[ARC5] Receiver Tube Substitution

[Richard Knoppow]

    Your experiments are interesting. Hallicrafters, of course, modified 
the RF stage for both bias and for correct plate and screen voltage, 
which are slightly different. Its likely the input and output impedances 
are different so the transformers would need some change. The older 
tubes were pushed up beyond 30Mhz in many receivers but I think it was 
wishful thinking plus some advertising hype to claim something like an 
S-20R would go up to 40Mhz. Another problem is mixer noise. The old 
mixers like the 6SA7 and 6A8 are very noisy tubes and need a lot of gain 
ahead of them to prevent them from dominating the noise figure of the 
receiver.  I tried substituting a 6SB7Y for the 6SA7 in an RCA AR-88. It 
made no difference whatever as a straight across change but, of course, 
it might if bias and other parameters were changed. In the case of the 
AR-88 the two stages of RF with 6SG7s and low loss coils on the higher 
bands results in quite low noise.
    No simple answers for this one.

On 12/17/2016 5:43 PM, Kenneth G. Gordon wrote:
> On 17 Dec 2016 at 14:08, Richard Knoppow wrote:
>
>>     While the 6SG7 will plug in place of a 6SK7 and work OK I am not so
>> sure its noise will be any lower.  The bias for maximum gain is about 1
>> volt while the 6SK7 needs 3 volts so in a receiver designed for the 6SK7
>> the bias will reduce the gain and Gm of the 6SG7. I have not tested this
>> but think the resulting noise would be the same.
>
> All the above is pretty much the case, especially the biasing issue.
>
> Furthemore, internal capacitances are enough different that simply plugging a 6/12SG7 in
> place of a 6/12SK7 in the RF amp results in the "Align Input" capacitor of an ARC-5 not
> working correctly.
>
> Replacing the 12SK7 with a 12SG7 in several refurbished ARC-5 receivers here resulted in
> no noticeable improvement and the "Align Input" issue I mention above.
>
> However, I also have a BC-454 here which was converted to cover the 10 meter band
> sometime around 1950 or so. I finally got that one to work, and the 12SG7 does make a
> noticeable improvement in the noise level on 10 meters, even without adjusting the biasing
> so that it is correct.
>
> Even so, the 12SK7 RF amp still shows a sensitivity on 10 meters of 1 microvolt or better
> MDS on CW.
>
> I am reasonably certain that any advantage to using the 12SG7 over the 12SK7 would not
> become apparent at any frequency below about 20 MHz.
>
> Ken W7EKB
>
Richard Knoppow
1oldlens1 at ix.netcom.com
WB6KBL