Yes. Maybe we should qualify the statement about testing tubes in a radio by
first defining that the radio has no other problems. After all you wouldn't
expect a good test result from a tube tester that didn't work properly. If you
replace your 1L6 and the radio works you would probably button it up and
pronounce it fixed. Months later when it fails again you know better. For
expensive or rare tubes it's always good insurance to check out the
environment that the tube operates in and see if it's being properly fed
and cared for. Modifications to the set can be detrimental if done improperly.
V
Daniel Schoo (o o)
Electronics Design Engineer ( V )
Fermilab, Batavia, Illinois, USA .......m.m......Dan's Cockatoo Ranch
vvv
1. Getting an old radio working properly may involve a lot more than
just trying to identify a faulty component and replace it. One radio I
tangled with last fall was a set that had been playing fairly well
until it lost most of it's RF sensitivity. Yes, there was a bad screen
bypass cap on the RF preamp stage that took it's feed resistor with it.
But a little investigation showed that the set had all sorts of
problems in other places, and it ended up with 100% replacement of
paper caps and 90% replacement of resistors. Every single circuit in a
fifteen-tube radio had at least one component-related problem, and only
one of those circuits needed a new tube, replaced for a defect that
would have been hard to find on a tube tester (6SQ7 diode quit
conducting after the set had played 15 minutes).
2. Modifications: You are on your own making any modifications to the
original circuit. You have to consider carefully what balance was in
the original design that is not particularly obvious, and that is going
to be upset by changing a circuit parameter. Large communications sets
popular with "hams" are particular victims of half-baked improvements
swiped from sources like QST, the ARRL Handbook, Radio Engineer's
Handbook, and Radiotron Designers' Handbook that are completely wrong
for the radio and screw up its performance royally.
Replacing a selenium rectifier with a silicon diode and adding
resistance to bring voltages back down to the correct values is fairly
safe and easy to do. Adapting a power transformer with a different
rectifier plate voltage is also not too difficult, although you will
need to deal with filter design and resistance dividers so that the
circuits see the correct voltages and source impedances from the power
supply. Fooling around with an oscillator, converter, detector, or
audio voltage amplifier requires a hard-core design and performance
study. My current "straighten out the mess" project involves a local
oscillator that had a voltage regulator tube added to the circuit,
identical to a change shown on a manufacturer's schematic and parts
list. It had very unpleasant side effects on the oscillator/RF
tracking, and made a good-performing, but somehwat unstable, radio into
a very bad performer. I completely removed the ARRL handbook
detector/noise limiter circuit and replaced it with one that presented
the manufacturer's original impedance to the last IF transformer and
which matches the first audio ciruit as designed by the manufacturer.
That made a big difference in the set's sensitivity, selectivity, and
audio response.
Modifications? Hang onto your hat, because you are very likely to
invent trouble where it didn't exist.
--
***********************************************************
Hank van Cleef vanc...@bga.com vanc...@tmn.com
***********************************************************
I agree entirely, and think maybe something needs to be added to the FAQ:
"When you bring something home, FIRST, before even plugging into a variac to
reform the caps, check for Partly Baked Mods."
Two of the first five T/Os I pulled the chassis on had Si rectifiers and the
stock resistor. Will work just great for a while. Am reminded of the story
in one of Ripley's books of faqirs in India who would cure partial blindness
by draining the eye. Would pocket the rupees and be off. Victem would see
for about a day after which they were permanently and completely blind. Now
I am going to have to pull every chassis I have and check since tubes ain't
cheap.
Am coming to the opinion that modern tubes can take a lot of abuse so long
as the filaments are pampered and fed just what they need and no more. Turn
on seems to occur at about 90% of rated filament voltage for a good tube.
Does anyone know of a reason not to underdrive say 5% nominal as a matter of
course for longetivity ? Does this thought make any sense (have no actural
tables for tubes) ?
A. Padgett Peterson, P.E.
Cybernetic Psychophysicist
Totally Obsessed with TransOceanics
My other car is a Pontiac too
We also walk dogs
PGP 2.7 Public Key Available
This *was* done in the vacuum tube computer era, when getting 20,000
tubes all working simultaneously was quite a feat.
In "Reference Data for Radio Engineers" (Sams, 1979) there is an
"actuarial graph" for bright-tungsten filaments. It shows an emission
decrease of 30% and a life increase of 90% at 5% below rated voltage.
--
John Doty "You can't confuse me, that's my job."
j...@space.mit.edu
: Am coming to the opinion that modern tubes can take a lot of abuse so long
: as the filaments are pampered and fed just what they need and no more. Turn
: on seems to occur at about 90% of rated filament voltage for a good tube.
: Does anyone know of a reason not to underdrive say 5% nominal as a matter of
: course for longetivity ? Does this thought make any sense (have no actural
: tables for tubes) ?
It makes sense, though I have to admit that as someone who started out
when tubes were plentiful and cheap, I have a hard time getting the idea
into my heart that later (octal and miniature) tubes are scarce. :)
Someone else has already mentioned use of reduced filament voltage in
vacuum-tube computers. In big broadcast transmitters, we *routinely* run
the filaments in the finals almost as low as they'll go; the procedure is
to put in a new tube, and run it at rated filament voltage for at least a
day and no more than four days; then you back off the voltage 'til the
power output just begins to drop*, run it up a tad past that point, and
leave it there. Over time, you keep an eye on it and run it back up as
emission fades. The 9.5V nominal on the PAs in the big rig usually start
out after burning in at 8.8-9.0V, and end up a couple of years later
approaching 10V. (Don't even ask about current, around 150A!)
Those tubes cost around US$6000.00 each, so coddling them is worth it.
Before they were second-sourced, prices got as high as $7500! ...You
can see why I don't have the same emotional response to doing that for a
$25.00 receiving tube; that doesn't mean it's not a good idea.
73,
--Bobbi
______________________
*Actually, in television you look for "sync compression," which
amounts to the same thing, the tip of sync being the highest-power part
of the transmitted waveform. The power meters are heavily damped, so
looking at an oscilloscope display of the off-air signal is a more
responsive indicator.
This is pretty much the procedure for scanning electron microscope
filaments where a replacement filament may not cost a great deal but your
instrument might be down for the better part of a day, or more, during
replacement.
sdb
Not all of them, just the ones that were pricy and had few applications to
begin with. Was told a loooong time ago that I am too "future oriented" and
spent a lot of time designing control systems for fighter aircraft where
reliability is of paramount importance (have also designed more than a few
"overdrive" systems for those instances where mission accomplishment is
worth more than a U$50 million aircraft). Also used to road race in my
illspent yout.
Put them all together and it comes out to "never stress things any more than
you need to so when you need to, there is a reserve". (Blew an injected
Covette engine sky high once - all the smoke in the world type - fifty feet
*after* crossing the finish line ahead of a 906 Porsche. Was worth it).