On Fri, 8 May 98 00:59:05 -0800 Rich Measures <measures@vc.net> writes:
>>
>>>
>>>>>4E27s from the same lot were quite gassy, presumably due to
>>>imperfect
>>>>>metal/glass seals.
>>>>
>>>>Are you now confusing or mixing up "gas" Rich. Gas due to
>imperfect
>>>>seals is not the same as gas released internally from the
>materials.
>>>
>>>The point is that gas is not gradually absorbed by a getter -- i.e.,
>
>>>the
>>>rauchian disappearing gas theory does not wash. However, the basic
>>>problem is that gas is seldom found in tubes with bent filament
>>>helices /
>>>a filament-grid short.
>>
>>But you still did not answer the question Rich..sounds like a
>rauchian
>>evasion IMO.
>>A getter has nothing to do with "gradual" gas absorbtion. The anode
>is
>>the culprit and the getter is the cure according to my understanding.
>>Maybe we are just saying the same thing in a different way.
>>
>I am saying that I have seen gassy 3-500Zs, and the gas does not
>disappear, and that filament/grid shorted 3-500Zs very rarely exhibit
>gas
>leakage.
I still feel that you are combining outside atmospheric gas due to seals
AND gas released by the anode into one generalization Rich.
They are not the same.
Yes, seal leakage will show up as a glow with a hi-pot tester and can
often be baked out. OTOH gas released internally would not show up on the
tester, it shows no color and leaves no traces after the arc event. With
a plasma zap at the relatively low voltage of a 3-500Z I would not expect
to see a point contact type of arc mark. In BIG tubes that would probably
not be the case but 3200V is not really considered HV !
I have never seen arc-marks in a filament/grid shorted
>3-500Z
>
>> In a tantalum plate tube the plate must show color for the getter to
>>even begin to do its job. A typical SB-220 owner will rarely, if
>ever,
>>get his amp into the condition necessary to harden the tubes.
>>Note that I keep stressing the SB-220 since it is probably the most
>>common 3-500 amp ever built and in use today. For the time being I
>dont
>>care squat about any other tube...lets just agree to stick to the
>3-500
>>and the SB-220 since we and many readers are all very familiar with
>it.
>>
>
>>>
>>>>I also suspect that your microamp hipot tester will not be capable
>of
>>>>creating the conditions necessary to recreate a gas discharge
>>>event...
>>>
>>>It matters not whether the high potential supply being used to
>measure
>>>
>>>gas leakage is capable of 1mA or 1000A.
>>
>>I disagree. A microamp hipot tester will not be capable of any plasma
>>generation..if you want to try and duplicate the fault conditions.
>>
>My highpot tester has plenty of current capability to create a gas
>plasma
>in gassy tubes. The bluish plasma glows plainly in a darkened room.
That is certainly not a plasma. No current is being drawn and you can see
the same glow in an old 6L6 that you can see in a 3-500Z.
>
>>>
>>>>but I will leave that to Arlen and others with a formal education
>to
>>>comment
>>>on.
>>>
>>>>Sometimes it takes a 2x4 approach to get a persons attention Arlen.
>I
>>>>think your message was well put and I know that I learned from it.
>>
>>>>As has been mentioned here numerous times...how do you know the
>>>condition
>>>>of the resistor minutes, days or weeks before the arc? Unless you
>>>can
>>>>recreate the event in a controlled experiment I fail to see why you
>>>>continue to harp on the resistor.
>>>
>>>because the suppressor resistor is virtually shorted out by
>>><1-milliohm
>>>of copper buswire that typically has 0.08 uH of L, which should
>>>protect
>>>it from DC arcs in the tube, but which would not protect it from
>>>bursts
>>>of vhf energy. .
>>>
>>>>In the case of the SB-220 would not the almost instantaneous
>>>discharge of
>>>>the 3200VDC ( in standby which is where most arcs appear to occur)
>>>>not cause damage to the resistor?
>>>
>>>no, no
>>
>>YES YES IMO
>>
>You have not satisfactorily explained how a 50 - 100 ohm resistor that
>is
>shorted out by 1 milliohm of copper buswire can be damaged by a DC
>anode
>arc which mysteriously leaves no arc craters in the filament-grid
>shorted
>tube.
I offered my own idea on that Rich several days ago...so we still
disagree but neither one of us is in a position to show 100% proof.
With the availability of the full engineering libraries of several
universities on the Web, I would think that a few of our better educated
brethren could find an answer...so that we may finally change to a new
subject.
I'm waiting for Ian and others to kick in with some more tetrode and UHF
info...after all QRO does not stop at 30 MHz.
>>>
>>>>Would not the field developed in the
>>>>suppressor L be sufficient to "bend" the carbon in the resistor
>causing
>>>>the case to crack?
>
>no
You offer no reason why not. It appears possible to me but I dont have
the knowledge to take it much further.
>
>>>>If one were to believe that the magnetic field
>>>>caused the filament to bend then the same reasoning would hold for
>the
>>>>resistor...at least it makes sense to this uneducated dummy.
>>>>
>>>not to yours truly
>>
>>Why not? If I'm wrong please educate me. I promise to not ask a
>months
>>worth of useless questions!
>>
>Carbon granules do not bend. Wire bends.
Does not any element react to stimulus? So it doesn't bend...how about
expand? What exactly does happen to the carbon element when it is
subjected to extreme momentary stress when surrounded by a huge magnetic
field?
Both of us stress the use of metal oxide resistors since they can absorb
overloads much better than carbon.
>>>
>>>>Rich has also failed to ever explain how an amp in standby, biased
>well
>>>>beyond cut-off can suddenly have a parasitic event. Particularly
>in
>>>>an amp that is completely stable when keyed with no drive and the
>Tune
>>>>and Load caps are varied in an attempt to force a parasitic.
>>>>Not to be confused by external arcs in tank circuit components
>while
>>>>the amp is in a RF amplifying state.
>>>>
>>>Only 5% or so of the parasitic oscillations I hear about fall into
>>>this category,
>>
>>Yet, almost 100% of my customers and reports here indicate the bang
>is
>>during standby. Again...ONLY the SB-220 is discussed here. Heck,
>throw in
>>the TL-922 also if you want.
>>
>>> however, the open contact gap in a typical cathode bias
>>>relay is only 25 thousandths, which is easily broken down if the
>tube
>>>conducts for a microsecond or so for who knows whatever reason.
>>
>>I may be missing your point but I doubt if the relay path is ever an
>>issue during a catostrophic arc....
>
>Are you contending that current that causes the damage to the
>cathode/filament passes through the 10k to 100k ohm cutoff bias R?
An excellent point Rich...and no, I cannot accept that the arc passes
thru a 100K 1/2W resistor in the SB-220 without ever a sign of damage.
>The
>only paths of current to the cathode are through the relay contacts or
>
>through the cathode cutoff bias resistor. . I doubt that the current
>
>needed to blow a 1A grid-gnd RFC
In the SB-220 the grid RFC's are 160ma CCS rated. They are 1A + in the
TL-922.
could be passing through the cutoff
>bias
>R, which leaves only the relay contacts.
Yet I have rarely ever seen a bad SB-220 relay but I must agree for the
moment that it does appear to be the logical path. The biggest problem
IMO is that no one that I am aware of has ever been able to completely
"capture" the event...much less duplicate it. I have tried with a
modified PS and 3800V and still cant force a BANG. I believe that you
also stated similar frustration.
.
>
>>I've never replaced one (a relay) for that reason.
>
> There is typically no fatal damage to a relay whose contacts have
>sustained a brief arc.
I concede that point Rich...at .025" spacing it would not take much
voltage to arc.
73 Carl KM1H
>
>
>cheers
>Rich...
>
>R. L. Measures, 805-386-3734, AG6K
>
>
>--
>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
>
_____________________________________________________________________
You don't need to buy Internet access to use free Internet e-mail.
Get completely free e-mail from Juno at http://www.juno.com
Or call Juno at (800) 654-JUNO [654-5866]
--
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
|