[AMRadio] WRL Globe Champion 350: Back On the Air with 8643s

[Kim Elmore]

I have to crow a little bit and this is probably the pace to do it: 
with some help and encouragement from Jim, W5JO, I've successfully 
completed converting my dad's (W5JHJ SK) Globe Champion 350 to use 
8643s in the final, retiring the 8063/AX9909's even though they still 
made full output. (If anybody wants a good pair, these are for sale.)

I found nine 8643s through an estate sale of W0ZUS that Ron, W8ACR, 
helped manage. My dad used to have a set of spares for the AX9909s, 
but apparently used them at some point as they were not in evidence 
when I went through his things. So, while the Champ made full output, 
if anything ever happened to to those AX9909s I would have to use 
some sort of substitution. At 76 W plate dissipation per tube (2 x 38 
W) and 8 W screen dissipation, the 8643s are a much sturdier tube yet 
electrically almost identical to the AX9909s. Thus, they are perfect 
for the job!

Jim had just finished a complete rebuild of a Champion 300 and 
performed the conversion to 8643s as part of the job. I've seen it 
and it's stunningly gorgeous! That spurred me out of my complacency 
and I got busy.

Earlier, I had replaced all of the rectifiers with SS units, replaced 
all the electrolytics and removed the last of the selenium 
rectifiers, installing a Zener a bias supply for the 809s and 
rediscovering the B+ supply for the VFO and accompanying bias supply 
for the PA that my Dad had installed.

Jim did the same but the rectifier in the 5U4GB slot shorted, taking 
the LV transformer with it. Both he and I decided to fuse the LV 
transformer separately so that, if such a thing ever happens again 
the fuse save the transformer before the transformer sacrifices 
itself to save the fuse. Jim learned through trial and error that the 
proper fuse for the LV transformer primary is 6 A. The original 10 A 
fuse remains in place

The 8643 substitution followed exactly K9RJs write up in Electric 
Radio, since the 8643 is nothing more or less than a very stout 
version of the 5894/AX-9903. Neither Jim nor I used the Fahnestock 
clips, but instead found some nifty plate caps I found at Surplus 
Sales of Nebraska. Once drilled out about 0.130", these were a 
perfect fit and should get rid of plenty of heat. keeping the seal 
temperatures well below the maximum even under the most long-winded 
transmissions.

Again following Jim's lead, I removed a pi-type low pass filter in 
the audio chain that feeds the grid of the 6AQ5 that drives the 809 
modulators, and replaced it with a 0.01 uF disc ceramic. This opens 
up audio passband a bit -- we both left the couplates in place.

Jim used some band switch wafers from a DX-100B to convert the output 
tanak for a pi to a pi-L on 160 m. I chose a different approach, 
using instead a 2:1 Sevick transmission line transformer on the 
antenna output, instead. Thus the CHamp "sees" 100 ohms when feeding 
a 50 ohm load. This works very well on all bands and 160 m tunes 
smoothly. I'd been using it in an outboard enclosure for some time, 
having found that the Champ can't quite match the 200 ohms presented 
by a 4:1 Sevick transformer on 10 m.

I neutralized the transmitter with the new tubes and found, to my 
surprise, that the required adjustment was significant. The spacing 
on the neutralizing capacitor is now around 3/32" or so. I also 
lubricated the fan that helps cool the tubes while I was in there.

Finally, at Jim's suggestion, I added an RF sniffer probe in the 
under-chassis output enclosure, so that I can measure the output 
frequency more easily with the B&K 1827 counter that my dad had used 
for at least a couple of decades. I use a panel-mount BNC and 
sacrificed the input for use as a 400 W input linear amplifier, since 
the rig will never be used as a linear amplifier.

Output remains unchanged, those the 8643s seem a bit easier to drive 
-- in fact, they take very little. I stumbled along the way, leaving 
the grid of one tube floating (I failed to solder the lead that 
paralleled both grids). That lead to much consternation, as it seemed 
to tune normally and load fine, making plenty of output, but would 
peg the plate-current meter if drive was removed, such as when the 
key is open in CW.

Late this afternoon, I had a QSO with Jim on 75 m, he using his 300 
and me using my 350; I was using my HQ-170 receiver and I think he 
was using an HRO receiver. It was, frankly, glorious!  Substituting 
something else for the AX9909s was something my dad and I had 
pondered for 10-15 years, but neither of us ever got around to it. So 
now, it's done and I'm tickled silly. I've had a lot of fun getting 
to this point. It's now done and just in time for winter! I'll be on 
the air with it over the fall and winter as time permits (I'm still a 
a working stiff and have daddy duties). I look forward to meeting a 
lot of you on the air!

73,

Kim N5OP