[AMRadio] Cabinet for the T-368 exciter?

[D. Chester]

Probably the best bet is to make one.  I modified one and use it as my 
station VFO, capable of driving any of my rigs.

I still found the drift objectionable when the PTO is turned off during 
stand-by, particularly when working 40m CW, so I modified mine to let the 
oscillator run all the time even when the rest of the unit is turned off. 
The problem with doing that is that the oscillator is audible in the 
receiver, which can be a problem with weak signals when  the band is quiet. 
I built a metal shielded enclosure to encase the  whole thing, PTO, 
buffer/multiplier unit and all.  I used about 1/16" sheet aluminium and some 
1/4" square rods of aluminium stock salvaged from something long forgotten. 
I cut side and top panels, using the original bottom cover, rear panel and 
front sub-panel behind the mechanical dial assembly.  I carefully drilled 
and tapped holes for 4-40 screws, using the square stock as a skeleton to 
hold the sheet metal panels together at the corners.  The front and rear 
sub-panels in the unit are thick enough that I was able to drill and tap 
directly into the edges, without the square stock.  I used standard circuits 
for TVI filtering and bypassing for all power supply, filament and control 
leads going in and out of the unit, except that the inductances and 
capacitances were optimised for the 1.5-20 mHz range instead of for VHF.  I 
was able to reduce the rf feed-through to the  receiver to negligible with 
the receiving antenna connected, although it is faintly audible when the 
antenna is removed from the receiver.

I  replaced the type 6000 tube with a 6AG7.  I had to re-wire the octal tube 
socket, but the capacitances of the two tubes are nearly the same.  I put 
only about 150 volts on the 6AG7, and that gives me about 200-250 milliwatts 
of power out, which is comparable to the output from a ham type VFO like the 
Johnson 122 or Heathkit VF-1.

The next problem was coupling the output to the transmitters.  In the 
original T-368 setup, a short length of coax is used to directly couple the 
plate of  the output  to the grid of the final amplifier, and its 
capacitance was not enough to cause a problem with resonance at the output 
coil. Some of my rigs are as far as 10' away from the operating table, and 
that much coax loads down the output circuit with too much capacitance, and 
the alignment parameters changed with different lengths of coax to the 
different  rigs.  So I decided to make mine  link  coupled.  I salvaged a 
spare bandswitch wafer from a T-195 multiplier/buffer unit I had on hand, 
which is very similar to the one in the T-368.  I disassemble the 
bandswitch, cut two of the hollow tubes used as spacers between wafers to 
accomodate the additional wafer, and re-assembled the bandswitch.  I then 
took out the output coils and removed the shields, then wound a coupling 
coil over the cold end of each coil and brought the new  lead out the bottom 
of the  coil.  I don't recall if there was already an extra banana plug at 
the coil bases, or if I added one or simply brought the wire out through a 
hole (I made these mods about 20 years ago).  The new wafer section was used 
to select the coupling coil.  As I recall, I used about 2 turns for the 
highest frequency coil, 4 for the next, 8 for the 3-6 mHz coil and 16 for 
the 1.6-3 mHz coil.  I removed the original BNC connector and relay at the 
top of the buffer assembly, and relocated the BNC to the rear of the unit as 
the output jack.  With the modified output, I can get about 1/4 watt of rf 
into a 50-ohm dummy load.

Before the mod, the unit would track perfectly over each one of the tuning 
ranges.  I was amazed that I could put an RF voltmeter at the output and it 
would show a constant reading no matter what band or frequency I tuned it 
to.  Adding the coupling links somewhat screwed up the linearity, and I 
could never get the ranges to track as well as they did before the 
modification.  I possibly could have experimented with  the exact number of 
turns of coupling coil on each output coil, and been able to get the unit to 
track perfectly over each frequency range, but I didn't go to the trouble, 
since I use each frequency output range of the unit for only one ham band: 
1.5-3 for 160, 3-6 for 75/80, 6-12 for 40m, and 12-20 for 20m.    I found 
that I could align the output coils well enough that the unit would still 
track perfectly across the entire ham band in each frequency range, but the 
output would fall off at the extreme ends.  Since I don't do pirate 
broadcasting on shortwave and have no use for frequencies outside  the ham 
bands, the tracking error made zero difference to me, so I didn't bother. I 
haven't tried using it on 30m or 17m, but hopefully I would still be able to 
get enough output on those bands to drive whatever transmitter I might use.

The final modification was to add a Jackson Brothers planetary dial drive to 
the tuning knob, because the stock tuning rate was too fast on the higher 
frequencies. The reduction drive magnifies the small amount of backlash in 
the tuning mechanism, but it still allows for more precise tuning to 
zero-beat a signal, particularly on 40m, than with the tuning knob running 
straight through.

Using a drill press, I drilled a large number of holes in the top panel of 
the sheild I added, in a pattern that places ventilation holes above each 
tube, to reduce temperature build-up inside the unit and possible warm-up 
drift. I used a small drill, something around a size #48, to make the 
ventilation holes.

Don k4kyv