[ARC5] Receiver Tube Substitution

[Richard Knoppow]

    The BC-779 and other Super Pro receivers used a 6K8 mixer, even 
noisier than the 6SA7. It was about the best of a bad lot at the time. 
There is enough gain from the RF stages to overcome the noise. Actually, 
the main purpose of having two RF stages is to reduce image response. 
The Super-Pro is an interesting design. The basic model is limited to 20 
Mhz where the image rejection and losses are reasonable. Hammarlund 
built a model that trades the low broadcast band for another HF band 
going from 20 to 40Mhz. They also changed the RF stages from series to 
shunt fed to get the DC out of the coils and raise the Q so the image 
rejection is still reasonably good. The peculiar cam operated switch 
was, I think, mainly a response to National's sliding turret arrangement 
but does not keep the lead length constant as the National method does. 
This obviously also inspired the rotary turret in the SP-600 where lead 
lengths are very short and constant for all bands.
    The AVC and BFO arrangement in the Super Pro are also interesting. 
The BFO is isolated from the AVC so does not desensitize the receiver. 
The complaint about low BFO injection may be due to an attempt to keep 
noise low but Hammarlund could have easily used a product detector. PDs 
were not at all common when the SP was designed, there was probably no 
demand for them.  The main shortcomings are lack of regulation of the 
plate voltage for the oscillator so that there is frequency pulling with 
variation of RF gain, and lack of temperature compensation. The latter 
was addressed in the last of the series, the SP-400.  However, if run 
continuously they are very stable. The variable bandwidth IF is unique 
AFAIK. It is a true variable mutual inductance arrangement so that the 
pass band stays centered and symmetrical with variation in width. The 
SP-600 and RCA AR-88 used tapped mutual inductance coupling coils which 
are nearly as good and simpler to use in practice.
    The Super-Pro also used the patented Hammarlund crystal filter 
starting with the 200 series. This is the best of the single crystal 
filters and was later used in the SP-600 and also by Collins and TMC.

On 12/18/2016 1:39 PM, Arden Allen wrote:
>> ......The BC-779 has two RF amp stages ahead of the 6SA7, which as you
>> state
> is a very noisy mixer.  ..........
>
> As you all know multi-grid tubes used as mixers are noisier than
> triodes. One of the tradeoffs was for gain.  Which sounds naïve because
> you need even more gain ahead to get the same signal to noise ratio as
> what a triode would provide.  I guess it was all about automatic gain
> control and RF-OSC isolation which pentagrid mixers (6SA7, 6SB7Y, 6BE6,
> etc.) did well for AA5 radios which had only two gain stages ahead of
> the detector.  With the additional grids of those converter tubes even
> more noise was generated due to the partition effect, i.e., electrons
> playing bumper cars around the grids.  The added grids were needed for
> isolation between the RF tank at the control grid and the local
> oscillator tank.  That avoided unnecessary interaction between the two
> tanks which included radiation of the oscillator signal from the
> receiver antenna.  However above 10 MHz the isolation becomes
> progressively less with frequency.  The RF tank pulls the oscillator
> when an antenna trimmer is adjusted and the AVC bus is biased negatively
> from rectification of the oscillator signal at the RF grid.  No doubt
> for years the conundrum caused Trouble in River City for designers
> trying to come up with cost efficient designs.  A 6SG7 ahead of a 6SA7
> proved to be a successful approach for moderately priced multiband
> receivers like my Halli SX-110 which I converted to product detection
> and does a fair job on the bands.
>
> Arden Allen
> KB6NAX
>
> He who is cruel to animals becomes
> hard also in his dealings with men.
> We can judge the heart of a man by
> his treatment of animals.
> —Immanuel Kant
>
>

-- 
Richard Knoppow
1oldlens1 at ix.netcom.com
WB6KBL