On Thu, 3 Apr 2008, Rick wrote:

An inrush current limiter
in series with the filament only costs a dollar or two and can easily pay
for itself a hundred times over.

Hang on here. Do you know what 813s cost these days? Still only around $10
!!
I got two at a hamfest last week. Got them hooked up to a filament xfmr,
and let them cook for an hour. Boy are they sure beautiful sitting there
glowing in the dark. They seem to be saying to me, "Hey we've been waiting
for 60 years to get a chance to make some RF!"
I also have 2kv waiting. So next step is to jury rig them up and see if
they draw plate current, with a little bias control, before I commit to
putting them in a new homebrew amp.
Since I just finished an 8877 amp this winter I am a little cautious and
aprehensive about hamfest tubes, because when I first fired up the 8877 with
4kv I got just that - fire.

Fuses in primaries and secondaries of plate transformers are your friends.

Variacs feeding the primary of the plate transformer (and with an ammeter)
is a good idea if you start out with low voltage on the primary, and pay
very close attention to all of your meters.

It is better to have more meters than not enough meters.

Sometimes an ohmmeter check on pins of tubes, BEFORE you do the smoke
test, is wise (I have done smoke tests where I got smoke, too).

Seems the "I think it is good" tube a guy gave
me actually had a grid-plate short.

I have had the experience, at hamfests, where, after money goes to the guy
and the thing goes to me, then he tells me "Oh, by the way,..(fill in
the blank)...."

Took a couple weeks to clean up THAT
mess. Fun stuff.
Thanks for all the replies, OTs.

After deciding to build a linear, the decision came down to the 813 for
the following reasons:

1. Large glass tubes need no special socket or chimney.
2. Tube manual says no forced-air cooling is needed.
3. Sockets are cheap. Plate capes are cheap.
4. Tube is cheap.
5. Filament cathodes are "instant on" whereas heater cathodes need warm-up
time.
6. Large tubes generally mean old-fashioned large grids and that means a
larger margin for error (i.e. overloaded grid dissipation) than those new
small sexy expensive small metal-ceramic tubes with dinky, flimsy, small
grids that vaporize if you accidentally overdrive them for more than 1-2
milliseconds (I've heard guys talk about this on the air).
7. I have no fan on my 2x813 G-G amp, and it is very nice and quiet
(unlike three other commercial linears I've had with objectionable fan noise).

Pictures of some of my stuff at: http://w4pon.freeshell.org
Not all that pretty but it all works and I know how to fix it because I
built it.

73

Rick K2XT