Amps
[Top] [All Lists]

[Amps] ?negative screen current

To: amps@contesting.com
Subject: [Amps] ?negative screen current
From: John Lyles <jtml@losalamos.com>
Reply-to: jtml@vla.com
Date: Tue, 31 Dec 2013 12:47:44 -0700
List-post: <amps@contesting.com">mailto:amps@contesting.com>
WA4LAV's explanation agrees with what I have read and heard talking to tube manufacturers. A lot of information can be gleaned from papers in the archives of IRE, IEEE, AIEE, Wescon meetings, and so forth, but it isn't all in one text book.

In high power tubes, sometimes the name radial beam power tetrode is used whereas we know that the beam power tubes like 807, 6L6, and 6146 all use beam forming plates that schematically look like suppressor grids in a pentode. In these tubes the elements are in rectilinear arrays, and the cathode/heater is a wire wound around vertical bars that emits a cloud of electrons off the broad sides. However, cutting open high power tubes shows that the grids are all cylindrical meshes where electrons are radially supplied from a central filament or cathode basket or cylinder. In some designs there is deliberate effort to align the grids on top of one another as Bill said. These tubes have high performance, and low interception by grids lowers their thermal dissipation too.

But second order effects are further described by beam optic calculations, and there are some conditions of voltages on G2 and anode that can cause beam spread or even focusing into spots on the anode. These are typically not at the desired operating point, but may be swept through during sinusoidal voltages on the elements. And lets face it, in big tetrodes, there is no such thing as exactly zero RF on a screen grid, due to lead inductance, no matter what bypassing is applied.

I once had a condition where we were loosing 4CW250,000 tetrodes regularly, they changed their operating point over time and would start to loose control (grid) voltage.They acted like they were gassy. An old guy (SK, former Eimac engineer) at Econco cut one open and showed me that there was warpage of the screen grid at various places, each one corresponding to a mark on the inside of the anode. Here electrons were focusing into spots and causing release of secondaries as well as copper ions, that were flying back to the grids. This change caused the grids to run hotter and eventually those spots buckled and warped. I believe it all started from non-perfect symmetry in the tube construction, and a bad combination of voltages. We solved our problem by raising the cutoff bias from -350 to -600 VDC as there was a tiny (mA) beam there all the time with inadequate cutoff. With adequate cutoff, there was no reason for the beams to cut copper.

RCA, in their glory days, designed a whole series of very interesting high power tubes with unit triodes and unit tetrodes. These had little single filament line arrays that would get constructed and stacked in parallel in radial array around the tube. They were able to align the grids by construction this way. The 4616 tetrode had the filaments on the outside, beam pointing inward to a cylindrical copper and tungsten anode core. The super power triodes like 7835, 4617 and A2346 have it reversed, central filament bars and anode is an outer cylinder. Each filament in the 7835 runs about 70 amps DC and there are 96 unit triodes in parallel, to form a megawatt class tube.

They also had what they called shielded beam tubes, which are a little more complex than the triodes, but they enclosed the little elements with a copper side wall and proper alignment to reduce screen interception to very low amounts.

In the big Thales tetrodes I mentioned in prior email (having pyrolytic graphite) the screen and control grid meshes are carefully constructed and aligned so that they are in the shadow of each other. Amazing that all this is made by chemical vapor deposition and then cut with computer controlled machinery or lasers. Tubes have some a LONG way since Deforest and Fleming, long live 'em!


 Message: 1
Date: Tue, 31 Dec 2013 03:16:24 +0000
From: "Fuqua, Bill L" <wlfuqu00@uky.edu>
To: Peter Voelpel <dj7ww@t-online.de>, "amps@contesting.com"
        <amps@contesting.com>
Subject: Re: [Amps] ?negative screen current
Message-ID:
        <B7E8B5B4A202074084E2515A7B10A7F32D68EBCF@ex10mb02.ad.uky.edu>
Content-Type: text/plain; charset="Windows-1252"

   You have to remember that Wikipedia consist of contributions from people 
that don't always have all the
facts.
   In most tetrodes the incident electrons that strike the screen grid far out 
number the secondaries that
get captured by the plate. However, in some tetrodes there is an attempt to put 
the screen grid in the
shadow of the control grid so that little or no electrons strike it. However, 
under some combinations of
plate voltage and screen voltage some electrons do strike it but not right on, 
you can imagine the shadow
closing in on the screen element. When this happens multiple secondaries are 
produced that are not
recaptured by the screen and are attracted by the more positive plate.
   Now, whenever an electron strikes a metal object grid or plate it does 
produce secondaries if
it has enough kinetic energy. At several hundred Volts of accelerating 
potential multiple secondaries
are produced. It is just that most of the time these slow moving secondaries 
are recaptured by the
element that emitted them.
   It is a problem that I have seen with my transmitting tube tester under the 
right conditions and
has nothing to do with the dynatron region where the plate voltage is near or 
below the screen
voltage.
  73
Bill wa4lav


_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps

<Prev in Thread] Current Thread [Next in Thread>
  • [Amps] ?negative screen current, John Lyles <=