[Boatanchors] 8122 Suitability/Selection (was: 6146

[James Liles]

Good evening Mac:

I won't recommend a version of the 8122 for use in the NCL-2000 but can give 
you what info you need to choose wisely.  Because this will probably solicit 
questions regarding other than the NCL-2000 the 8122 needs some explaining.

The 8122 is a close spaced high power tube that every manufacturer has had a 
good deal of difficulty holding to specs.  Burle is the only maker who 
qualified the tubes to help generate a suitable design depending on the 
amplifiers requirements.  Burle color coded the plate cap with a Red, Blue, 
Green, or Black color.  They set up a test set that applied 700 volts to the 
plate, 250 volts to the screen, adjusted the control grid bias for 185Ma of 
plate current.

The Red tip tube bias was from  -8.0v to -10.9v
The Blue tip tube                        -11.0v to -13.9v
The Green tip                              -14.0v to -16.9v
The Black                                    -17.0v to 19.0v

That’s an 11v spread so why is that and why not throw the spec sheet away?

The 8122 is built with very close spacing and in such a way as to keep the 
control grid within the shadow of the screen when looking toward the plate. 
This does two things.  first it allows the control grid maximum control and 
second the capacity between the plate and control grid is at a minimum.  If 
the control grid is ever so slightly outside of the screen shadow there are 
two consequences.  The capacity between the plate and control grid increases 
causing more neutralizing voltage for stability and secondly the screen 
begins to exert more influence increasing plate current.  With this in mind, 
the type of amplifier is what determines which versions are appropriate.

I am quite familiar with the Hallicrafters SR-2000 and Test profile 8122's 
to use in that radio.  With the profile I can accurately predict a matched 
pair, neutralizing requirements, and what output you can expect.  As for the 
NCL-2000, I have never owned or used one but am quite familiar with the 
design and can give you enough info to choose.

The SR-2000 and I believe the NCX-1000 are AB1 amps and in sideband mode 
generate ALC when grid current is drawn.  They also use high impedance drive 
to the finals.   The NCL-2000 is a proprietary AB2 amp which I believe has a 
circuit that will generate ALC when the control grid reaches +15v.  It also 
uses a low impedance input (Swamping) that makes neutralizing a non-issue. 
Carl knows the National products far in a way better than I do and can 
clarify any misconceptions that you may find here.

The fact that you can plug any matched pair into the NCL-2000 and expect it 
to work properly in sideband mode is questionable.  Here's why and maybe 
someone can run a test.  The Black tip 8122 will reach saturation before the 
control grid reaches the 0v level and some Green tip 8122's as well. 
Knowing that the NCL-2000 will allow the control grid to reach well above 0v 
should be troublesome for some 8122's.  In sideband mode, I would expect to 
generate ALC before saturation.  I can't tell you for sure when that will be 
in the NCL-2000.  I would solicit a detailed description regarding the 
generation of ALC in the NCL-2000 then make your choice.

As for the SR-2000, and I believe the NCX-1000, both are AB1 and implement a 
high impedance drive.  Both are quite sensitive to the 8122 version and will 
not tolerate the Black version and to use the Green tip need rework for the 
bias and neutralizing circuit to work properly.  The Blue tip is ideal and 
the Red you will not find.

The answer to all of this is to profile the 8122 so as to know what bias is 
required for idle current, at what bias is the tube saturated, and what is 
the current level at saturation using the amps design voltages.

The SR-2000 generates ALC when the control grid reaches 0v and if the 8122 
is saturated at -3v or lower you have a linearity problem.

Glen, the 8072 that you speak of is indeed an 8122 without a heat sink.  I 
have tried to remove the heat sink from an 8122 but they are welded in 
place.  They are also beryllium copper which you do not without great care 
grind or cut -- very dangerous if ingested.  If anyone has a source of heat 
sinks for the 8072 please let us know.

NOTE:   If any of you have a comment or question and are able to articulate 
it in a friendly manner I will be happy to respond.  If not please create 
another thread.

Kindest regards Jim K9AXN

------------------------------ ORIGINAL 
MESSAGE --------------------------------

Date: Wed, 6 Jul 2011 12:52:32 -0500
From: D C *Mac* Macdonald <k2gkk at hotmail.com>
Subject: [Boatanchors] 8122 Suitability/Selection (was: 6146
discussion on )
To: <james.liles at comcast.net>, <boatanchors at mailman.qth.net>
Message-ID: <BLU129-W97E8BE23ED29273C0716DF15E0 at phx.gbl>
Content-Type: text/plain; charset="iso-8859-1"


Hi, Jim (and others).

Might I ask how Burle 8122s can be correctly
chosen for use in the National NCL-2000 amp?

I have previously been advised that original
NOS RCAs are no longer available at decent cost.

* * * * * * * * * * *
* 73 - Mac, K2GKK/5 *
* (Since 30 Nov 53) *
* Oklahoma City, OK *
* USAF, Ret (61-81) *
* * * * * * * * * * *




From: james.liles at comcast.net
To: boatanchors at mailman.qth.net
Date: Wed, 6 Jul 2011 12:36:29 -0500
Subject: [Boatanchors] 6146 vs 6146A vs 6146B or 6146W

> SNIPPAGE <

The reason that I followed with interest is the history of the 8122 seems to
draw a parallel. Some amplifiers will work with any 8122 and others will
only work with about 3/4 of the production without modification, and others
cannot be made to operate properly with nearly 1 in 6 8122's. There is a
broad range of bias and neutralization attributes involved. Burle is the
only manufacturer that classified them with a four color code to enable
intelligent choice.

Kindest regards Jim K9AXN

------------------------------