(c) 1999,2008 Peter McCollum
The RT/A-3, a part of the BN-2
The RT/A-3 is based on a 'standard' RT-3 transmitter, but with
the following modifications:

RT/A-3 transmitter. Note the missing bandswitch
control, the modified crystal socket, and the yellow stripe on the case. Author's
collection.

Close-up view of the RT/A-3's modified crystal
socket, with a crystal installed.

Another style of RT/A-3 modification (probably
later). The markings are printed directly on the panel, and the crystal holder
is different.
The BN-2 is apparently a designation for a beacon transmitter
which includes the RT/A-3, and was used to signal for air drops, or to mark an
airstrip. To support this theory, consider the following:
Many types of aircraft in
the 1950's were equipped with navigation equipment that tuned in the area of
the broadcast band; such as the ARN-59 which tunes 190-1750 KC. So, the RT/A-3
could have been used by clandestine teams in remote areas as a way of providing
a navigation beacon for supply air-drops, or to mark a temporary airstrip. The
RT/A-3 would have been very familiar to the agents who were already trained
with the RS-1. Because of the crystal-socket modification, it could have been
delivered to the field with a crystal already installed (a wide variety of
frequencies may not have been needed for a beacon that was used only
occasionally, and the signal could be easily 'hidden' in the AM BC band). With
the key 'locked down', it would be easy for a single person to set up and
operate the transmitter with a GN-58 generator; and transmitting an unmodulated
carrier allows it to be more easily overlooked by the enemy. The January 1998
issue of "Air & Space" magazine has an article about the CIA's
air operation in Tibet in the late '50's. The author mentions a case where a
C-130 was returning, low on fuel, and the pilot needed to find a CIA-run emergency
airstrip in Thailand that had "only a non-directional radio beacon".
The RT/A-3 could fit a situation like this.
The RT/D-3 is a variation of the RT-3 transmitter which includes
the addition of a burst-coder connector, similar to the T-784/GRC-109. On the
ID label, the "RT-3" has been covered by a small foil sticker that
reads "RT/D-3". On the lid, the letters "RT-D-3" have been
stenciled in yellow, about 1/2" high. There is an "MWO 39"
marking on the panel, just like many RT-3 units. This modification adds an RF
choke between the final tank circuit and ground, thus providing a DC path to
ground for the antenna. Two examples of this model have been seen so far - both
are late-production RT-3 units.

The panel of the RT/D-3 transmitter. Note
burst-coder connector in the upper-right, and the "MWO-39" marking in
the lower-left. Author's collection.
This is another variation of the RT-3 transmitter. The
difference is in the frequency coverage - the "E" model covers 3-30
MC, instead of just 3-22 MC. The bands are 3-5.5, 6-9, 10-16, and 17-30 MC. The
original tuning chart has been covered with a new chart (black on white,
instead of white on black) to indicate settings up to 30 MC. On the panel, the
original "RT-3" marking has been painted over in black, and
"RT/E-3" has been painted over it. The "E" model does NOT
have the burst-coder connector like the "D" model. Three examples
have been seen.
The reason for adding the 22-30 MC coverage was to allow more
propagation options when operating at certain times of day, or over certain
distances, where the lower frequencies were not always acceptable.
The conversion of a standard RT-3 to an RT/E-3 is perhaps
described by “MWO 43”, as this marking is seen on one example. The circuit
modifications are as follows:
·
Remove 3 turns from L3 (the
final tank coil for bands 3 and 4), to extend the frequency range of band 4.
·
Reverse the end connections
on L1 (the oscillator tank coil for bands 3 and 4). This moves tap closer to
the end, to extend the frequency range of band 4.
·
Change R6 and R7 from 15K to
2.7K each (the oscillator plate load resistor). This will increase the plate
current in the oscillator.
·
Replace R8 with a wire (the
final grid-leak resistor is now only an RF choke).
·
Change C8 from 18pF to 15pF.
Bench testing at about 18 Mhz shows that the RT/E-3 has
noticeably higher power output compared to a standard RT-3.
RT/E-3 serial #5724 was found to have a 6AG7 oscillator instead
of a 6AC7. Are the two types intended to be interchangeable in this
transmitter, or is one a mistake? Since the oscillator runs at a higher plate
current due to the modifications, it is possible that a 6AG7 was the intended
tube.

The ID markings on an
RT/E-3. Author's collection.