> Over the Last thirty years or so, I have designed and/or built literally
>dozens of power amps, both at home and on the job. These span the range of
>audio thru L-band, from a pair of 1625's (??) to a multi-KW L-band solid
>state job. Of all of these, my 4CX1000A amp has given me the most grief.
>It's been over ten years since the last time I rebuilt it, so I thought it
>might be fun to do it again. (Sure!)
> I have been going thru the last several months of list archives looking
>for clues, hints, revelations, etc, for passive grid tetrode amps hoping
>that I might find something that I overlooked in past design efforts. I am
>amazed at the amount of knowledge and experience (and EGO) possessed by the
>list members in general. In spite of this, I could not find answers to two
>points that are currently bothering me.
> First question: Many references have been made to grid damage in the
>4CX1000A and similar tubes; what are the detailed symptoms of said damage?
>(Assuming the tube still sorta operates)
The grid in the 8168/4cx1000A is barely visible to the naked eye.
Presumably, when the 0w [sic] grid dissipation rating is grossly
exceeded, the tiny grid wires melt.
> #2: I want to include NFB cathode resistance this time around. Looking
>at the Measures' and Daughters' versions, I see no mention of filament
>chokes. With the cathode either connected to the filament at the factory
>(4CX1000A) or by preference (4CX1600), I don't see how this can work.
- Because when RF cathode feedback is used, the peak potential between
the heater and the cathode is roughly no more than 25v, which is usually
well within the typical heater/cathode max. v rating. .
>filament is either bypassed with caps or thru the filament transformer
>windings and therefore the cathode resistance is bypassed as well.
Only if the heater is connected to the cathode, would this present a
problem. However, little L would be called for if the cathode were
wired to the heater. .
>missing something here? If a filament choke is required as I suspect, then
>that is yet another reason why one should never consider a tetrode for a HF
>amp unless one gets one with handles for FREE!
Multi band tetrodes-with-handles amps are somewhat easier to build than a
g-g buffer amp. driving a g-g amp. With a TWH amp., one can drive the
beast directly with a transceiver.
- Some profess that tetrode amps are overly complex because they need a
grid supply and a screen supply. However, thanks to modern power FETs,
an electronic-switched grid bias supply is simple to build, as is an
electronically-regulated screen supply. Examples of both circuits are
on my Web site at:
- In *Amplifiers* there is an explanation of what I believe are
worthwhile design considerations in Class AB1 amplifiers. Two of them
are selecting the appropriate type of screen supply (to avoid possible
destruction of the screen), and assuring that there is minimal L between
the grid termination R and the grid (to minimize the possibility of VHF
- In a tetrode amp., the tuned input circuit basically consists of a
small roller coil, an RF toroidal, Z step up transformer, and a grid
termination R. However, in the above example of a g-g amplifier driving
a g-g amplifier, each amplifier needs a tuned input for 160m, 80m, 75m,
40m, 30m, 20m, 17m, 15m, 12m and 10m, for a grand total of 20 tuned input
pi-networks with a Q of roughly 2 -- in other words, relatively speaking,
a ponderous pain in the posterior.
R. L. Measures, 805-386-3734, AG6K
FAQ on WWW: http://www.contesting.com/ampfaq.html
Administrative requests: amps-REQUEST@contesting.com