is hearing your signal as he is tuned to it. It does not tell you anything about how other off your channel are hearing your buckshot. Neutralization does not fix this problem. It is a separate issue. Fortunately, this situation is relatively rare, and when it appears, it is easier to fix. Low frequency parasitics and distortion products are best tested on the air with a local ham friend who has a software defined radio (band scope function) or tunes the rest of the band in use while you are testing.

The RF amplifier can have parasitic oscillations at VHF or UHF owing to inductances in the grounding or bypassing. This is why in my article on the Johnson Adventurer, I shortened all the screen and cathode bypass capacitor leads and improved the grounding at the 807 socket. W8JI also discovers a problem in the Valiant owing to grounding of the Plate Tuning Capacitor, even though the Valiant IS neutralized. The resistor and concentric inductor included in the plate lead (and sometimes the grid lead) of an RF amplifier are put there to quench the tendency for VHF parasitic oscillation. The effect is to lower the frequency of the tendency to self oscillate by adding inductance to the plate circuit, while loading it with a resistor to reduce its Q. This reduces the frequency of the tendency to parasitic oscillation to the range where neutralization is effective. This VHF UHF parasitic problem is often mistaken for a neutralization problem. Neutralization fixes the tendency to oscillate on the operating frequency. It may help with VHF parasitics in some circumstances, but they are separate technical issues which must be addressed in tandem. Often the resistors in the VHF parasitic chokes change value due to heating and aging. When they increase in value, the circuit Q goes up. The plate capacitor often arcs, causing pitting that must be addressed. Worse yet, sometimes the band switch is destroyed. Occasionally the plate RF choke explodes noisily. Maybe the final RF tube fails for unknown reasons.

Reactive loads (high SWR) can cause a transmitter to take off from lack of neutralization or improper attention to VHF parasitics. That is why I use only RESONANT Coax fed antennas at my station. If the line needs the help of a transmatch to reduce SWR at the band edges, I perform all adjustments in RECEIVE mode using a MFJ-212 matchmaker. Think it about it: a misadjusted transmatch temporarily puts widely varying reactive loads on your transmitter. The most risky time for a transmitter is during tuneup. Your finals will thank you with long dependable service.