Which tube tester?
 

Hi fellas,

I find myself in the need for a tube tester and am watching three at the
moment:

- Heathkit TC-2
- Eico 625
- Accurate Instrument 157

The Accurate Instrument is powered straight from the line which unnerves
me a bit :-\

Which of the three have you any experience as reasonably reliable???

Many thanks!

--
Gregg "t3h g33k"
http://geek.scorpiorising.ca
*Ratings are for transistors, tubes have guidelines*

Gregg wrote:
Hi fellas,

I find myself in the need for a tube tester and am watching three at the
moment:

- Heathkit TC-2
- Eico 625
- Accurate Instrument 157

The Accurate Instrument is powered straight from the line which unnerves
me a bit :-\

Which of the three have you any experience as reasonably reliable???

Many thanks!

Which of the three are Dynamic Mutual Conductance tube testers?

The Heath, and the Eico are emission testers. I know nothing about
the Accurate Inst. unit.

Emission testers are basically only good for telling if the cathode
is still active.

If you really want to test tubes, you have to have a dynamic mutual
conductance tester. There were plenty of examples available from
Hickok.

-Chuck

Many people believe that a tube tester needs to be a transconductance type
to be accurate. I have numerous examples of both.

My experiences through hundreds of tests are that the emission tester is
accurate in almost all cases. As a matter of fact, I usually grab whichever
tester is handy. If a tube is bad, either tester will usually tell you.
Emission testers may actually be better in some situations - such as
determining life left in the tube.

I can recall only two tests in hundreds where the transconductance tester
led me to a problem hidden by the emission tester. In one case, a Drake
transceiver had a tube that had developed cathode interface and had about a
1 megohm internal leakage, which had developed because of a circuit designed
by Drake, which had a very high impedance grid circuit that recommended by
tube designers. Although I had substituted another tube, it also had
cathode interface. The transconductance tester allowed me to do some
detective work to find the problem. A third tube solved the problem.

In the 30's, many people built their own tube tester. I have done that.

I may get a lot of flack for my opinion, but I have my flack shield on and I
can only report what I experience.

Colin K7FM

COLIN LAMB wrote:
Many people believe that a tube tester needs to be a transconductance type
to be accurate. I have numerous examples of both.

My experiences through hundreds of tests are that the emission tester is
accurate in almost all cases. As a matter of fact, I usually grab whichever
tester is handy. If a tube is bad, either tester will usually tell you.
Emission testers may actually be better in some situations - such as
determining life left in the tube.

Poppycock! Transconductance testers combine the functionality of an emissions
tester with a simple test for low frequency AC gain.

Further, transconductance testers have a sensitive test for gassy tubes.
Something that an emissions tester cannot do (except in extreme cases).

Your experience just shows that old tubes are rather reliable.

I can recall only two tests in hundreds where the transconductance tester
led me to a problem hidden by the emission tester. In one case, a Drake
transceiver had a tube that had developed cathode interface and had about a
1 megohm internal leakage, which had developed because of a circuit designed
by Drake, which had a very high impedance grid circuit that recommended by
tube designers.

The condition known as "cathode interface" comes about because the interface
between the metal cathode substrate, and the highly em missive oxide cathode
layer, becomes degraded. The interface develops a higher DC resistance than
it is supposed to have. The problem is somewhat vexing because the capacitance
of the cathode interface remains as high as when the tube was new. The most
noticeable characteristic of "cathode interface" is that the tube has reduced
DC gain, and normal HF AC gain. If you put a square wave through a DC coupled
tube that has "cathode interface" the the top and bottom horizontal lines of
the square wave will droop towards zero. If you compensate the tube circuit
so that no droop occurs, there will be a pronounced spike at the leading and
trailing edge of the squarewave's rise and fall time. Cathode interface was
rarely a problem before the advent of high performance vacuum tube oscilloscopes.

What you discovered was a gassy tube. When gas enters a vacuum tube, the gas
causes a current path from the cathode to the grid. This makes the grid go
more positive than normal. In a circuit that has high grid impedance, the
tube will significantly change its bias point. A fairly important thing in
grid leak biased tubes (that are often found in receivers).

Although I had substituted another tube, it also had
cathode interface. The transconductance tester allowed me to do some
detective work to find the problem. A third tube solved the problem.

In the 30's, many people built their own tube tester. I have done that.

I may get a lot of flack for my opinion, but I have my flack shield on and I
can only report what I experience.

Colin K7FM

If you are going to spend the money and waste the space on a tube tester,
you ought to get one that will actually do a good job of testing tubes. Today
that is even more important than it was 40 years ago... tubes have become
obscenely expensive, and most tubes you will be exposed to have long since
passed their expected useful lifetimes. Conditions like gas are quite common,
and should be tested.

-Chuck

Hi,

- Heathkit TC-2
- Eico 625
- Accurate Instrument 157

The first two have identical circuitry; pick whichever is in better
shape or cheaper. Avoid the Accurate at any price.

Alan

Behold, Alan Douglas scribed on tube chassis:

Hi,

- Heathkit TC-2
- Eico 625
- Accurate Instrument 157

The first two have identical circuitry; pick whichever is in better
shape or cheaper. Avoid the Accurate at any price.

Alan


Thanks all!

:-)

--
Gregg "t3h g33k"
http://geek.scorpiorising.ca
*Ratings are for transistors, tubes have guidelines*

Some tube manufacturers specified a maximum grid leak resistance. It was
often about 2 million ohms. Drake found that much higher resistance was
great for avc performance.

However, over the long run, it was also asking for trouble.

In theory, what Chuck says is true - that transconductance testers are
better than emission testers. However, I can just report that in hundreds
of tests I have just found a couple of cases where there was any practical
difference.

I often end up getting sucked into helping others get their receiver going
again. I do not charge for it and I want to get back to my own projects.
The emission tester is much faster to use and I do not really care if the
tube is 79% good or whether it has 10% less gain. The radio does not work
when I start and I am looking for black or white, not shades of gray.
Often, it is not the tubes, but rather another component. The tube tester
simply eliminates the tube as the culprit for total failure.

I suppose if I were looking for a tube with the best performance for a 2
meter moon bounce preamp, I would use a transconductance tester. However,
in that case, I would not use a tube at all.

I could not imagine running all the tubes of an S-85 through the test
procedure for a transconductance tester - unless I had to.

Colin K7FM

COLIN LAMB wrote:
Some tube manufacturers specified a maximum grid leak resistance. It was
often about 2 million ohms. Drake found that much higher resistance was
great for avc performance.

However, over the long run, it was also asking for trouble.

In theory, what Chuck says is true - that transconductance testers are
better than emission testers. However, I can just report that in hundreds
of tests I have just found a couple of cases where there was any practical
difference.

I often end up getting sucked into helping others get their receiver going
again. I do not charge for it and I want to get back to my own projects.
The emission tester is much faster to use and I do not really care if the
tube is 79% good or whether it has 10% less gain. The radio does not work
when I start and I am looking for black or white, not shades of gray.
Often, it is not the tubes, but rather another component. The tube tester
simply eliminates the tube as the culprit for total failure.

I suppose if I were looking for a tube with the best performance for a 2
meter moon bounce preamp, I would use a transconductance tester. However,
in that case, I would not use a tube at all.

I could not imagine running all the tubes of an S-85 through the test
procedure for a transconductance tester - unless I had to.

Colin K7FM

Hi Colin,

The setup for a transconductance tester has only one extra step over the
emission tester: you have to set the grid bias. If you do the gas test,
you get two more steps: set plate resistor to specified plate current,
press button.

Emissions tester:

1) set heater voltage
2) set socket connections
3) plug in tube (start warming up)
4) set line voltage adjustment rheostat
5) set plate load resistance pot.
6) do shorts test
7) press emission test button

Transconductance tester:

1) set heater voltage
2) set socket connections (two rotary selector switches on Hickok)
3) plug in tube (start warming up)
4) set line voltage adjustment rheostat
5) set plate load resistance pot
6) set grid bias pot
7) do shorts test
8) press Gm test button

+ gas test:

1) press Gm test button and hold
2) adjust grid bias for 100umhos
3) press Gm and Gas test


Since most of the time you don't need to do the gas test, the additional
step is hardly a noticable increase in effort considering the increased
quality of the test.

I have tested every tube in every Tektronix scope that has come my way,
and it is no more burden than doing the test in an emission tester would
be. Since I have to pull all the tubes to wash the scope, I might as well
test them. The complete test for all tubes in a 585A scope takes maybe
20 minutes... most of the time is spent waiting for heaters to warm up.

-Chuck

Many transconductance testers have multiple switch decks to move each
element around. This represents a number of additional steps, over the
emission tester.

And, I often can get the radio back up and running in 20 minutes - and speed
is important since I do not charge for it.

Chuck no doubt is more thorough than I am - but I am only servicing an old
radio and not a 50 Mhz precision scope. Incidently, one of my close friends
designed many of the circuits in the 585, and single handedly designed the
519 and 130, and I have a row Tek scopes and other Tek equipment.

Colin K7FM

COLIN LAMB wrote:
Many transconductance testers have multiple switch decks to move each
element around. This represents a number of additional steps, over the
emission tester.

And, I often can get the radio back up and running in 20 minutes - and speed
is important since I do not charge for it.

Chuck no doubt is more thorough than I am - but I am only servicing an old
radio and not a 50 Mhz precision scope. Incidently, one of my close friends
designed many of the circuits in the 585, and single handedly designed the
519 and 130, and I have a row Tek scopes and other Tek equipment.

Colin K7FM

Hi Colin,

My first emission style tester had a bank of switch levers, one
for each of the 12 possible pins on a tube. Each switch had
several possible positions:

1) Heater +
2) Heater -
3) Plate
4) Cathode
5) Open

The Hickok testers arrange the pins by using a bank of rotary switches,
with a separate rotary switch for each of the following elements:

1) Filament
2) Filament
3) Cathode
4) Grid
5) Plate
6) Screen
7) Supressor

For setup, it becomes a 6 of one, half-dozen of the other situation.
Either way, you have to account for all of the elements in the tube.

As to fixing a radio in 20 minutes, or less... Sure, since most radios
have fewer than 20 tubes (KWM-2 has 18). A minute a tube is very realistic.

I've done it both ways, and for me it takes as long to use an emissions
tester as it does to test the tube correctly with a transconductance
tester. There isn't a significant difference in the number of steps
either way. There is a major difference in the quality of the test.

-Chuck

I'll be a contrarian and ask, Do you really need a tester?

I have restored many tube radios, including some boatanchors. I own a couple
of testers, but don't use them often. Usually, the tester only tells me
something I could find out another way ("That's a dud -- duh").

As Colin mentioned in another post, marginal differences between tubes may
not make any noticeable difference in how the radio performs. Many times, I
have replaced a "weak" tube with a new one, only to find that the radio
performs exactly the same.

There are exceptions, of course. The horizontal and vertical circuits in a
TV may be fussier about tube condition than the circuits in a simple,
consumer-grade radio.

One case where a tester is very useful is if someone gives you hundreds of
loose tubes. It can quickly sort out the duds from the keepers.

The best test of all is in a working circuit for which that tube was
designed.

You can test tubes without a tester if you have some working radios and
known-good tubes in the house.

For example, you can test a suspect tube by putting it in the same type
application in a working radio. If the radio still works, the tube is
"good."

You can also do the reverse. That is, substitute a known-good tube in your
subject radio and see whether the radio works better. If so, the suspect
tube was "bad." If not, it was "good" (or at least you have ruled out that
tube as the *sole* problem -- many unrestored tube radios have multiple
problems such as leaky capacitors.)

So, ask yourself, Is this something that I really need and will use often?
Or will it just be another Thing cluttering up my life?

If you're determined to get one, I'd buy a cheap emission type tester and
try it out for a while. You can always buy a fancy tester later, if that
seems necessary.

Just my $0.02 :-)

Phil Nelson
Phil's Old Radios
http://antiqueradio.org/index.html

Behold, Phil Nelson scribed on tube chassis:

I'll be a contrarian and ask, Do you really need a tester?

Hi Phil,

I do no repairs, but design from scratch, so yeah, it'd be handy ;-)

I have over 1,000 tubes with many onesies, so I can't really sub too
easily. Example, a tube regulator drove me nutz. It was a bad 6AU6,
'cause when I rewired for a 6CB6, it worked :-)

Cheers!

--
Gregg "t3h g33k"
http://geek.scorpiorising.ca
*Ratings are for transistors, tubes have guidelines*

Switching tubes in a radio is a good simple test, but still requires some
basic organization. I remember using a radio to test a batch of 6CB6 tubes.
I tested about ten old unboxed tubes and found two of them were bad.
Unfortunately, I had grouped the tubes into different groups of similar
performance. A couple bad, a couple very hot, some middle of the road and
so on. When I got done, I had forgotten where the bad tubes were, so had to
start over again.

Before a friend died, he requested that I purchase his collection of 14,000
used tubes - which I did. Fortunately, my wife was away and I got them home
and stored before she got home. It was weeks before she figured out there
was something new in my collection.

But, having 15,000 tubes (I had 1,000) is something like a curse. The last
time I needed a particular tube, I could not find it and had to buy one.
Found it 2 months later.

73, Colin K7FM

Hi,
Colin wrote:

Incidently, one of my close friends
designed many of the circuits in the 585, and single handedly designed the
519 and 130, and I have a row Tek scopes and other Tek equipment.

The 130 L-C meter? Great little instrument: I still have one on my
bench, and examples of both the old and newer models.

Alan

Yep, the 130 L-C meter. Cliff needed something like that for a project and
there was nothing available, so he whipped one up. It was so useful, the
other engineers soon wanted one. Then Product Design got ahold of it.

I have his personal 575, too.

Colin K7FM

The RCA Tube Manuals have an excellent summary of tube testers.

Many transconductance tube testers test the tube under static conditions,
using dc on the grid. This is a simpler technique, but not as accurate as
tests made under actual conditions. That is the reason that the best test
is to install the tube in the circuit and see if it works properly. Some of
the very expensive tube testers, like the Western Electric card tester, will
test the tube under conditions likely to be seen in the actual circuit. The
problem with one of those testers is that the tester is one large heavy
suitcase and the cards used in the tester is one large heavy suitcase.

The dynamic tube tester uses ac on the grid and an ac ammeter of the
dynamometer type. Transconductance is equal to the ac plate current divided
by the input-signal voltage.

The summary of the RCA text is as follows:

"The tube tester, therefore, cannot be looked upon as a final authority in
determining whether or not a tube is always satisfactory. Actual operating
test in the equipment in which the tube is to be used will give the best
possible indication of a tube's worth."

Colin K7FM

COLIN LAMB wrote:
The RCA Tube Manuals have an excellent summary of tube testers.

Many transconductance tube testers test the tube under static conditions,
using dc on the grid. This is a simpler technique, but not as accurate as
tests made under actual conditions.

I have never seen a DC transconductance tester. All of the Hickok units
that I am aware of are "dynamic mutual conductance" testers.

That is the reason that the best test
is to install the tube in the circuit and see if it works properly. Some of
the very expensive tube testers, like the Western Electric card tester, will
test the tube under conditions likely to be seen in the actual circuit. The
problem with one of those testers is that the tester is one large heavy
suitcase and the cards used in the tester is one large heavy suitcase.

Agreed, but I'd sure love to have one anyway.


The dynamic tube tester uses ac on the grid and an ac ammeter of the
dynamometer type. Transconductance is equal to the ac plate current divided
by the input-signal voltage.

That is the method used by all of the testers based on the Hickok designs.
Testers designed by Hickok, but made by others, include, I177, TV2, TV7,
and I believe the RCA suitcase testers.

The summary of the RCA text is as follows:

"The tube tester, therefore, cannot be looked upon as a final authority in
determining whether or not a tube is always satisfactory. Actual operating
test in the equipment in which the tube is to be used will give the best
possible indication of a tube's worth."

That is, of course, the reason why any tube tester is basically a luxury. They
do a good job of showing when a tube deviates drastically from the norm, but
say nothing about whether the tube will/won't work in an actual circuit. Tektronix
warns technicians not to replace tubes just because they test weak, but rather
to replace tubes that prevent the scope from meeting alignment specifications.

-Chuck

Hi,
The static transconductance test is more commonly known as the
grid-shift method. It's very old, dating from the 1920s, but can be
extremely accurate, depending on the amount of grid shift (easy with
modern digital meters). Grid-shift went out of favor in the US in the
1930s but most of the British AVO models use it, except with rectified
AC on the plate instead of DC.

The Hickok AC-47 from 1930 had a dynamometer movement (6.25mA F.S.)
and is actually a very fine instrument but was dropped in the mid-30s
because of expense and the difficulty of keeping it up to date with
adapters. Other than one Sensitive Research design, the last
dynamometer model was Hickok's "laboratory model" 700. That beast was
obsolete the day it appeared in 1951.

The Hickok Cardmatic models---123, military USM-118B, Western
Electric KS 15874---can measure either emission or transconductance,
sometimes both (on TV sweep tubes for instance where extra cards are
provided for each test).

Alan