G3SEK says:
>Those tactics are repugnant. They are also directly responsible for
>depriving this reflector of public contributions from EIMAC engineers -
>as one of them has told me by private e-mail.
Having interaction with the Eimac engineers, marketing, and sales folks for
about 18 years now, I have worked with Mr. Brandon and others. And it is
true that they hold this forum at arms length and don't jump in the pool to
swim or sink. Remember, these companies make their livelyhood making and
selling power tubes, and cannot afford to partake actively in the center of
our controversies. Ham radio is a part of their market. We shouldn't expect
them to step forward and defend or take sides. It would be nice and helpful
to hear their experience, I agree. Below, I want to try and defend the tube
manufacturers. I do not have financial interest in any one of them.
K5PC says:
>I also think that one can safely assume that Eimac determined by TEST,
>not a paper assumption based on other unrelated tube specs
>that a 50 volt >5 Joule arc in a 3-500Z or 8877 will cause permanent
>damage to the small wires in the grid and cathode elements. Naysayers
>should run about 10 Joules at 50 volts through one on their tubes and
>report back. Warning: Eimac's warranty does not cover fault related
>failures either by accident or on purpose!
It is my understanding from discussions with CPI/Eimac, RCA/Burle, and Thomson
tubes that the Joule limits on dumping energy into tube electrodes is
indeed directly related to the size of the grid wires, the cathode bars,
and the filament. It is not always correlated to the average power
dissipation capability of the elements. Even though the fundamental
definition of energy is watt seconds, we are comparing an arc in a tiny
area to the overall average of the grid basket. For instance, some of the
Burle tubes have water cooling internal to the structure, behind the grid
suppports. In this case, the dissipation may be huge, but the individual
wires that make up the grid are still only a mil or three in diameter. So,
if a big fat arc comes about, it can vaporize the fine wire, or rip out a
bond in the array (where the wires cross and are tacked).
Each tube company has their recommendation, the Eimac one was already
discussed here. Application Bulletin #17 on Fault Protection appears to be
well stated. For low energy operation, with oxide cathode tubes up to 1500
Watt rating (such as 8877), four Joules of delievered energy will not blow
a 40 AWG wire. For thoriated tungsten tubes of higher power, Eimac speaks
of 50 Joules as the limit for total energy delivered in an arc in the tube.
The 3-500Z, 4CX1500A, 5CX1500A, 3CX3000A7, 3CX10000A7 and even 3CX20000A7
are listed as exceptions, and probably other tubes with handles should also
be protected at the 4 J level. It is exactly the fine grid wire and
filament structure that they are trying to recommend that you protect.
Burle has a slightly different twist, still recommending to keep the fault
energy under some limit. For example, the 8807 20 KW Cermelox tetrode (a
popular old VHF TV amplifier tube) says 100 Joules. They state in the data
sheet to use a 30 Ohm minimum R in series with the plate supply, and 50
Ohms in the screen and grid. For the 8501 tetrode, 100 Ohms in series with
the screen is stated "Must be used". Ten to 50 Ohms in the plate is
recommended.
In Burle technical paper TP-105, a method of testing ones protective system
is shown, using a piece of 1 mil thick aluminum foil and a piece of solder
on an insulated stick. If a hole > 0.01 inches is blown in the foil when
the solder is brought close to the foil, then it will damage tubes. We have
tested this method and the Eimac preferred method (also Thomson) for a
28-32 guage wire across the HV supply. These correlate
fairly well (sorry, difficult to quantitatively measure this). Again, the
tube manufacturers are really concerned with the energy in the arc, and the
ability to blow holes in the grid wires.
Interesting is that in the Seimens transmitting tube data book, they
actually tabulated the copper test wire diameter, and the A^2 x seconds
rating for every tube individually. It goes from 0.1mm wire, up to 0.3 mm
wire.
Philips/Amperex/Richardson have the same type of table for their
communication tubes.
One should stick to Eimacs recommendation, if you want your tubes to be
covered under their warranty. Stick to Svetlana, Seimens, Thomson, EEV or
whoever you bought them from. Of course, if you prefer to fly with your own
rules, go ahead, it's a free world. Just don't expect the manufacturer to
be sympathetic if it breaks. And they don't give a darn if it is a
parasite-induced arc over, or "rocky-point" or gas, it's the total
follow-on energy from the power supply that should be the primary concern
in selecting a series R or a crowbar circuit. You can add a plate
overcurrent trip or fuse the HV as you like, if you think it will protect
the bottle better from whatever gremlins you have.
-------------------------------
K5PC asked:
>Gas arcs are still a mystery to me. Perhaps the chemists on the reflector
>can tell us what the make-up of the gas is in a tube, and if an arc would
>change its properties, or eliminate it altogether. Apparently there are tubes
>that survive arcs; gas, point-to-point (Rocky Point) and external flash overs
>and go on to live a long useful life. Some don't. The term "gettering" has
>been used often in reviving old tubes. How do we know that an arc cannot be a
>vehicle for this process?
and L. Wald asked:
>Just how good is the vacuum in the tubes?
For super power triodes, we have a pair of internal Vacion pumps running
all the time. Actually they are not Varian Vacion but Perkin Elmer ion
pumps - Vacion is a tradename of Varian I think. We can
monitor the vacuum while the tube is running RF, and we do witness
occasional arcs inside the tubes.
The background vacuum varies as indicated by the ion pump power supply
current. The tubes will be sitting at about 10^-7 Torr, then a burp occurs
(an arc), then it jumps to 10^-5 or so. It recovers rapidly (minutes)
either due to the getter action or the ion pumps. In these large ceramic
metal tubes (beer keg sized WITHOUT handles, with crane lifing hooks), one
cannot leave the tube on the shelf for years without vacuum degradation.
The warranty requires checking the vacuum with external power supply every
few months. We put a power supply on every tube's ion pump, and store them
coninuously like this. When we first turn on the RF, the vacuum does a
dance until it settles back to the background level. It is imperative that
we start slowly, and not slam full RF power (full plate swing) on it until
the outgassing can be accomidated by the ion guages and the getter.
Sorry, don't have any information on the vacuum performance of glass versus
ceramic/metal tubes. I believe that it is worse in larger tubes, as
mentioned above.
CPI/Eimac has another bulletin (AB 21) on Condiioning of Large RF Power
Tubes. It discusses the outgassing, hipotting recommendations, and
debarnacling tubes if you are interested. Normally these are not concerns
for smaller power tubes.
John
K5PRO
--
FAQ on WWW: http://www.contesting.com/ampfaq.html
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
Search: http://www.contesting.com/km9p/search.htm
|