[Milsurplus] 348 loose ends

[Mike Hanz]

Unserviceable but Repairable wrote:

> <>The 348J triode RFA. It's coupled to 2nd pentode RFA via RC shunt-feed
> where 15k resistor 'way less than plate resistance so gain low. I think
> so-done so 1st stage adds little to system gain but, instead, provides 
> xtra
> selectivity 'causa the xtra tuned ckt. Besides, this seems last of series
> & there was a giant 348-factory turning all the rest of the '348 system'
> at a furious pace & this allowed use of old (6K7, etc designs) w. new 
> (6SK7,
> etc. designs). 


and

> <>Will provide 10 6J6s to 1st guy to prove otherwise I turned up a 
> ringer in that the 1st RF Amp of final -J, -N, -Q
> productions is actually degenerative. this I bet because the gain of
> the 'single-ended' tubes like 6SK7 were greater than progenitors like
> 6K7. In fact I offered 10 6J6s to the 1st who refututed that.
> No takers & thread perished


Probably because of the work involved to do the gain analysis.  Don't 
need no 6J6s, and with some caveats you may be in the ballpark anyway.  
The transconductance of the 6SK7 isn't that far from the 6K7 (2000 vs 
1650), so I'm not convinced the changed design rationale for easy 
substitution is a strong argument - the plate resistance is in the same 
general region as well.  The redesign appears to me to have three 
primary purposes: 1) to reduce costs, 2) to move to more logistically 
supportable tubes for the future, and 3) to go to a design that didn't 
require B+ through the antenna coils, which relates more to maintenance 
cost than anything.  The latter has two benefits - if there's a failure 
and the tube draws too much current, the failure doesn't involve 
replacing the coil, but rather a cheap resistor.  The second benefit is 
that the coil can be wound with smaller wire and isn't as touchy about 
moisture or insulation, i.e., less expensive - purpose # 1.

There doesn't seem to be a significant difference between the triode 
connected stage and the pentode connected stage from an external circuit 
comparison standpoint.  If I had to guess, I'd say it was just a simple 
artifact of adjusting the overall receiver gain budget to meet the 
sensitivity requirements.  The second RF amp uses a 10k resistor as the 
B+ dropping resistor while the first stage uses 15k, but those values 
clearly weren't chosen to match plate resistance or any rule of thumb 
I'm familiar with.  The use of a 6SA7 as the mixer increased the 
conversion gain over a 6J7, so the RF amps could use less gain and still 
meet the sensitivity spec.  I suspect the design was a reasoned tradeoff 
on those elements more than anything.

73,
Mike