I've finished building it...what an excellent tester...... except when
testing el34's. I only have used el34's; 20 of them, Sovtek, Chinese (mesa
boogie) but mainly Phillips and Telefunken and they all test around 7-8
Ma/v. Data says anode current should be 100Ma where the rat tester says use
50 Ma (this is the max that it can manage) are all my EL34's duds or is it
the rat tube tester unable to supply enough current. I have many many used
tubes and have found that even tubes used every day all day for 30 yrs (
6V6's in a cinema amp) can still come very close to new tube specs. Its hard
to beleive that ALL my EL34's are duds. as they sound great in my son's
guitar amp.

Zenon wrote:

I've finished building it...what an excellent tester...... except when
testing el34's. I only have used el34's; 20 of them, Sovtek, Chinese (mesa
boogie) but mainly Phillips and Telefunken and they all test around 7-8
Ma/v. Data says anode current should be 100Ma where the rat tester says use
50 Ma (this is the max that it can manage) are all my EL34's duds or is it
the rat tube tester unable to supply enough current. I have many many used
tubes and have found that even tubes used every day all day for 30 yrs (
6V6's in a cinema amp) can still come very close to new tube specs. Its hard
to beleive that ALL my EL34's are duds. as they sound great in my son's
guitar amp.

It doesn't seem that the tester is excellent if the highest Ia available is 50
mA,
so why not extend its capabilities?

EL34 (and EL84) are supposed to measure 11 mA/V,
but at 50 mA, not many EL34 manage it.
Gm, Ra and U are the 3 parameters of which the U is the most constant for all
tubes,
since it depends on the element spacing, which is fixed.
But the Gm and Ra values are quite variable. At low bias currents, Ra tends to
rise.
Now Gm = U / Raor all tubes.
So if the U of EL34 is 130, and Ra is 15k at low bias current, then Gm = 130 /
15,000 = 0.00866 amps per volt,
= 8.7 ma/V.
But at 100 mA, Ra has risen to 12k, so Gm = 10.8 ma/V.

I don't use a valve tester.

I place the tube in a circuit where its wanted to do a certain thing, and
make sure it works OK with low noise and microphony, and if the tube is OK,
the gain it will have will be what is expected. The CRO tells me more about
these
than a tester will.
In an old radio, the IF amp is often a vari-mu tube, whose Ra and
Gm vary enormously, but where I see the set generating
a healthy agc negative bias on a small input signal, all is probably well.
Frequency converters sometimes wear out as fast as power tubes, but replacing a
suspect dud
with a known OK tube soon finds out if its the tube or surrounding circuitry at
fault.
Its mostly the surrounding circuit.
Tubes often get the blame for the failures of bits around them.
I do have a one channel preamp where small triodes can be tried for gain
with different loads, and we know that for all tubes,
voltage gain, A, = U x RL / ( RL + Ra ).
With two different loads, two different gains can be measured, and the equations

above can be filled out so by very simple algebra the values of the two
unknowns,
Ra and U can be worked out.
Then GM can be worked out from the formula above.

A noise and distortion meter also tells me which tubes are best in a given
circuit.

My ears tell me which sounds best.

Its possible to have a CCS load to any given tube under test,
and with such a load, U = voltage gain for all tubes.
When testing pentodes such as 6AU6, the U may be several thousand,
since Ra is hundreds of thousands of ohms, and Gm is 3 to 5 ma/V.
So using a CCS means measuring gain could be dicky, since
the input voltage for 10vrms output of gain might only be 3.3 mV.
But using the two load method, where one load is 50k, and the other is 10k,
it is all a bit more accurate.
Its not hard to set up a tube with a DC load resistor of say 50k,
and then have a cap coupled load to reduce the load value to 10k,
say with 12.5k added to make a parallel load of 10k.
This keeps the DC conditions identical for the test,
and thus the test is valid. Keep the anode signal small, and well away from
serious thd.

Patrick Turner.

Thanks for the info. (I'll have to read the last bit about CCS a few times
so that it sinks in) Circuit looks like it can be modified easily to give
100Ma. I'll give it a go.

"Patrick Turner" wrote in message
...


Zenon wrote:

I've finished building it...what an excellent tester...... except when
testing el34's. I only have used el34's; 20 of them, Sovtek, Chinese
(mesa
boogie) but mainly Phillips and Telefunken and they all test around 7-8
Ma/v. Data says anode current should be 100Ma where the rat tester says
use
50 Ma (this is the max that it can manage) are all my EL34's duds or is
it
the rat tube tester unable to supply enough current. I have many many
used
tubes and have found that even tubes used every day all day for 30 yrs (
6V6's in a cinema amp) can still come very close to new tube specs. Its
hard
to beleive that ALL my EL34's are duds. as they sound great in my son's
guitar amp.

It doesn't seem that the tester is excellent if the highest Ia available
is 50
mA,
so why not extend its capabilities?

EL34 (and EL84) are supposed to measure 11 mA/V,
but at 50 mA, not many EL34 manage it.
Gm, Ra and U are the 3 parameters of which the U is the most constant for
all
tubes,
since it depends on the element spacing, which is fixed.
But the Gm and Ra values are quite variable. At low bias currents, Ra
tends to
rise.
Now Gm = U / Raor all tubes.
So if the U of EL34 is 130, and Ra is 15k at low bias current, then Gm =
130 /
15,000 = 0.00866 amps per volt,
= 8.7 ma/V.
But at 100 mA, Ra has risen to 12k, so Gm = 10.8 ma/V.

I don't use a valve tester.

I place the tube in a circuit where its wanted to do a certain thing, and
make sure it works OK with low noise and microphony, and if the tube is
OK,
the gain it will have will be what is expected. The CRO tells me more
about
these
than a tester will.
In an old radio, the IF amp is often a vari-mu tube, whose Ra and
Gm vary enormously, but where I see the set generating
a healthy agc negative bias on a small input signal, all is probably well.
Frequency converters sometimes wear out as fast as power tubes, but
replacing a
suspect dud
with a known OK tube soon finds out if its the tube or surrounding
circuitry at
fault.
Its mostly the surrounding circuit.
Tubes often get the blame for the failures of bits around them.
I do have a one channel preamp where small triodes can be tried for gain
with different loads, and we know that for all tubes,
voltage gain, A, = U x RL / ( RL + Ra ).
With two different loads, two different gains can be measured, and the
equations

above can be filled out so by very simple algebra the values of the two
unknowns,
Ra and U can be worked out.
Then GM can be worked out from the formula above.

A noise and distortion meter also tells me which tubes are best in a given
circuit.

My ears tell me which sounds best.

Its possible to have a CCS load to any given tube under test,
and with such a load, U = voltage gain for all tubes.
When testing pentodes such as 6AU6, the U may be several thousand,
since Ra is hundreds of thousands of ohms, and Gm is 3 to 5 ma/V.
So using a CCS means measuring gain could be dicky, since
the input voltage for 10vrms output of gain might only be 3.3 mV.
But using the two load method, where one load is 50k, and the other is
10k,
it is all a bit more accurate.
Its not hard to set up a tube with a DC load resistor of say 50k,
and then have a cap coupled load to reduce the load value to 10k,
say with 12.5k added to make a parallel load of 10k.
This keeps the DC conditions identical for the test,
and thus the test is valid. Keep the anode signal small, and well away
from
serious thd.

Patrick Turner.

I've finished building it...what an excellent tester...... except when
testing el34's. I only have used el34's; 20 of them, Sovtek, Chinese (mesa
boogie) but mainly Phillips and Telefunken and they all test around 7-8
Ma/v. Data says anode current should be 100Ma where the rat tester says use
50 Ma (this is the max that it can manage) are all my EL34's duds or is it
the rat tube tester unable to supply enough current. I have many many used
tubes and have found that even tubes used every day all day for 30 yrs (
6V6's in a cinema amp) can still come very close to new tube specs. Its hard
to beleive that ALL my EL34's are duds. as they sound great in my son's
guitar amp.


Hi there,

I sort of purposely limited the current available from the tester.
The rationalle I used at the time was that it could still test all
types (see below) and still not be harmful when misapplied
by folk just learning about tubes. That's also one reason the
B+ is purposely "saggy". (The other is for shorts etc).

With respect to testing the EL34s, there's a conversion
chart in every RCA tube manual published that lists how
tube parameters vary with voltage and current changes.
(It's on a different page in every manual, darn it). For instance,
to convert between 100mA and 250V plate and screen (the
spec point for 6CA7/EL34), and 45mA, the chart shows that
the transconductance ought to be about 77-78% of the 100mA
value. (8.4mS).

This suggests that all the tubes you measured between 7 and 8
mS are just fine. If they were down to 4 or 5 mS, then this
would be considered "used" (but still operable in many
circumstances).

As measured, they're perfectly fine. As has been pointed out,
successful application is a good test. However a tester is a
good way to easily weed out the stock.

Have fun with the tester.


Best Regards,
Steve

Check my web page .. A
HREF="http://members.aol.com/sbench101/"http://members.aol.com/sbench101/A
Remove the .gov to EMail me

"Steve Bench" wrote in message
news:20040526112707.15636.00001836@mb- Check my web page .. A

HREF="http://members.aol.com/sbench101/"http://members.aol.com/sbench101/A

Remove the .gov to EMail me

Steve, Please post that URL again. I'm having trouble with it. Thanks,
Paul Pinyot

Steve, Please post that URL again. I'm having trouble with it. Thanks

Note there's no WWW in AOL's member pages.

http://members.aol.com/sbench101


Best Regards,
Steve