All,
I can't speak for the gas release, but I can speak as the owner of a 2 x
4CX250B
144MHz amp that used to "let go" randomly before I got a handle on the
problem...
I'm talking of 25+ years ago here... a UK design known as the "plumber's
delight"
which was basically a push-pull pair much like the W1SL except the anode
design
was in 2" copper tube....
Anyway, the "failure mode" in the one that I acquired (and debugged) was
that
the grid bias (switched between Vbias and Vcutoff) and the screen (switched
betwen Vscreen ~330V and Ground) was controlled by a pair of double
pole double throw (DPDT) relays -- one per valve -- big octal plug in
units...
The problem(s) were that:
a) while the relay contacts were "flying" (approx 10-20mS) the valve's
electrodes
were effectively open circuit and found their own potential due to
electrostatic
forces on them
b) the two relays were independant -- both controlled from the same source
but timing could vary.
With this amp in its native form (as I received it) it would perform ok for
several
months and then randomly (usually on release of PTT) go "phutttt" and blow
the
anode fuse... sometimes making a soot mark on the screen ring of the tube
socket, very occasionally taking out a tube!
It took me a while to get my head round what was going on (I was 16, new to
high power RF, etc.) but I worked it out... you had something like 2200V DC
on the anode, 0V on the cathode and on TX/RX changeover both the control
grid and screen grid were left open to float for some millisceconds whilest
the
relay contacts were in flight... basic physics suggests that the charge on
the
electrodes will tend towards the average/mean (based on distance/seperation/
capacitance, etc.) of the other electrodes
Basically the amp had the tendany (possibilty) to go in to run-away for a
few
millisedconds each time it swicthed betwen tx and rx. .. I'm stilll
supprised it
didn't fail more often!
I fixed this amp by:
a) hard wiring Vscreen to both tubes (bypassing the relays) - always at
~ 330V DC
b) increasing Vcutoff to about -120VDC
c) using on relay to control Vgrid to both tubes and arranging for a 100K
"pull down" from Vcutoof to the grids such that they could never be open
circuit...
Problem went away... no more random "phuttt" noises and blown fuses - I was
real pleased with myself :o)
Mike G8TIC
----- Original Message -----
From: "Ian White GM3SEK" <g3sek@ifwtech.co.uk>
To: <amps@contesting.com>
Sent: Tuesday, July 12, 2005 10:14 AM
Subject: Re: [Amps] Tetrodes
> wrote:
>>
>>On Jul 11, 2005, at 11:36 PM, Ian White GM3SEK wrote:
>>
>>> wrote:
>>>>>>
>>>>>> ... by stopping the flow of current into the screen PS.
>>>>>>
>>>>> If the "glitch" is an arc from anode to screen, the mains fuse won't
>>>>> protect the screen at all.
>>>>
>>>> A tube that arcs from the anode to the screen is already a goner, Ian,
>>>> so there is nothing of value to protect.
>>>
>>> That simply isn't true. Well designed power supplies can protect the
>>> tube, the anode supply and the screen supply against a wide range of
>>> faults, including arcs. Many of these faults are completely
>>>recoverable.
>>
>>What condition would make the anode arc to the screen?
>
> As many folks will remember from previous repetitions of this topic,
> reasons for anode-screen arcs include:
>
> 1. Loss of load at a phase angle that results in extremely high anode
> voltages. With certain amps, you can make this happen anytime you wish:
> remove load, apply full drive - BANG.
>
> If the RF voltage is high enough, the anode flashes over to the next
> lower-potential object within sight. For a tetrode, this is usually the
> screen, so the screen supply experiences a very large negative current
> input from the tube.
>
> For a GG triode, the RF arc will hit either the chassis if the arc is
> external, or the control grid if the arc is internal. An external arc to
> chassis probably won't harm the tube, but an interna arc to the grid
> certainly can.
>
> The so-called "glitch resistor" in the B+ supply is a vital part of
> every tube amplifier, because it limits the peak current: it's simply
> Ipk=V/R. This resistor is not intended as any kind of fuse - it simply
> limits the maximum possible current for a few vital milliseconds, until
> something *else* has time to switch off. Eimac Bulletin 17 covers this
> very adequately.
>
> If the power supply provides adequate over-current protection for the
> screen and anode supplies, and adequate over-voltage protection for the
> screen bypass capacitor, then arc faults due to loss of load can be
> completely recoverable. Re-connect the load, press the RESET button and
> the amp will come back online.
>
> 2. Bugs in the tube chimney. Don't ask me how they got in there, but in
> FD situations I have sometimes found an earwig or a moth bridging the
> anode and the screen. Again there's an arc and the amp trips out, but
> this time it won't come back online until the charred corpse has been
> removed.
>
> 3. Internal arcs due to gas release. I know that Rich doesn't like to
> accept this; but I go along with the vast majority of tube manufacturers
> and professional RF engineers who have seen enough evidence to know that
> it happens - so we protect against it. Again, these faults are usually
> self-clearing, and the amp will come back online when the RESET button
> is pressed.
>
> 4. Any other reason the amp may choose! The most important practical
> point is to accept that amplifiers occasionally DO arc, so they need
> some protection.
>
>
>
> --
> 73 from Ian G/GM3SEK
> _______________________________________________
> Amps mailing list
> Amps@contesting.com
> http://lists.contesting.com/mailman/listinfo/amps
>
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|