[Boatanchors] WWII Navy RAL-8 Receiver Project

[David Stinson]

Got the RAK working and finished, covered in wrap
to keep dust (and mice) out of it and back on the
storage shelf.  It was in pretty nice shape and
didn't need much work- ground fixes and alignment,
mostly.  

Now I'm working on the RAL-8, and that one is in
rougher shape.  Unlike the RAK, the low-pass audio
filter can be switched out-of-circuit, so no need
to bypass it for decent audio as we did with the
RAK.
https://photos.app.goo.gl/EawDbZrmw5AuoBXh8

Needed the usual De-Ox-It treatment, lubrication,
some tube, capacitor and resistor replacements and
the grounds refurbished.  Haven't gotten to
alignment yet.  The band switches needed a good
dose of De-Ox-It.  Slow-bring-up of B+ displayed
no unusual current leakage or spikes.  Powered-up
to check basic functions.  Receiver did sorta-play
on all bands except band 7 (5-8 MC).  The Detector
would not oscillate.

With the RAK, internal resistors divide the 180V
B+ and apply it to the Detector and this seems to
be "good enough" to run the set.  This is not the
case with the RAL; you need to supply High B+
(normally 180V but can be anywhere around there as
long as it's regulated) to terminal 6 of the
power/audio connection strip and 90V Low B+ to
terminal 5.  Fortunately, this is very easy and
cheap to do; you don't need a dual-voltage supply.
A single 1N53xx series 5-Watt Zener diode, widely
available for pocket change, can do the job.  
https://photos.app.goo.gl/pFKzvEUZFukAUpLo8
In this photo, terminal 9 (ground) is at top, 8
and 7 are Audio Out, 6 is High B+ (180V from a
regulated variable supply in this case) and 5 is
Low B+.  The Zener is a 1N5375, chosen because it
was the closest I had available.  This Zener is
rated at 82 Volts.  The diode drops 82 of the B+
volts, leaving an nice clean 100V for terminal 5
from a single B+ supply.  The diode runs cool
since the Detector draws little current.

Say you have a supply that will source 150V B+,
which will run the RAL just fine.  You need to
drop 60 Volts to supply the Detector Low B+.  You
could do it with a two-resistor voltage divider,
waste power and generate heat, or go with a
cheaper, easier alternative.

Check the spec sheet for the 1N53xx series of
zener diodes:
https://www.onsemi.com/pub/Collateral/1N5333B-D.PD
F
Scroll down to page 3 of the PDF.  Note that a
1N5371 is rated at 60 Volts and is available in
quantities of 1 from Mouser for the terribly
expensive sum of 45 cents.  Connect the cathode of
this diode and your 150V B+ supply to terminal 6.
Connect the Anode of the diode to terminal 5 and
you will have a nice, working 90V for your
detector.  Unless something shorts the 90V buss,
the diode will outlast us both.

As noted above, the receiver did play except on
band 7 and the Detector would not oscillate.
Regeneration is controlled by varying the screen
voltage on the Detector and this voltage was low.
The main problem was a very leaky C118, a 1 mFd
oil-filled cap which is the Detector screen
bypass.  This turned-out to be only oil-filled cap
in the entire set to be defective.  Decided to
"err on the side of caution" in the Detector
stage, replacing resistors more than 20% out of
tolerance and the low-value silver micas.  The
Mica-Mold .01 bypasses all checked like new, so I
left them in place.  If one fails, it will tell
me.  
The 2.2 Meg Grid resistor in the Detector was
fine, but the 1 Meg grid resistor in the preceding
stage (R107) was 1.8 Megs and was replaced.
Powered-up and the Detector now oscillates
normally.  However, band 7 was still dead.

With the top cover off the tuning capacitor
compartment, I connected an Ohmmeter to the
stators of the three tuning capacitor sections and
ground, each by turn.  This reads across the
selected band's coil section for each band.
Turned the band switch and all coils showed proper
low-resistance paths except for the Detector
stage, on band 7, which showed open.  Turned the
radio over and removed the bottom cover over the
RF stages.  Then carefully removed the panel with
the set screws in it that keeps the band switches
in alignment.  Be very, very careful here.  If you
twist or pull any of those switch segments, they
have to be mechanically re-aligned and you will be
very unhappy.  Don't "gorilla" anything.

Used and ohmmeter to check continuity through each
band switch set of contacts (2 per switch on band
7) but didn't find a problem.  However, on switch
S111, I noticed there is a mica, C142, 510 pFd,
which is connected only when on band 7.  I
carefully unsoldered one end and connected the
ZM-11 cap tester.  Sure enough, the little bugger
was leaking badly.  The manual says the function
of this cap is to "Detuning cap on primary winding
of the LF Detector Coil when using Band 7."  I
don't know why but it must be important.  Maybe
the stage oscillates and blocks on band 7 without
it?  Replaced it with a 560 pFd silver mica, put
the switch alignment plate back on and band 7 came
alive.  

Did a few other misc. fixes.  Replaced the 100K
grid resistor in the Audio Output stage, corrected
cold solder joints at a couple of tube sockets,
disconnected the .01 uFd RF bypasses across the
Audio Output transformer secondaries to ground to
improve high-frequency audio response.  Will be
playing the set a few hours tomorrow as a "burn
in" to make sure there aren't any surprise smoke
signals.  Rig sounds really nice when the Regen
control is set properly.  

Which brings me to a request:
https://photos.app.goo.gl/C5wBpK7j6EbvW6YW8
Need a couple of knobs.  One large, one small.
Enough RAK/RAL sets were scrapped that surely
someone out there has a couple of knobs.  I
*think* I have the Output meter.  Still digging.
I do have have freq chart holder, just removed it
for painting.

Also:  The nomenclature tags on the case are the
fiber-board type.  They've got that "faded" look
that so common with these:
https://photos.app.goo.gl/h2G31zrpTCZGtWC77
Don't know if the ink fades of if it's MFP that
gets cloudy.  Has anyone succeeded in getting
these tags to look decent? 

Hope this encourages someone to fire-up their set.

GL OM ES 73 DE Dave AB5S