While many or most tube failures are caused by age, there have also been bad
tube manufacturing runs over periods of time. More by far with 8877’s than any
other tube. While the 8877 is a good tube with excellent IM characteristics and
very high gain, it is probably the more problem plagued tube Eimac manufactures.
The failures and problems are actually why MRI equipment in the 80’s and
early 90’s quickly moved away from 8877’s. Field failures in 8877’s are a
major cause of ETO losing a contract manufacturing medical amplifiers.
While one west coast amateur blames all failures on amplifier design issues
(and oddly enough sells what he says is the only cure for problems), the truth
is power grid tubes have multiple causes of failure, virtually none of which
have anything to so with parasitics. Like it or not, tubes are very difficult to
manufacture. No surprise, they are a very common point of failure even with
conservative operation. The fortunate thing is when you get a good tube, as long
as you watch grid current, the tube last a very long time. One thing you can do
to hurt the tube in a very short time is to run excessive grid current.
Excessive grid current is bad news.
Newer 8877 tubes from San Carlos have had very few problems compared to older
tubes from Salt Lake City. Bad runs of 8877’s included almost the entire year of
1987 and sporadic smaller batches since then. I have put brand new 8877’s in my
amplifiers, had the tube fail within a year, and without making any other
changes except the tube had a tube from another production run last several
years of daily use. Some runs, like it or not, are not as good as subsequent or
preceding runs.
The main production problem with 8877’s was alignment of the grid and cathode.
In the late 1980’s, a heat-dam problem caused the cathode to move. The tantalum
barrier that serves as a support and as a gettering mechanism would shift and
misalign the grid, allowing it to lay against the cathode. During one period of
time nearly 90% of new tubes arriving at Ameritron would fail during 24 hour
heater cycling tests, and that was without anode voltage applied!
Another common grid problem is misalignment of grid wires in the tube. The
cathode emission bands of the 8877 are deposited in layered rings or stripes.
There are about 140 bands around the circumference of the short but large
diameter cathode. Between each electron emitter band, Eimac positions a very
thin gold plated grid wire. The function of the grid wire is to create an
electrostatic field. This field controls anode current, and the placement of the
grid wire outside the electron stream means very few electrons hit the grid. The
result of this is very little grid current flows in normal operation, and the
tube has very low distortion and very high gain.
If the grid or even if a few wires in the grid are misaligned, or if the grid
is allowed to go significantly positive during operation (high grid current),
high velocity electrons will strike the grid. High velocity electrons have
a large amount of kinetic energy. When these electrons strike a grid wire,
minute amounts of the gold grid coating can be released or “evaporated”.
This slow deterioration can occur from kinetic energy of electrons striking the
gold even if the grid does not become excessively hot. Even a few thousands of
an inch alignment error compromises long term life of the tube!
The gold eventually re-deposits. It often collects on cool areas, where it can
cause an arc path across insulation in the tube. The result is a severe HV arc
from anode to grid or other elements. Sometimes the arc will clear the fault
path and the tube will function again, but once the problem starts the tube will
generally slowly get worse until it eventually becomes unusable.
Second less likely and less destructive failures are G/K shorts caused by poor
cathode/grid alignment or caused on rarer occasions by conductive debris in the
tube. This failure generally is NOT accompanied by an arc.
To obtain maximum life from an 8877:
Do not cycle the heater excessively
Do not run more than 50 mA longer term average grid current
Do not exceed 150 mA grid current even for very brief periods
Do not apply HV or draw cathode current unless the heater has been within the factory specified range for 3 full minutes. If the heater
is started very slow, the tube warm-up time should be extended
Never run the tube when heater voltage is
low. Doing so will poison the cathode
Never reduce filament voltage outside of specified allowed voltage range
at the tube pins in an effort to extend emission life
Vacuum Tubes
since December 15 2004
|