[Hallicrafters] reverse engineering of the SX-28 LF amplifier

[Roy Morgan]

At 11:36 PM 3/14/02 +0100, Emanuele Girlando (IW1DHI) wrote:
>Dear Gurus,
>I am trying to reverse engineer the LF amplifier of the SX-28.


Congratulations.. I applaud your effors!

...This working condition is far away from the one
>recommended by tube manufacturer. Why this choice? Why the choice to stay 
>so low in all the values?

Because those conditions met the desired performance of the amplifier.  The 
needed output voltage and gain of the audio amplifier are set by the drive 
requirements of the 6V6's and the available output voltage of the detector 
circuits, respectively.  The manufacturer's "recommended"  or "typical" 
conditions are for reference only..those conditions were not needed to 
accomplish the needs of in SX-28.

>As per original diagram the plate load for the first section is R37 or the
>LC circuit CH2-C43 depending on the SW10 position.

This is the tone control, or part of it.

>This configuration leads
>to two completely different working conditions and introduces a level of
>asymmetry with the second section plate loading (plain R36=R37). Is it
>normal? Is it a mistake in the diagram?

Not really "normal" but not a mistake.  The two positions of SW-10 just set 
up to different conditions for audio signals.. don't worry if the DC or 
steady state conditions change somewhat too.  Consider the circuit from two 
different points of view: DC standing (steady state) conditions, and then 
audio small signal conditions.. It helps a lot to separate the two 
completely, especially in class A operation.  An important point in the 
design was, I assume, that high plate loads in the 6SC7 stage with 
moderately high plate supply voltage, lead to high ac voltage gain.

>  If correct, how it works?

Do not confuse published typical operating conditions with a set of other, 
different, choices made by competent engineers with different design goals 
in mind.

The second section of the 6SC7 is a simple inverter.. the grid voltage is a 
portion of the output of the first section always.. whatever voltage is 
present at the plate of the first section is inverted in the second section 
(sorry I can't read the pin numbers well).  This develops a push pull 
signal, one side 180 degrees out of phase with the other one, to drive the 
6V6 grids.  The resistors R-39 and R-41 deliver to the grid of the second 
section a voltage, which when amplified by the second section, produce at 
the pate of the second section a voltage equal but of opposite phase 
compare to the  signal at the plate of the first section. Thus the 6V6's 
are driven with a balanced push pull signal.  This is the purpose of the 
driver stage.

The assumptions in this design are:
  - The load of the driver stage is linear (That is, the 6V6's do not draw 
grid current - they may operate in Class B1, but not B2)
  - The R-31 R-41 division ratio is fixed, and the resistors do not drift (!!!)
  - The gain of the second section is known and does not change very much 
with tube age and plate voltage variations.
  - High fidelity in not required, thus loss of high frequency response is 
fine in the trade off with additional gain from high plate load resistors 
in the driver stage.

>I don't have an original SX-28 to check with.. my own has been tampered with
>everywhere...

Rip out all modifications and return the set to the original 
configuration.  Start all over again.  Use new resistors and caps. You will 
like the results.

PS:
<Soap Box Mode ON>
Audio Power Required:
Note the rated audio output of the radio and the associated distortion 
figure.  I challenge anyone with normal hearing to listen more than a few 
seconds to a radio delivering more than half a watt to an efficient 
speaker.  It will literally drive you out of the room.  You do not need 8 
watts of audio power!

Paper Capacitors from 1938:
Measure the 6V6 cathode current with the original paper capacitors at the 
grid inputs (C-45 and C-46).  Then replace them with modern caps.  Measure 
the cathode current again.  Ponder how long your 6V6's would have lasted at 
twice or three times the normal rated cathode current.  Consider the price 
of replacement tubes and output transformers.   Become a believer that 1938 
paper capacitors are all leaky and will put your tubes in danger.  Think 
carefully about adopting the policy of replacing such capacitors before you 
even turn on an old radio in the future.

Audio Output Tube Life:
I suggest you increase the 6V6 cathode resistor, R-42, to set the standing 
cathode current to about 80 percent of what the original design gave.  This 
will reduce the maximum available power output somewhat but will increase 
the life of your 6V6's by a whole lot. Notice the standing plate current 
and plate dissipation with the B+ increased by modern line 
voltages.  (What!! you don't nave a line bucking transformer in operation? 
Shame!  See below.)  Measure the actual audio output power into your 
favorite speaker and set it for one watt average and see how loud it is.. 
Then consider how loud it would be at the set's rated output level.  Think 
about the fact that an average set of earphones run at one watt input will 
blow up in less than five seconds and take your eardrums out first.  Then 
think carefully if you would rather have your 6V6's get tired fast and have 
that power available, or "make do" with slightly reduced maximum power and 
have your tubes last a long, long time.  It's up to you.. No, you cannot 
buy any of my spare 6V6's.

Overall Voltage Stresses:
Next, measure your normal house line voltage.  Measure your tube filament 
voltages at that line voltage (at the sockets).  Figure out what percent 
overvoltage they are running at (I predict it will be from 5% to 8%.)  Then 
consider the expected tube life at that overvoltage. Then go to the 
following link and learn about "Compensating for High Line Voltages":
<http://www.r-390a.net/faq-HiVolt.htm>

<Soap Box Mode OFF>

Happy SX-28 restoration, and may it play for decades more to come.


Roy
- Roy Morgan, K1LKY since 1959 - Keep 'em Glowing!
7130 Panorama Drive, Derwood MD 20855
Home: 301-330-8828 Work: Voice: 301-975-3254,  Fax: 301-948-6213
roy.morgan at nist.gov --