I have seen 3-500Z tubes mounted in the johnson ceramic sockets. The Heath
SB-220 uses them.
Will the 4-400 tube handle the airflow the same way? Usually you see the air
system sockets used so that the air will flow up through the bottom. Will
the filament connections remain solid in a 275 ceramic socket with a muffin
fan blowing across the tube?
? I have seen the sockets raised a bit so that the filaments get cooling.
Just wondering if the 4-400 would work that way as well? Any experience
with this?

Doug N4IJ

DOUGLAS SNOWDEN wrote:
I have seen 3-500Z tubes mounted in the johnson ceramic sockets. The Heath
SB-220 uses them.
Will the 4-400 tube handle the airflow the same way? Usually you see the air
system sockets used so that the air will flow up through the bottom. Will
the filament connections remain solid in a 275 ceramic socket with a muffin
fan blowing across the tube?
? I have seen the sockets raised a bit so that the filaments get cooling.
Just wondering if the 4-400 would work that way as well? Any experience
with this?
=========================================
During the late 1970s ,the old 405 lines TV system in the UK operating
in Band 1 (approx 45 -65 MHz) was decommissioned and some equipment
landed at junk sales held at radio clubs . I managed to obtain two 1 kW
amplifier racks (out of 7 released from a local 7 kW TV repeater )

The rack had 2 QY4-400 (tetrode)valves with the (flat,ceramic) sockets
fitted approx 12 mm below the chassis on stand-offs . The space under
the chassis was fully enclosed (almost air -tight) and was fitted with a
pressure switch (with a large diaphragm to sense and switch a low pressure).
A centrigugal fan blew air under the chassis which escaped via the valve
socket openings thus cooling the valves. There were no valve 'chimneys'

Upon fan failure the pressure switch would initiate the rack's shutdown.

A simple but effective way of cooling the valves.

Frank GM0CSZ / KN6WH

That helps Terry. I guess I will follow pretty much the way you described it
except will try and put a chimney on the tubes. I plan on driving the crap
out of them! Naw! Just joking. I try to build and use to last.

Doug N4IJ

wrote in message
oups.com...
Doug, Here's what I remember from several RCA and Collins Broadcast
transmitters and a KW amp I built: Those ceramic sockets SHOULD have
holes in them of about 1/2" diameter. Line up a 4-400A with the socket.
Notice that the BASE of the 4-400A has holes that line up with the
socket holes. Notice that the 4-400A has a metal base which has a
series of holes around it. It's very important to put lots of air thru
those holes to flow around the tube base pins. If you're going to run
lots of dissipation (definite color on plate) then the usual approach
is to mount the socket just below the chassis on standoffs or about
3/16 to 1/4" as I recall. THe SIZE of the hole in the chassis then is
set to have about 3/16 to 1/4" of space around the outside of the metal
tube base, so there is additional airflow which will hopefully stay
fairly laminar around the upper part of the tube envelope. NOTE:
Normally that metal base ring is grounded by several metal clips.

SOME systems used the "Air System Chimney" (A pyrex glass open-topped
"bottle" around the tube) to direct the air around the tube envelope,
although 1000W OUT Broadcast transmitters from Collins and RCA did not
use them, and their high level modulators put out 750 watts of audio,
at, say, 50 percent efficiency, so there was about 375 watts of plate
dissipation per 4-400A at 100% sinewave modulation. Which at one time
was OK, because the actual duty cycle on "Typical Program Material" was
much lower. But about 1970, as I recall, watching the plate color
during "Magic Carpet Ride" by Steppenwolf, with an asymetrical limiter,
20 db of syllable-rate compression, plus the pre-processing of the
airplay cart by a demented Program Director, the plates looked a lot
like 100% modulation during an Audio Proof of Performance test :-)

Regards, Terry King ...On The Mediterranean in Carthage

DOUGLAS SNOWDEN wrote:
I have seen 3-500Z tubes mounted in the johnson ceramic sockets. The Heath
SB-220 uses them.
Will the 4-400 tube handle the airflow the same way? Usually you see the air
system sockets used so that the air will flow up through the bottom. Will
the filament connections remain solid in a 275 ceramic socket with a muffin
fan blowing across the tube?
? I have seen the sockets raised a bit so that the filaments get cooling.
Just wondering if the 4-400 would work that way as well? Any experience
with this?

Doug N4IJ


I built a linear using two 4-400's. I used the Johnson ceramic
sockets (square type) that did have some holes which lined up
with the ones in the bottom of the tubes. I mounted the sockets
under the chassis FLUSH with the bottom of the chassis. I pressurized
the chassis with a 4" boxer type fan mounted under the tube sockets.
This fan was one of the high speed types that could move lots of air.
I also mounted a pair of the same type of fans next to the
bottles on the top of the chassis, blowing air from outside
into the chassis past the tubes. The air flow went through the base holes
and around the glass bottle. This is similar to the way the
SB220 linear was build, as with many HB amps. The tubes did not
suffer any heat effects, while the pair ran near the legal limit.

BTW, 4-400's work ok in GG, but they like more HV than their triode
counterparts. While a pair of 3-500's work well with 2500v, the 4-400
likes more like 3300-3500v to get the same output and IM figures.
So your tank circuit will have more "L" and less "C".

Doug, a suggestion about "Longevity" for 4-400's (Actually, any
thoriated-tungsten filament power tubes):

These tubes are very sensitive to filament voltage. Make sure the tube
pins and socket are nice and clean and shiny (don't use any
abrasives!).. Measure filament voltage at the socket. This is all to
make sure the tubes REALLY get the correct filament voltage to the
filament... It's 28 Amps for two tubes!

Tubes that are new or in good condition do NOT need the full rated
filament voltage for full output. But if you do not have good AC line
regulation at your location, it's hard to get it exactly right.

In Broadcast use, I modified some RCA and Collins transmitters by
adding a "Sola" type voltage regulator on the AC line, and an
adjustable autotransformer ("Variac" or "Powerstat") to feed the
filament transformer. With all new tubes, I could measure the
distortion percentage on the transmitted audio, and lower the filament
voltage until the distortion just started to creep up. And set a little
margin. As I recall, a 5.0V tube would initially run fine at 4.6 to
4.7V. On one transmitter, the tube life went from 4000 hours to over
9000 hours (a year, 24-7). Of course, the filament voltage was slowly
raised, and the tubes were 'retired' when the filament voltage for low
distortion got up above 5.0 volts. In noncommercial service, you could
get some retired 4-400a's to work for another 1000 hours or more by
raising the voltage to 5.1 or 5.2

-- Plenty of air
-- Clean bright tube pins and sockets
-- Minimum required filament voltage

were the main considerations.

Have fun!

On Sun, 5 Mar 2006, DOUGLAS SNOWDEN wrote:

I have seen 3-500Z tubes mounted in the johnson ceramic sockets. The Heath
SB-220 uses them.
Will the 4-400 tube handle the airflow the same way? Usually you see the air
system sockets used so that the air will flow up through the bottom. Will
the filament connections remain solid in a 275 ceramic socket with a muffin
fan blowing across the tube?
? I have seen the sockets raised a bit so that the filaments get cooling.
Just wondering if the 4-400 would work that way as well? Any experience
with this?

My recollection from Eimac tube spec sheets is that there must be
forced-air directed upwards from the non-tube side of the tube socket and
there are specifications for flow rate and pressure. Both the 3-500 and
4-400 (100 watts less plate dissipation) call for 5 v at 15 amps (IIRC) on
the filament and that is a lot of heat. I don't know if the 3-500 is
pushed closer to the limits on this but Eimac made a special socket (and
chimney) for the 4-400 but many ARRL handbook amplifiers went for a much
simpler socket. And, the Johnson Thunderbolt (I had two of them at one
time) used a simple ceramic socket plus an under-chassis phonograph motor
type of fan to blow air up from the bottom as well as an above chassis fan
motor to blow more air around the pair of 4-400s for a 1 KW DC plate input
level (with dull red cherry glow on cw carrier and a little less glow on
SSB voice).

Beyond that, if you are thinking of running 4-400s instead of the 3-500z,
you also need to worry about bias voltage a little and you would have to
resolder the socket pins because the 3-500 is a triode and the 4-400 is a
tetrode (need to ground the screen, as well as control grid). But I have
run 4-400s in GG in a homebrew amp in the past.

Doug N4IJ

ORIGINAL MESSAGE:

Straydog wrote:

there must be forced-air directed upwards from the non-tube side of
the tube socket and there are specifications for flow rate and
pressure.


~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
"Flow rate and pressure" are nice, but the *only* important thing is
the temperature of the tube at various points as specified by the tube
manufacturer. If you are serious about proper tube cooling, use either
a non-contact thermometer of some kind or thermally-sensitive paint to
ensure your tube is within temp specs. Flow rate and pressure are a
means to an end, not an end in themselves.

Bill, W6WRT

wrote:
Doug, a suggestion about "Longevity" for 4-400's (Actually, any
thoriated-tungsten filament power tubes):

These tubes are very sensitive to filament voltage. Make sure the tube
pins and socket are nice and clean and shiny (don't use any
abrasives!).. Measure filament voltage at the socket. This is all to
make sure the tubes REALLY get the correct filament voltage to the
filament... It's 28 Amps for two tubes!

Actually in amateur service filament voltage is not that big a worry. I
think the excessive worry comes as a carryover from BC service where
tubes are operated far below ratings and can be operated for almost
9000 hours per year.

In BC service, excessive voltage shows up in a noticable period of time
since the tubes are operated 24/7. The voltages and currents of other
elements are a much lower percentage of maximum compared to amateur
service, and filaments are not cycled. As such, emission failures are a
common failure mechanism.

In amateur service, it takes many years and thousands of fialment
cycles to accumalate the hours a BC tube sees in one year. Tubes in
amateur service much more often fail from seal leakage, element
structure defects, and occasional abuse.

Out of hundreds of tubes I've seen fail in amateur service over the
past 30 years, I can count the number of emission failures on my two
hands.

I'm not saying filaments should be run extra hot, but I am trying to
point out what worries BC stations is often not a concern in amateur
use.

By the way, two of the emission failures I saw were in 8877 tubes that
were operated too cold. The owners had reduced filament voltage below
4.75 volts. NEVER run a MOX cathode tube at reduced voltage unless you
also plan to limit peak emission current.

My advice is to run all tubes in amateur service at RATED filament
voltage or in the allowable tolerance range for voltage.

As for cooling, seal cooling is critical in glass tubes. So is periodic
operation to keep the tubes gettered.

http://www.w8ji.com/vacuum_tubes_and...e_failures.htm

73 Tom

wrote:
wrote:

Doug, a suggestion about "Longevity" for 4-400's (Actually, any
thoriated-tungsten filament power tubes):

These tubes are very sensitive to filament voltage. Make sure the tube
pins and socket are nice and clean and shiny (don't use any
abrasives!).. Measure filament voltage at the socket. This is all to
make sure the tubes REALLY get the correct filament voltage to the
filament... It's 28 Amps for two tubes!


Actually in amateur service filament voltage is not that big a worry. I
think the excessive worry comes as a carryover from BC service where
tubes are operated far below ratings and can be operated for almost
9000 hours per year.

In BC service, excessive voltage shows up in a noticable period of time
since the tubes are operated 24/7. The voltages and currents of other
elements are a much lower percentage of maximum compared to amateur
service, and filaments are not cycled. As such, emission failures are a
common failure mechanism.

In amateur service, it takes many years and thousands of fialment
cycles to accumalate the hours a BC tube sees in one year. Tubes in
amateur service much more often fail from seal leakage, element
structure defects, and occasional abuse.

Out of hundreds of tubes I've seen fail in amateur service over the
past 30 years, I can count the number of emission failures on my two
hands.

I'm not saying filaments should be run extra hot, but I am trying to
point out what worries BC stations is often not a concern in amateur
use.

By the way, two of the emission failures I saw were in 8877 tubes that
were operated too cold. The owners had reduced filament voltage below
4.75 volts. NEVER run a MOX cathode tube at reduced voltage unless you
also plan to limit peak emission current.

My advice is to run all tubes in amateur service at RATED filament
voltage or in the allowable tolerance range for voltage.

As for cooling, seal cooling is critical in glass tubes. So is periodic
operation to keep the tubes gettered.

http://www.w8ji.com/vacuum_tubes_and...e_failures.htm

73 Tom

If oversize filament transformers are used (never know what kind of
surplus junk hams use!) there is a danger of filament failure
due to starting in-rush current. My home brew linear started these
tubes up with a power resistor in series with the primary of the
filament transformer for a few seconds, which was then shorted out
by a relay. A transistor r/c delay circuit did the dirty work.
This lets the tube filaments preheat a bit before applying full
voltage.

Hi, Tom

Thanks for the hint about gas in large glass tubes.
Also thanks for the link to your web site, plenty of wonderful info!

One question please - I have 8 type 3-500Z which
have not been used for 20 years. You mentioned operating
them at low plate voltage and positive grid voltage,
so that the plates show color (for gettering action).

What range of plate voltage/ grid voltage do you recommend,
and what current limiting resistor would you suggest
in the plate circuit?
Do you use a current limiting resistor in the grid circuit also?
I plan a test fixture to recussitate
the 3-500Z tubes before risking them in an amplifier.

73,
Ed Knobloch

wrote:
snip
As for cooling, seal cooling is critical in glass tubes. So is periodic
operation to keep the tubes gettered.

http://www.w8ji.com/vacuum_tubes_and...e_failures.htm

73 Tom