[Boatanchors] A 1936 HRO Means a Weekend Project

[David Stinson]

Well my HRO Sr arrived.  Clean as a pin and very pretty.
It's serial number L114, which means it was built in early 1936.   
http://home.netcom.com/~arc5/HRO/hro1.jpg

Doesn't have a tube one in it, LOL.
Guess the estate folks thought those old tubes would
buy them a retirement; good luck with that ;-).
That's OK because I've got most of them 
and I'll get to plugging them in shortly.

It came mounted in a floor rack with lots of room.
Well.... there's nothing for it but to build a 
little mid-1930s CW rig to go in the rack with it.
So since the AN/ARC-2 is presently on hold,
I started digging-through my collection of 
1920s and 1930s  parts.  

I settled on a simple 1935 single-tube pentode crystal oscillator.
I wanted it to *look* like a mid-1930s "home brew" rig-
wooden baseboard with ground-busses
screwed to it and the whole "scrounged out of the junk yard" 
effect.  I had just the tube, too- A Raytheon RK20 pentode.  
It's tall and bright and all-around awesome-looking.  
Trouble is...  I can't find the darned thing! 
So for now, I decided to go with a late-30s (1937) 
GE version of the same tube- save a different fil voltage- 
the 814.  It looks a lot like the RK20 and 
will be easy to swap-out when I find the Raytheon.

While the grid and screen circuits remained the same,
I built three different versions of the plate circuit:
Series feed-link out, shunt feed-link out
and a Pi-network.  The link outputs worked fine.
I'm still de-bugging the Pi network.
Here are diagrams of the two successful configurations:

http://home.netcom.com/~arc5/HRO/1935tx.jpg
http://home.netcom.com/~arc5/HRO/1935tx2.jpg

I forgot to include the world's gawkie-est 
parasitic plate choke in the diagrams, 
but you'll see the awkward thing in the photos.
I probably should put one in the screen lead,
but haven't so far.
Also- I had to couple "C-Special" a little closer
than the description might indicate, due to 
my early Bliley crystal having activity issues.
I had to take it apart and clean the nearly 
inch-across blank to get it to work at all,
but seems OK now.

I "fudged" on the tank coil and used ARC-5 
transmitter parts.  I'll wind a "real" 1930s tank 
later, after I've worked all the other bugs out.
Also need to replace that disk-ceramic plate
blocking cap, which is like bringing a Formula One
to an Egyptian chariot race.

For the successful link-coupled tanks, I used the
PA coil from a 2.1-3.0 MC transmitter.  Did this 
because its link has more turns than the one
for 3-4 MC.  I removed the powdered-iron 
core and shorted three turns at the top.
Using a 400 pFd cap, this allowes me to 
resonate the tank on both 80 and 160 meters
and the extra link turns mean I can actually
feed a 50-ohm load.  

I'm going to mount meters on another panel
that will go across the top but leave the 
middle open so I can see the tube.
Here's some views.  I left the "loading" cap
mounted as I'm not giving-up on Pi-output
just yet.

http://home.netcom.com/~arc5/HRO/814rig4.jpg
http://home.netcom.com/~arc5/HRO/814rig3.jpg
http://home.netcom.com/~arc5/HRO/814rig2.jpg
http://home.netcom.com/~arc5/HRO/814rig1.jpg

You'll notice an insulation-coated wire coming from 
a white stand-off and sort-of "twirling" around the 
plate lead.  This is "C-Special."  Plate-Grid isolation
is so "good" in these old tubes (in the 814, the 
plate/grid interelectrode capacitance is like  1/4th
of a picoFarad!), you have to provide
external feedback to get it to oscillate.  
Adjustment is a little tricky-  High-activity crystals
don't need much, low-activity crystals need a lot.

For God's sweet sake, remember there's B++
on that plate lead!

To tune up, you hook the rig to a dummy load,
uncouple the loop and slowly turn the tank cap 
until the tube oscillates.  Peak the output then-
and this is important- tune the cap for slightly 
less capacitance- 5-10% from peak output 
is usually about right.  This is required to get
the feedback phase correct and give a clean note.  
Tuned to peak output or on the higher-C side 
will cause chirp and/or failure to reliably oscillate.  
The 400 pFd tuning cap makes this a little
"touchy;"  150 or 200 pFd would have been better
if I were just going to stay on 80 meters
and 150 is about as big as you can use if
you build this rig for 40.
On the lower-C side of the peak you can get
a nice, clean note if your crystal is active and
you have enough coupling in "C-Special."
Now start tightening the link coupling and 
re-tweaking.  If you're using a 50-ohm load
and the rig stops oscillating, you need to tighten
"C-Special."  Reactive antenna loads can cause
the rig to stop oscillating even if C-Special is "right."

With 650 volts on the plate and 250 on the screen,
the rig will output 15-20 watts.  I'll be going 
to a kiloVolt once I'm happy with all the "tweaks."
That should get it up to 40+ watts.
I may decide to just keep it at 20 watts, 
since I'm making contacts without trouble.
My first contact on the rig was a station in North Carolina,
which isn't bad.  If you hear me on 3522 some evening,
give me a call.

Oh yes one neat thing:  My first 814, for no reason I know,
developed an open filament.  I have another 814 but 
it's a little "gassy;"  the getter is just about all used-up.
After letting the tube sit and draw about 25 mils at 250 volts
for a few hours, the "gas" seemed to go away and it works fine.  
I got to looking at the really nice, full getter on my now dead-fil
814 laying in the bottom of the trash can 
and decided I had nothing to lose, so hooked 12 volts to the 
filament pins and started banging on the tube.  Lady Luck smiled,
because after about the 8th try, the fil ends touched 
and welded back together.
I left it lit for a few hours and the weld seems to have "taken."
A few off-n-on cycles and the fil is still lighting and the tube is
working normally.  I'll see in the morning if it survived the night.
It will get gentle handling for sure.

73 DE Dave AB5S

Ronnie:  Can you post this over to Glowbugs?  Thanks.