[Collins] 30S-1

[Dr. Gerald N. Johnson]

On Tue, 2007-09-11 at 16:32 -0400, jeremy-ca wrote:
> The 4CX1000A has a zero Ig  rating. As such it was too fragile for some 
> applications and the failure rate was high. In addition the IMD was poor 
> when run at close to its maximum ratings.

And that 0 grid current rating also made it not suited for class AB2, B,
or C operation with their associated higher plate efficiencies. But it
probably worked better at 100 MHz than the 4-1000A (or a pair of 4-400A)
for FM broadcast at about 1 KW output. And surely was easier to make
work than the glass tubes.

Anyway, the original 4CX1000A data sheet claims it has good intermod
characteristics, and only shows RF and audio linear service
applications.
> 
> The 4CX1500B surprisingly has a lower Ip maximum. It has a new control grid 
> with a 1W rating; it requires a different Eg than the 1000A for the same 
> level of AB2 operation. Screen requirements are different also. Typical 
> operation runs a EG2 of 225V vs 325 for the 1000A. IMD is substantially 
> better and equals the best of the triodes.

Most of the 4CX1500B and 4CX1000A data sheets I have read the same. Some
numbers are different, but most of the applications text is exactly the
same.
> 
> Obviously it requires a change to both grid and screen voltages in order to 
> take full benefit of the improved IMD.
> 
> Ive run both tubes on 2M and performed detailed IMD tests. The differences 
> when each is setup properly is substantial and in the 15dB range.
> 
> I don't know the blower rating of the 30S1 but going from 1000 to 1500W 
> dissipation requires a big jump from 18 cfm @ .23" drop to 34 cfm @ .60" 
> drop. Thats a huge difference in order to keep the seal temperatures within 
> rating. Its probably of no consequence on CW/SSB but RTTY and digital mode 
> operators should be aware.

More than a few early Eimac external anode tubes (4X150 for example)
wouldn't stay within their temperature ratings with any practical amount
of air flow. Those were the ones with the smooth sinusoidal inner fin
that had no louvers. After Collins mechanical engineers pointed that
problem out the anode coolers with louvered inner fins were introduced
and then the 4CX150 came up with 250 watts plate dissipation with
different flow and pressure drops like the 4CX260B. The pressure drop
surely increased with the louvered inner fins for even the same air
flow. Back to the 4CX1000A and 4CX1500B data sheets, the 4CX1500B will
dissipate 1 KW with essentially the same air flow as the 4CX1000A but
needs a significant increase to dissipate 1500 watts. One might infer
from that that the 4CX1000A would stay at a reasonable temperature with
that increased air flow too while dissipation 1500 watts. At that point
it may be good that its a breech lock socket, otherwise the air pressure
may tend to blow the tube out of the socket.
> 
> Carl
> KM1H
> 
> 
When the reports are that that 4CX1000A didn't do so good on intermod
while the data sheet proclaims its super for IMD, and that the 4X150
didn't have 150 watts dissipation and the 4CV100,000C had to have the
drive pushing the screen current to the dissipation limit to make 125 KW
per tube carrier in AM service (while checking the tube performance with
the Eimac computing tools left it a couple KW shy on output), these
events show some evidence that Eimac data sheets might have been a bit
optimistic at times for tube performance. I know in the 4X150 and
4CV100,000C that Collins was involved and we shipped many a 4CV100,000C
back with far fewer than rated hours but all crapped out. Supposed to
get 250 KW from a pair, the data sheet showed that, but the computations
from the characteristic curves came up short and we needed a few more
for output network and harmonic filter losses. Eventually newer tubes
arrived (with the same numbers) with increased screen dissipation
ratings. Must have been some of those sent back had melted down screens.

Between the tubes almost good enough and the vacuum variables almost
good enough and getting worse from shipping impacts we took a year
longer than the contract in developing the 821A-1. We had to teach Eimac
and Jennings how to make those parts by burning them up, the largest
each had ever made, at the time.

It was an interesting tube application, with 12.5 KV DC on the plates
modulated to twice that and with the RF peaks, the electrons saw better
than 50 KV acceleration voltage at those peaks and over the distance
they traveled they reached a velocity of 0.1C. The tubes I saw apart,
had dents in the copper anodes from the collision of those speedy
electrons with the anode surface. There was a clear shadow of the
screen.

Our trouble was that the tube selection was limited and Art wouldn't
hear of using tubes from other makers even if other maker's tubes
actually worked.

The latest Collins magazine says those 821A-1 cost 1/2 million each? I
recall the selling price as contracted was more like $120,000 each.
About the purchase parts cost. We spent more than $250K each in
engineering time and I don't know how much in production though
eventually all the production hardware passed through my hands for final
test, updates, and preparation for delivery. Ya, 1/2 million each was
about right for cost. Too bad for Art that he gave them away to break
into a new power level...
-- 
73, Jerry, K0CQ, Technical Advisor to the CRA
All content copyright Dr. Gerald N. Johnson, electrical engineer