[Amps] Failure of a pair of FU728F

[Jim Garland]

Two years ago, I completed my first tetrode amplifier 
(http://www.w8zr.net/160amplifier/160overview.htm), a duo-band 
(80m/160m) design running three Russian GU74B tetrodes. The experience 
taught me a lot about the different design requirements between tetrode 
and triode amps.

 From this experience, I learned that tetrodes require a stiffly 
regulated screen voltage (340V for my amplifier), benefit from tiered 
control grid voltage regulation (different operating bias for SSB and 
CW/digital modes, and higher cutoff bias), and need more sophistication 
in their safety trip and flashover protection circuits, than do triodes. 
In other words, tetrode amplfiers are harder to design, build, and debug 
that their triode equivalents, although this added complexity is offset 
somewhat by the simplicity of tetrode's passive grid-driven input 
circuitry. (Triode amps normally require bandswitched tuned input 
circuits, while tetrode amps use a simple 50 ohm resistive input.)

Despite the difference in design complexity between triode and tetrode 
amplifiers, however, their tuneup and operation are rather similar. 
During tuning, one has to repetitively dip the plate current and 
increase the loading, while slowly increasing the drive level, all the 
time keeping an eye on the grid current and, in the case of tetrode 
amplifiers, the screen current. For a tetrode amp, the screen current is 
usually a more sensitive indicator of proper tuneup than the grid 
current. Tetrode amplifiers normally run in class AB1 (zero grid 
current) and AB2 (a few mA of grid current). Since the grid current is 
usually about zero, there really isn't much point in monitoring it, 
except to see that it doesn't soar too high because of underloading or 
overdriving.

The important point in monitoring screen current is to keep it modestly 
positive when the amplifier is tuned for its rated key down maximum 
output. The screen current will then go negative if the drive power is 
reduced; a negative screen current is normal and doesn't signify 
mistuning. In my amplifier, I monitor the screen current with an LED, 
but don't measure it on a panel meter. The LED flickers red when the 
screen current goes positive, gets brighter as the current increases, 
and trips the amplifier off-line when it exceeds 50 mA.

It is possible to tune a tetrode amplfier at lower power levels than its 
full-bore rating, but I don't do that because it's too easy to overdrive 
the amp and exceed the screen and grid current ratings. The tuning just 
becomes too finicky. In my opinion, it's better always to tune the 
amplifier at its maximum design rating, and then reduce the drive to 
scale back the output power. (By the way, this recommendation also 
pertains to triode amps.) Doing so will assure that the amplifier is 
being operated with the proper tank circuit Q and plate load impedance. 
The downside is that the amplifier efficiency is lower when the drive is 
reduced, than it would be if it were tuned for maximum output at a lower 
drive level. In other words, my GU74B amplifier generates more heat if I 
tune it for 3kW output, and then reduce drive to get 1.5KW, than it 
would if I tuned it for maximum power output at 1.5KW to start with.To 
me, this added heat is a small price to pay to assure myself that the 
amp is running optimally with respect to all parameters, producing the 
cleanest signal, with large safety margins for the screen and control grids.

73,

Jim W8ZR


On 9/20/2019 8:36 AM, Ron Youvan wrote:
>   John KK9A, john at kk9a.com wrote:
>
>> I believe that the tuning of a tetrode is different than a triode - 
>> correct?  Are they tetrodes less reliable or more finicky?
>
>   Tetrodes, if operated properly are a rugged as a tetrode, but they are
> more sensitive to heater power, G1 drive power and screen current so
> more care is required, in the design and operation, but gain and other
> operating parameters are better.