[ARC5] Smart People: Triode Connecting Pentodes

[Kenneth G. Gordon]

On 12 Nov 2013 at 14:08, David Stinson wrote:

> I've been told that the screen should be directly connected to the
> plate through a parasitic suppressor (100 ohms)

That cannot possibly be correct for a power tube. Receiving tubes, maybe: power tubes, no.

> and that the
> screen area is effectively added to the plate.

As I said, for receiving tubes, possibly. In fact, that is done in some receivers, the BC-
779/SP-200 for one in the audio driver and audio output stages. All three tubes are 6F6s with 
the screens connected directly to the plates. The plate voltage is something like 375 VDC, 
while the screens are rated at only 250 VDC, but those receivers ran that way for many years 
with no ill effects.

> These are running at 500 volts on the plate at about 100 mils,
> which is well within the tube *plate* ratings.

OK.

> However, the screen is max rated at 250 volts 12 mA, 
> or 3W max input.  My source says this isn't an issue 
> when the screen is connected to the plate, but I'm not so sure.

Good on ya: that cannot possibly be correct. Have you measured the screen voltage while 
the tube is in operation? I am certain you will find the screen voltage equal to the plate 
voltage, with a consequent vast increase in screen current. Not good.

> The 7984s work OK for the first few minutes, then start folding on me.

Why am I not surprised?

> I'm betting the screens are going.

How could they NOT?
   
> Who are the experts in this, and what are your points of view?

I am most certainly no expert. However, it stands to reason that with only a 100 ohm 
"parasitic suppressor" to limit screen voltage/current, the screens will be operating well 
beyond their specs.

Can you see "color" on the screens? If so, they are over-dissipating.

What I would do is to use what-ever resistance is needed, in this case around 20K ohms (if I 
have my calculations correct) to bring the screen voltage down to nearer normal, then bypass 
that resistor with a paralleled "large" capacitance to make sure the screen is at the same RF 
potential as the plate. Probably a 0.01 mfd disk ceramic would do the trick, keeping leads as 
short as possible.

Others here who are far more knowledgeable than I will most certainly correct me. :-)

BTW, we have still not clarified which of the two methods of "triode-connecting" a pentode 
would be best in your particular differing applications: 1) connecting the screen to the plate, 
or 2) connecting the screen to the grid (G1). Both result in a triode-connected tube, but 
according to Bill W6SAI 1) results in a low-mu triode, while 2) results in a high-mu triode.

It seems to me that in any case, the high-mu connection would be best for oscillator use.

Bill also does not recommend the low-mu connection if the tube is used in grounded-grid 
circuits since the screen dissipation will be "...too high...", especially for certain types of 
tubes. In addition, the GRID current of tubes like the 4-65 and 4X150A can be as much as 
several AMPs when those are used in the screen-to-plate grounded-grid connection.

Ken W7EKB