Amps
[Top] [All Lists]

[AMPS] Glitch protection relay

To: <amps@contesting.com>
Subject: [AMPS] Glitch protection relay
From: jtml@lanl.gov (John T. M. Lyles)
Date: Wed, 29 Aug 2001 16:38:37 -0600
This note was sent with two JPEG files with photos, to 2 and W8JI. 
Here is plain copy. I can forward the photos upon request.

-----------

I have inspected the insides of probably a dozen big tubes in my 
work, and most - if not all - of them have arc marks inside. Some 
were still operating when pulled and opened, as they had excessive 
hours and needed rebuilding. We pay a standard fee to the 
manufacturer to open each tube and analyze the failure modes and 
condition before rebuilding as the rebuilding process is expensive, 
but cheaper than a new 7835 grounded grid triode. I usually witness 
the opening when a failure mode isn't understood. Autopsy is the best 
method to find out what killed it.

Attached are two photographs showing two different tubes at the 
manufacturer. In one, for serial number P9, you can see a 1/4 inch 
chunk of the grid missing, my pen pointing to the hole. This was 
caused by numerous plate to grid arcs. We obviously had problems with 
the protective crowbar for this particular amplifier. We also have 
vacion guages mounted on each of these big tubes, and can observe the 
pressure surges when they outgass. It always occurs on turn on when 
the filament is ramped up to full temperature from cold. We call the 
first hour of RF power conditioning, as we have to slowly increase 
the plate V and the duty factor to keep the vacion reading below a 
certain pressure. There are no parasitics which will happen during 
this first hour and then subside. It is strictly a materials 
function. New tubes are conditioned for over 100 hours to burn off 
barnicles and outgas the oxygen free copper and other materials as RF 
current heats them.

In the second photo, two anodes are compared, one with cracks in it 
due to a manufacturing defect called grain growth. The lower tube has 
no such cracks, but instead one can see numerous arc marks around the 
large radiused edge of the anode. To see it with you computer, you 
should zoom in with your picture viewer program around the upper edge 
of the lower tube. There are clearly visible black marks and smudges, 
with dings into the copper.

I recommend that hams consider other methodologies which can cause 
arcs besides an "all inclusive" parasitic theory; in industry we 
don't subscribe to that theory unless other reasons are found to 
promote that hypothesis. Even then, the smoking gun is difficult to 
find.  The Fyler/GE paper on parasitics from the WLW 500 kW 
transmitter that I sent to Rich about 4-5 years ago told of many 
instances of parasitics which had to be snuffed. A reading of the 
complete paper reveals that they were breaking new ground with high 
RF voltages and standing waves in large circuit layouts. Flash overs 
were common in that rig, which was the first of its kind. Parasitics 
were really a problem. One certainly hopes that these are not the 
sort of problems which we have today, 80 years later. In our 
amplifiers at work it certainly isn't, and the photos prove that 
outgassing is the predominant cause of internal arcs in tubes with 
handles. Now the reasons for sudden outgassing are many, including 
changing RF loading, power supply voltage fluctuations, changing 
drive power, changing duty factor, changing VSWR (fast!) or loss of 
cooling. I am sure there are other reasons also, but in essence we 
are getting a thermal condition which will push the materials to a 
new plateau of operation, with subsequent outgassing until they 
settle down at the new level.

73
John
K5PRO


2 sez:

>//  Of the dozens of kaput tubes I've autopsied,  I have yet to find one
>with an internal arc mark.
>
>>If the tube and other components have enough HV breakdown to
>>operate at the desired frequency, they almost certainly have
>>enough for any unwanted oscillation.
>>
>//  ... whistling in the dark?

-- 

--
FAQ on WWW:               http://www.contesting.com/FAQ/amps
Submissions:              amps@contesting.com
Administrative requests:  amps-REQUEST@contesting.com
Problems:                 owner-amps@contesting.com


<Prev in Thread] Current Thread [Next in Thread>