I have a fully functional TV-7 and have recently acquired a Triplett
3444. After replacing the metallic rectifiers, the things work very
well. I noticed a post, here, a few days ago regarding the conversion
of the scale factors on the TV-7 meter to GM. I noted the GM ranges and
decided to compare the TV-7 with the 3444. Before even turning on the
equipment, I selected a new 6DC6 tube. Using the test criteria in the
TV-7 book, I calculated the expected minimum GM for the TV-7. I then
looked at the direct reading minimum GM value on the 3444 rollchart.
They are nowhere near each other. I then checked the new tube in both
tube testers to see if there was any correlation.. there was none. I
did find that both testers agree a tube is good and conversely showed a
known bad tube to be bad, when it was.

May I therefore assume that the parameters each manufacturer (Triplett
and Hickok, in this case) selected for a given test may be differing,
but in each test scenario if that manufacturer's respective minimum
value is obtained, the tube can be assumed to be "good"? The above
would dispose of a notion that are (were) published tube test parameters
that the manufacturers would implement in their designs. The big RCA
manual does allude to test standards but maybe they were for them
(RCA). Any insights would be helpful. I've been testing tubes since
the 50's but never had the opportunity to do this type of comparison.
Regards de K3HVG

Hi,

It has been stated since the first tube tester was made that the only
true test of a tube's functionality is in the circuit it is to operate.

Virtually every tube tester, even different models by the same
manufacturer, has different operating points for its tests. Since Gm
is very defendant on plate voltage, current, grid drive, etc. Unless
the manufacturer of tube tester "A" used the exact same operating points
as the manufacturer of tube tester "B" there will be different Gm's for
the same tube.

The test equipment that RCA used in its factories and labs was certainly
more sophisticated and rigorous than what Hickok made for the service
industry.

For instance, on the Hickok testers, the filament transformer uses the
same 6.3V tap for a 6AU6 as it does for a 6AS7. One tube draws 300ma,
the other draws 2.5A. The only way the Hickok testers have of
compensating for the vastly different current draw is to adjust a pot
in series with the power transformer so the primary of the transformer
has the correct voltage. As a consequence, the 6AU6 may be operating
at 6.4V on the filament, and the 6AS7 only 5.5V. The 6AS7 is operating
outside of the manufacturer's specs. So are all of your favorite big
beam power tubes: 6L6, 5880, 6550, KT88, ...

Tube testers are a compromise. The manufacturers of the testers try
out a lot of different tubes to create their tube databases. The Gm
values bear no resemblance to the values found in the tube manuals.

-Chuck Harris

I'm sure that Alan Douglas can give you droves of information on this
subject, or you could buy his highly regarded book.

k3hvg wrote:
I have a fully functional TV-7 and have recently acquired a Triplett
3444. After replacing the metallic rectifiers, the things work very
well. I noticed a post, here, a few days ago regarding the conversion
of the scale factors on the TV-7 meter to GM. I noted the GM ranges and
decided to compare the TV-7 with the 3444. Before even turning on the
equipment, I selected a new 6DC6 tube. Using the test criteria in the
TV-7 book, I calculated the expected minimum GM for the TV-7. I then
looked at the direct reading minimum GM value on the 3444 rollchart.
They are nowhere near each other. I then checked the new tube in both
tube testers to see if there was any correlation.. there was none. I
did find that both testers agree a tube is good and conversely showed a
known bad tube to be bad, when it was.

May I therefore assume that the parameters each manufacturer (Triplett
and Hickok, in this case) selected for a given test may be differing,
but in each test scenario if that manufacturer's respective minimum
value is obtained, the tube can be assumed to be "good"? The above
would dispose of a notion that are (were) published tube test parameters
that the manufacturers would implement in their designs. The big RCA
manual does allude to test standards but maybe they were for them
(RCA). Any insights would be helpful. I've been testing tubes since
the 50's but never had the opportunity to do this type of comparison.
Regards de K3HVG

Thanks to all who provided the answers! Tube testing, it appears, includes
a touch of art, along with the science!

k3hvg

Hi,
Chuck replied:

Tube testers are a compromise. The manufacturers of the testers try
out a lot of different tubes to create their tube databases. The Gm
values bear no resemblance to the values found in the tube manuals.

That pretty well sums it up. Incidentally the 3444 is a really
good tester, that runs tubes closer to tube-manual conditions. But as
noted, any Gm tester will rank tubes in about the same order, good to
bad.

I normally use a TV-7B and a KS15874 Cardmatic.

73, Alan

Subject: Tube tester query?
From: k3hvg
Date: 1/31/04 2:53 PM Eastern Standard Time
Message-id:

Thanks to all who provided the answers! Tube testing, it appears, includes
a touch of art, along with the science!

k3hvg

To this I add for your edification an article posted by Chris Haedt, who knows
tube testers better than most anybody except for Alan Douglas. He took a known
good 12AU7 and ran it through a large number of tube testers. He found the kind
of variation you did in your smaller sample. The article should be at
www.hickoktubetesters.com. The site may be down. If it is, let me know and I'll
send the article as an attachment.

Avery W3AVE
Potomac, Md.