I have an old 12AX7 that I tested recently. It tests as "good" on an
I-177 tester, with no shorts. However, on the gas test, the needle
moves upward two scale divisions, which amounts to a test failure.
(Less than one division is acceptable. The needle is moving from 100
to 200 on the meter, with 150 being the first division in between.)

The question: What, exactly, does this mean as far as the performance
of the tube is concerned? Will it work, but with degraded sound? Will
it show no apparent symptoms? Will it short out completely in the amp?

Any input would be much appreciated.

Thanks,

--E

Have you connected a heater of this tube to a power supply/ heater
transformer for several hours? If not- do so- you'll give the tube's getter
a chance to absorb these gases.

(Ether) said:

I have an old 12AX7 that I tested recently. It tests as "good" on an
I-177 tester, with no shorts. However, on the gas test, the needle
moves upward two scale divisions, which amounts to a test failure.
(Less than one division is acceptable. The needle is moving from 100
to 200 on the meter, with 150 being the first division in between.)

The question: What, exactly, does this mean as far as the performance
of the tube is concerned? Will it work, but with degraded sound? Will
it show no apparent symptoms? Will it short out completely in the amp?

Any input would be much appreciated.

The presence of gas in a tube could mean the getter is on its way out,
unable to bind the gaseous residu that's coming off the metal parts
inside the tube when hot.
The vacuum of the tube will get worse and the tube's parameters will
alter. It won't blow (yet), but it's most likely to produce more
distortion.

Sometimes, a tube that's been on the shelf for a long time, will read
bad on gas initially.
After some use, the getter ( if still intact) will have done its work
and the tube will behave like it should.

In some instances, the condition of the getter can be checked by its
color ( it may have turned brown) and even tree-like branches in the
getter.
A healthy getter is usually silverish. A black "coating" may obstruct
the view though.

Try "rejuvenating" the getter by heating the tube at a slightly higher
voltage (say 8 volts or so) and applying a moderate anode voltage.
If the gas readings don't improve after half an hour under this
conditions, one may assume the tube is lost.

--
Sander deWaal
"SOA of a KT88? Sufficient."

Ether wrote:

I have an old 12AX7 that I tested recently. It tests as "good" on an
I-177 tester, with no shorts. However, on the gas test, the needle
moves upward two scale divisions, which amounts to a test failure.
(Less than one division is acceptable. The needle is moving from 100
to 200 on the meter, with 150 being the first division in between.)

The question: What, exactly, does this mean as far as the performance
of the tube is concerned? Will it work, but with degraded sound? Will
it show no apparent symptoms? Will it short out completely in the amp?

Any input would be much appreciated.

Thanks,

--E

Even though I make tubhe amps for a living, I don't use tube testers.

I place the tube in a test circuit, and measure the anode current and
voltage
for a given load, and then measure the gain while watching the waveform on
the CRO.

If I measure a 12AX7 and with a 220k RL I get say a gain of about 75, then
the tube is OK for gain.
I will have a grid bias R of say 1M, and if there is more than 0.3
positive at the grid, the tube is slightly stuffed,
perhaps gassy. Usually tubes with a positive grid voltage measurement are
also noisy.

I sometimes measure the gain for two different values of RL.
Since U = Ra x Gm for all tubes, and Gain = U x RL / ( Ra + RL ),
I can work out the U, Ra, and Gm for any tube I measure, and see if it
tallies with the
data I have for such tubes. The pocket calculator which was invented after
the tube testers
tell me more than the tester ever will.

I also measure for noise. This requires the anode output of the 12AX7 is
coupled to a known low noise
preamp with a gain of say 1,000, and bandwidth of 20 kHz.
The 12AX7 grid must be grounded.
The noise of the preamp will be swamped by the noise of any additional
tube stage added
in front of the preamp, so that if I measure 1 vrms of noise at the preamp
output,
there is 1 mV at the preamp input, which couldn't have been generated by
low noise preamp
which itself will have only say 2 uV of input noise.
If the calculated 1mV of noise is divided by the 12AX7 voltage gain of say
75,
then there must be 13.3 uV of equivalent input noise at the grid of the
12AX7.

The noise voltage of say 1v at the output of a preamp should easily be
measured
by a voltmeter, which reads vrms signal voltages up to at least 20 kHz.
A DVM only reads to about 3 kHz, so something better is really needed.
If you don't know what noise is, plug a sample of what you are measuring
into
a sound amp and speaker, and it should sound like a large rumbly water
fall.

The test circuit should have a heater supply with a CT winding of 6.3vrms,

or a dedicated winding with a 100 ohm pot across it with the pot wiper
taken to 0V.
It should be possible to minimise the hum by nulling it with the pot.
But if not, there is a problem with the heater-cathode insulation.
If the heater is leaking current to the cathode, the noise will sound
dominated by the
hum from the mains frequency, and in fact the tube under test shouldn't
sound
hummy.

With a good 12AX7, the equivalent grid input noise, EIN, is usually
between 1.2 uV and 2.5 uV,
with the anode current of 0.6 mA, which seems optimal for this tube type.
Any 12AX7 with 13.3 uV is quite stuffed, and into the bin it goes.
Tube testers don't test everything.
Tubes which measure noisy when first plugged in won't ever improove.

After leaving an initially good tube sitting there for 1/2 an hour, I go
back to see if its
got any worse, or has any flicker noise problem, or increased +ve grid
voltage.
If it is just as good as it was after 5 minutes, it goes into the
box marked "good noise", but before they are tested for microphony.

Some gassy tubes might work a bit, but no such tubes are permitted
to stay in any amps of mine, they go out with the rubbish.

Microphony can also be tested in the same set as described, and
a badly microphonic tube will make a loud ringing sound at some audio F
with a long duration if gently struck with a pencil, and the result
listened to with a monitoring amp and speaker.
After testing 50 tubes, you'll get to know what's microphonic and what
isn't.
Low micro tybes will still produce a thud, but not the real loud twang
from a crook tube.
Such a tube is usually old, and had its mica plates go loose inside the
tube to allow
physical movement in the electrodes.
Some old tubes are both noisy and micro/p so out they go,
but some which are mildly micro/p but not noisy can be used
in power amp stages where the micro/p don't matter
since the SNR in power amp tubes is usually OK.

Its important to me that when I supply someone with an amp with NOS tubes,

they are actually still in new condition,
and have not gone gassy, and thus noisy in the years spent locked up in a
box on a shelf.

Much old gear which I find myself restoring may have had the tubes
replaced 30 years ago
and it didn't cure the noise or other problem, so the item was parked,
complete with a set of
perfectly fine tubes, so one wouldn't know if one didn't test.....

I have never found that a gassy or noisy or faulty preamp tube ever causes
damage in the circuit its in.
The worst case is that the tube shorts, and pulls the anode voltage to 0V,
therefore
heating the anode R a bit.
Spurious shorts or arcing internally is rare, but these are a worry if the
preamp
is hooked up to a SS amp which might ****e itself if it gets a pulse of
100v from a faulty tube circuit.
I have seen this caused by a faulty tube Counterpoint preamp, and the
problem was the
SS regulation of the preamp's 400v B+ supply, *and* the SS regulation of
the DC heater supply.
I try to have no regulators in my amps, or as few as possible, since they
are another darn thing that can go wrong.


Patrick Turner.

"Sander deWaal" wrote in message
...
The presence of gas in a tube could mean the getter is on its way out,
unable to bind the gaseous residu that's coming off the metal parts
inside the tube when hot.

If there's any metallic getter left, it's still doing its job. Except for
argon, of course. But since atmospheric gas is but 1% Ar, it would have to
go through 100 times as much air to get there... a pretty well worn tube to
say the least.

Heating the tube won't really do much, as the barium is about as reactive at
either temperature. I mean, yeah it's more reactive, but with so little
[absorbable] gas present, it doesn't really matter.

But you can if you want to... I built a little induction heater (see my
website under Electronics) and can heat 12AU7 plates red hot...hehe...

Tim

--
"I've got more trophies than Wayne Gretsky and the Pope combined!"
- Homer Simpson
Website @ http://webpages.charter.net/dawill/tmoranwms