I'd say your reception is normal. Back in the day, 100 foot plus of antenna and a good earth ground was standard practice. Also consider that the fewer stations back then were much higher power. A highly sensitive AM receiver today often picks up multiple stations on the same frequency, more so in the evening hours. The AM broadcast band has become somewhat of a low power waste land.

Kevin is right, a long antenna and good ground are necessary on a set like this. Also, are you sure your tuning condensers are all tracking properly? The tracking can be tweaked by loosening the pulley setscrews and adjusting the tuning shafts. Is your local/distance switch turned to distance?

A college buddy of mine has been doing DX listening at night with his Atwater Kent 40 that he got from his great aunt(IIRC) ever since I helped him get it working...Granted he does not expect room filling volume, and won't erect a long-wire antenna for it...

I'm now working on my second one of these, a Model 44F. F means its for 25 Hz.

I got the stuff in the power supply out of the tar yesterday and measured
the capacitors. The ones which are not filter capacitors are roughly
the same values as usually listed for the plain 44. The three filter
capacitors are much larger, physically and in capacitance.
They measure differently on my capacitance meter and by
measuring the response of an RC filter using them. The latter is
usually accurate. All three are in the 3.5 to 4 uF range, and
all look to be identical size. I plan on using values of 1.5, 10 and 10
uF for the recap values. I don't want to go too high on the first
one as this raises the B+ voltage ... and of course the set will be used
normally on 60 Hz. Those values will still work fine
at 25Hz of course, the B+ will just be a bit low, assuming the
same turns ratio on the monster transformer.

I also rebuilt the ballast, which was broken. I used 0.005 inch
Alumel wire and needed roughly the same number of turns
to get the same value, though I made it in fact 33 Ohms
due to our higher AC line voltage. Alumel is nice since
you can solder to it. The wire was reglued to the ceramic
core using a product called Red Devil Furnace Mortar.

I recapped it except for the micas, but left all resistors as they were good.

It worked fine the very first try.

I need help with the paper thing in the bottom. I'm enclosing a scan
of what I have, plus some reconstructions in Photoshop. They are BodoniMT
with a one pixel "stroke" effect. Its amazing how well a 2013 computer
reproduces a 1928 lead font.

But parts are missing ... I need the wording in those parts as well
as a picture or description of the lines and arrows that are missing. I will
reconstruct it and post a copy so people can make more of them. A
cell phone photo would do.

This might help.
 

This is the replication from my Model 37.

https://lh6.googleusercontent.com/-u...2520Thingy.jpg

A little better look at this link:
https://picasaweb.google.com/lh/phot...eat=directlink

Thank you! My Model 44 is slightly different, and if anyone has a photo of
that one for that I'd still like it, but if I don't get it, this and a bad image
in an original A-K brochure will do.

I found enough data to make one. I will try attaching a copy, as
a .zip file since it is a BIG image (over 11,000 pixels wide).

Also ... I found that the noise disaster using a wire antenna,
not matter how long, such that in the evening even some
local stations (one 25 watt and one 1 watt) were unreadable,
was fixed by manufacturing a 15x15 inch loop using 18 turns,
tapped at one turn to connect to the radio. This works as
well as ordinary AA5 radios with a built-in non-ferrite loop,
actually very well. In the daytime I get all the usual Chicago
stations 130 miles away.