[Heath] A possible AR-3 "upgrade"... lower sensitivity at 30 MHz.

[kgordon2006 at frontier.com]

On 26 Jan 2016 at 13:58, ChrisIwata at aol.com wrote:

> The overwhelming reason for lower sensitivity and poorer S/N at higher freqs on the those early 
> general coverage radios is they used the same 365pF (AMBC radio) tuning capacitor on all 
> the bands..
> 
> Those pentagrids all used Hartley oscillators. At the higher freqs, the L-C ratio was 
> unfavorable. On my AR-2, I measured 4Vrms on the oscillator grid for the AMBC band. This 
> dropped to 0.5Vrms on the 10-30Mhz band. So both the sensitivity and S/N ratio went 
> downhill. Plus the pentagrid ran hot, drew more juice on the higher freqs.
> 
> If they had added caps in series with the 365pF tuning cap (to effectively create a smaller 
> tuning cap) and increased the coil inductances on the higher bands, things would have been 
> better. But then you'd have a nighmare with tracking, trying to get the antenna coils to jibe.

Not mention that band-coverage would have been affected, possibly requiring more 
bands for the same coverage.

> The Franklin (not Vackar as I previously stated) oscillator overcomes this problem. It has to 
> do with how the feedback to create oscillation is derived: For Franklins, the tap on the coil is 
> unused. My Franklin produced 4Vrms on the AMBC band and still 3Vrms on the 10-30Mhz 
> band with that 365pF tuning capacitor.This is why Franklins are used often for signal 
> generators, good well above 100Mhz.
> 
> Chris, KL7DM 

Thank you very much for this extremely interesting and very valuable (at least to ME!) 
information, Chris! 

Although I have not had the necessary time to pursue such an investigation, I have 
always been very curious about the reasons for the lack of sensitivity of those HF 
receivers without RF amps, but only mixer inputs, while this has been a fairly common 
technique at VHF.

I'll investigate the Franklin oscillator asap.

I have long thought that the TRUE Pullen mixer (requiring 4 triodes, not just two, BTW) 
would make a very significant improvement in such receivers, but now, with your info, it 
is clear that that would be only PART of any solution.

In addition, the Pullen REQUIRES an RF amp ahead of it.  It does not do well with a 
direct antenna input due to the widely varying input impedances with changes in input 
frequencies.

Secondly, for those who might like to sub the 12CS6 for your 12BE6, an excellent 
company, headquarted in FL, ESRC, has those for sale for $3.00 each NIB.  I just ordered 
two.

The company is easily found via a Google search.

vy 73 for now,

Ken W7EKB