[SixClub] MY NEW AMPLIFIER

[Mike Andrews]

On Mon, Feb 13, 2012 at 04:26:20PM -0500, Pete wrote:
> LMFAOAPOM!
> Best thread in months.
> -Pete / NL7XM

Glad you're enjoying it, OM. 

Now, if you want to go _seriously_ QRO, John Lyles has a rig at Los Alamos
which is good for something like 2MW peak at 200 MHz, using a Thales TH628
diacrode, and has written about it on the "amps" list and elsewhere. This
critter looks like an oversized stock-pot in an Army chow hall, and needs
something like 325A at 9.5VDC to light the heaters. The plate dissipation
is "only" 2 MW.

Here is some of what he has written about it: 

"This is the final power, the driver is 14 kV at 25 amps. I don't have a
key for it, only a PLC command on a screen to energize it. Cannot send
morse with it, not fast enough, unless i key a pin diode on the front
end. There are no off the shelf dummy loads, no. The load is 16 foot
long coaxial pipe load that was commercial years ago. Instead of sodium
nitrite and water, i am using a corrosion inhibitor for loss and deionized
water. Return loss is -30 dB over a range of temperatures. It has closed
loop cooling through a heat exchanger, about 60 gallons a minute flow.
This transfers heat to the main cooling for the building, which is about
1000 GPM. I am using around 200 for the load, the IPA and the final plate
cooler. Its not a cheap rig. When I get some photos of the finished
operation and have them released for public, will post a link for look-see.

At VHF with a cavity, parasitics are no longer VHF but are likely to be
L band TE modes. So the suppressor is a different beast, not coil over
resistors!"

...

"For the prototype power amplifier I am preparing to test in a few weeks,
we just installed the filament cables. They are 1100 MCM cables, that
are as thick as your wrist. They are 15 feet long to the power supply,
which fills one equipment rack (20 kW filament). Tube damage from filament
overcurrent is unacceptable in this case, as the tube costs more than a
yearly salary or the cost of a home..."

...

"The design, construction and testing of a new power amplifier for 200
MHz (250 kW average power) is described in three papers going into the
2011 IEEE/Particle Accelerator Conference Proceedings. Here is a link to
the preliminary proceedings,

http://www.c-ad.bnl.gov/pac2011/proceedings/html/sessi0n.htm

Then go to Tuesday poster session (TUP) and look up, download the pdfs
for TUP133, TUP134 and TUP135. There is a mistake in the abstract for
the third one, that says I reached 480 kW of power, thats going to be
done with two tubes combined."

"A TH628 double-ended tetrode ("Diacrode") has been purring along all
day heating a water load with 26 kV plate voltage, and 110 amps of plate
current. At 200 MHz, it's been quite a leap, average power today was 180
kW in the load, 2 MW peak in pulses. A little RF leakage was causing havoc
before, now fixed. On Monday, it'll stretch out with around 250 kW average,
based on the good test results today, Friday, the 13th. I'm posting this
AFTER testing - worried about even trying today. Me, superstitious? I
remember being nervous about dipping the plate too slowly on a 3-500Z
at full voltage in my SB220 years ago. This monster makes me jumpy when
someone closes a book or drags a chair around the test bench... Is been an
experience not to forget, for this ham"

...

"2 MW out with 26 kV x 110 Amps (plus 60 kw of drive feedthru due to
cathode driven) is quite good efficiency, 68%. 180 kW is the average
(equivalent CW) power, as it is a pulsed amplifier. The peak power needs
high peak voltage. The high average power needs a lot of DC current and
makes heat. Both regimes are tested. No new photos online besides the 3
technical papers I posted links to, a month ago. "

...

"Techniques like multiphase cooling (Hypervapotron is one mfr's name)
to exchange anode heat with water at much higher efficiency, pyrolytic
graphite for grids, and improvements to cathodes for higher loading are
all improvements that are working their way into big tubes. These make
tremendous increases in power density. In cathode loading, a tube with
thoriated tungsten (going back to Langmuir and others at GE) can deliver
as much as 3 Amps per cm^2 of electrons in a well designed tube to get
lifetimes > 20,000 hours. Figure the size of the filament basket and you
understand how the peak current can be 1000 amps overall. Thats just not
possible in a transistor junction, no matter what type. Combining lots of
transistors gets there, however, at a cost.

For example, the TH628 from Thales has over 1 MW plate dissipation, with
less water flow and high performance into the VHF. The 4CM2500KG from
Eimac has over 2 MW plate dissipation, with slightly lower frequency
ranges but very high CW power. I wouldn't say that the few remaining tube
manufacturers are sitting still, but they are not cranking out new designs
yearly either. They are making measured step improvements to their existing
products, a lot easier than starting over with a clean page. These two
examples are sort of the crowns for these two companies to push performance
limits.

Others mentioned cold cathode research with field emission and such. Metal
filament and oxide-type tubes and cathodes provide excellent long life
performance, if properly rated for the application. It's when engineers
try to exceed the ratings that lifetimes shorten. Or manufacturers reduce
quality (rising gas levels for example). Similar lifetime-reductions happen
in solid state devices when application demands more than the device
designer expected, except much quicker - flash, pop.

As a designer of power amplifiers using both tubes and transistors, I see a
good trend in solid state, with 1200 watts or more per device. But I don't
expect to see a megawatt in a device, something that I can and do see with
new tubes. Ham radio will eventually quit using tubes except for nostalgia
reasons; we are still seeing a few 304TL and 450TH amplifiers getting built
due to having the parts and for fun. As long as their is a market, the Far
East will provide bottles for building simple amplifiers. Don't expect any
new tubes for this, however. HV power is getting harder to come by due to
cost of iron and copper. Switch mode technology comes just in time."

...

"For very high power, I cannot use resistor loads period and resort to
water column loads, where the water is doped with a weak ionic solution to
make it appear as a lossy medium, and the length of the pipe is such that
the reflected power at chosen frequency is very low. These are used for
3.125, 6.125 and 9.375 inch diameter coaxial loads. I've been testing QRO
for months now at 210 kW average, 2.1 MW peak, pulsed, with about 70 gpm
of ionic solution flowing through the column, 50 deg C outlet temperature!
I wish I could use some of this waste power to heat my home but its a long
way from the work QTH."

...

"I have been relearning this in the past 2 years, as I had to build a
big HV feedthru cap for my 200 MHz cavity amplifier at work. Nothing was
commercially available. I am passing 125 amps of plate current through
it, at 28 kV DC. I borrowed an idea from the bigger Broadcast Electronics
FM transmitters that I remember. I made an aluminum cylinder and then a
matching aluminum slug that fits into the center with a slight interference
fit. By wrapping the slug with Kapton, several layers, then heating the
outer cylinder, I pressed the center slug into place, and it sealed tight
when it cooled off. Then I potted the edges. I did the potting under vacuum
to ensure that there are no air pockets that will ionize and cause failure
of the Kapton. It has been working for a year now, fingers crossed..... "

A drive power of 60 KW. Peak power 2MW. At 200 MHz. 

Gents, _that_ is QRO.

-- 
Mike Andrews, W5EGO
mikea at mikea.ath.cx
Tired old sysadmin