[MAMS] [Mw] tubes for the 6155 Re:Back On 222mhz with Amp

[Dr. Gerald N. Johnson]

In the 1965 "Single Sideband for the Radio Amateur" it begins on page 
135 and says it was from "Understanding Tetrode Screen Current," July 
1961 QST. The QST article is available for ARRL members at www.arrl.org 
where you can search by title.

In the Collins 821A-1 (250 KW carrier output AM and also linears the PA 
screen current was used for ALC to control the drive and to set loading 
when automatically tuning the PA. Screen current is super sensitive to 
drive and in large tubes, getting full AM power required driving them to 
just below causing the screens to glow. The 30S-1 uses a screen meter 
for the main manual tuning indicator. Its set up to allow reading 
positive and negative screen current.

Part of what happens when loading is on the light side (plate load R is 
high) the lowest plate voltage tries to dip below the screen voltage and 
the electron stream moves to the element with the highest voltage 
cutting plate current momentarily and momentarily increasing screen 
current. That can be destructive. Heavier loading reduces screen current.

Eventually RCA and Eimac 4CX150s were given a 250 watt plate dissipation 
rating, but the initial products were only rated at 150 watts and a 
mechanical engineer (K uykendahl)in the high power transmitter 
department at Collins told me they had found those tubes overheated if 
they dissipated 150 watts. Those were the tubes with the smooth foil 
between the anode and the outer plate cylinder. What it took was the 
foil cut with louvers to get adequate heat transfer. The 8930 I believe 
has essentially the same cathode, grid, and screen as a 4CX250 or 4CX300 
but more cooling fin area. The gains of the 4CX250 and 4CX300 are the 
same, but the input capacitance of the 4CX300 is higher making it look 
kind of like a 4CX350. The 4CX350 has the same anode size as the 250 but 
the control grid to cathode spacing is smaller giving it greater gain 
and makes it incapable of standing any significant grid current because 
the grid is already operating as hot as it can stand just from radiant 
heat. The input capacitance of the 4CX300 is higher because of the 
extended twist lock base makes for longer grid leads.

The cathode capability can be found in the peak plate current rating. 
Cathode heating power would also be close to the same in different tubes 
with the same peak cathode current capability. The grid and screen can 
be compared by the Gm of the tubes, having the same current gain means 
the grids are the same size spaced the same distance from the cathode. 
The plate current rating is based more on plate dissipation than the 
inner elements of the tube.

One could probably achieve greater power dissipation with greater air 
flow, though the rated air flow would almost push the tubes out of the 
socket and can be noisy unless the PA is relegated to a remote 
controlled location.

Even with greater plate dissipation the control and screen grids are 
still limited in dissipation and cooling the plate doesn't help them. We 
melted down many 4CV100,000C developing the 821A-1 because it was 
fundamentally a 4CX35,000A air cooled tube guts with a vapor phase 
cooled plate and the grids were not up to the AM task. Eventually Eimac 
learned how to make sturdier grids though it cost us a year or more of 
development time in a contract with a late delivery penalty clause. We 
were running those tubes up to 60 kV peak plate voltage at combined RF, 
DC and audio each RF cycle and modulation peak. Someone computed the 
electron velocity at 0.1C. I know in one tube that we opened up, we 
could feel the shadow of the aligned grids in the inside of the copper 
plate. The electron impact had dented the copper a few mils. Also 
darkened it. We were running screen dissipation to their absolute 
maximum and barely making rated power.

Then while VOA invoked the late delivery penalty to save money (and we 
had bid the transmitters at about our purchased parts cost giving them 
the cost of engineering and fabrication to break into that new higher 
power business) VOA stored all the transmitters for 3 or 4 years while 
they had to go to congress for money to expand their buildings to hold 
the transmitters. Collins resisted the penalties on the grounds that VOA 
wasn't hurt, actually saved them warehouse rent for that year, and that 
vendors Jennings and Eimac learned how to make better high power parts 
that would benefit competing transmitter makers in future bids. I took a 
leave of absence and haven't been back before that was resolved so I 
don't know any details of that argument.

The 821A-1 essentially gave VOA two transmitters for the space of one. 
Their specification was that it should take two technicians no longer 
than 20 minutes to change frequency. What we delivered took no longer 
than 16 seconds for a wide frequency excursion and less for smaller 
frequency changes with no operator skill required. Before the station 
break, pick the new oscillator and crystal, then dial in the output 
frequency with a gang of thumb wheel switches. At the beginning of the 
station break life the protector from the "Tune Start" button, push that 
button and it took care of itself first removing modulation, then the RF 
drive, then preset the servos for the new frequency, switched to the low 
power plate voltage (a mere 90 KW out), peaked the drive, tuned and 
loaded, then if it had been operating on high power restored the high 
power plate voltage and then reapplied audio.

There were a bunch of vacuum variable capacitors plus four of the big 
tubes and there was a long term problem in vacuum capacitor life that 
made operating costs higher than some from the competition. In 1964 each 
variable capacitor and each PA/Modulator tube cost about $550. My gross 
junior engineer salary was $600 a month.

73, Jerry, K0CQ

On 12/31/2012 8:00 PM, John T Rose wrote:
> I believe it was some one in the midwest who had machined adapters to put
> use a 4CX250 tube.  But that was a long time ago.  The 4CX400/GS36B works
> nicely in these amps, but don't know if they are still available.  You do
> have to carefully bend the anode fingers as the 4CX400 tube anodes are
> about 50 mils smaller.  They will put out significantly more power and
> are potentially cleaner.  The Amperex DX-393 is another tube that these
> amps could use.  I have used all three of these tubes.  One thing to
> watch is the regulated screen current range is not that wide.  So
> depending on the tube and how it is tuned the tube can try to draw more
> screen current then the supply is capable of.  I.E. OK tuned with a
> carrier, but tries to draw too much current with a SSB signal.   As I
> recall these were built to run 50 watts carrier output AM.  In the early
> 60's an Eimac engineer wrote an excellent article on screen current in
> tetrode amplifiers in QST.  Don't remember the issue.  Really good
> reading if you have not tuned a tetrode amplifier.
>
> 73 John, WW1Z
>
>