[ARC5] Question on IF transformers...

[Kenneth G. Gordon]

On 19 Oct 2014 at 11:05, AKLDGUY . wrote:

> > Why are connections to the 1st IF transformers in the receivers up through the
> > 1.5 - 3.0 MHz receivers NOT tapped down on the IF coil, while the 2nd and 3rd
> > IF transformers ARE tapped?
> 
> At low frequency IF's, the 12SK7 will have more gain than it will at 2830 KHz
> in the 6-9.1 MHz receiver. Here, the first IF transformer is not tapped in
> order (my opinion) to get best signal to noise ratio out of this first stage.

OK. I wasn't aware that NOT tapping results in more gain, but after thinking 
about it a bit, I see where it could. Connecting the tube elements to the 
highest-impedance points of the transformers would (or should) increase the 
voltage there.

> The 2nd and 3rd stages are tapped to reduce the gain at this low IF and to
> improve the bandwidth. 

OK. That makes sense. :-)

> > In the 3.0 to 6.0 Mhz receivers, the 1st and 2nd are NOT tapped, while the 3rd
> > MAY be tapped,
> 
> Again, at 1415 KHz the 12SK7 gain is lower than it was at the lower IF's, and
> we want to get the best signal to noise ratio out of the first IF stage and
> max gain from the others. The "3rd MAY be tapped" may be an indication that
> this type of IF transformer is intended for use when the 12SF7 is fitted
> (maybe it has more gain than the 12SK7). 

Ah, yes. I hadn't thought of that either. I only checked the ARC-5 manual, not 
the SCR-274N manual. The ARC-5 manual shows a dotted line at that point 
on the schematic, whereas all others show a solid line.

> > ...and in the 6.0 to 9.1 MHz receivers, NONE are tapped.
> 
> Again, we want more gain at this even higher IF frequency. Presumably there was
> no benefit in tapping for improved selectivity (gain more important than a
> modest improvement in selectivity).

Well as someone, possibly you, pointed out, the 6 - 9.1 Mc receivers have no 
selectivity to speak of at all, so in those it would be moot to tap the coils.
 
> It's all about gain distribution and thus stability IMO, as well as
> bandwidth. Too much gain can result in feedback (breaking into oscillation)
> when tubes are changed, tolerances drift, etc. 

Yes.
 
> The writer had some experience in the telecommunications industry with
> distributed gain and bandwidth setting in broadband systems where, due to
> multiple repeaters in an 80 mile link, the gain of each repeater had to be
> flat within 1 dB (+/- 0.5) from 60 KHz to 4.0 MHz and had to meet the target
> gain at midband within 0.7 db. If you think we've got it tough, spare a
> thought for the telecomms engineers. 

Yes. When I was a broadcast engineer (TV and FM and AM) way back in the 
dark ages, I had to do that sort of thing too. Rather tedious, it was.

Thank you, Neil. Makes sense to me.

Ken W7EKB