[R-390] R-390A progress

[windy10605@juno.com]

Thought I'd send a note of where I'm at with the receiver.
It's a long note so just delete if you're not interested.

I'm going to put the repair scenario on my web site with 
pictures later.

73  Kees K5BCQ 
  
-------------------------------------------------------------------------
---------
Bringing another R-390A back to life 

R-390A purchased at a local swapmeet. Always wanted one because of 
the performance reputation and the amazing mechanicals. Serial #26xx
manufactured by EAC (a division of Hammarlund) per 1967 contract.
All the component serial numbers are roughly the same as the unit serial
number when you consider yield off a typical electronic sub assembly
manufacturing process. 

  Cabinet #40xx
  RF deck #44xx
  Gear train #38xx
  IF deck #43xx
  Osc deck #41xx
  Audio deck #32xx

The individual selling it said: "he was absolutely NOT going to take it
back home", "it had very low audio", "it didn't work right �. might be a
bad tube" (�.sure it is), and had two loose cables (repair effort ?). I
detected a level of frustration and he was quite happy to sell it to me. 

Replugged the loose cables and powered it up on a variac, nothing
unusual, glad to see good B+ (power transformer is OK). Dug around 
in the junkbox for a 600 ohm to 8 ohm transformer and actually found one.
Sure enough, very low audio with both the RF and AF gain fully advanced.
Tested all the tubes (all good), borrowed a manual, and later downloaded
the excellent Y2K R2 version. Found the two jumpers on the back for
remote AGC and RF Gain missing. After making some from aluminum 
scrap, there was plenty of audio noise, but still no signals.

The mechanicals are impressive but several of the slug racks would cam 
up and stick. Greased all the bearing points with a little white grease
and
still had two which sometimes stuck. Removed the two slug racks and
cleaned the slugs and the inside of the coil forms of all residue. They
all now follow the cams.

On the advice of several R-390A experts, pulled the electrolytics and
rebuilt them with new internal parts before they went bad (been there,
done that, didn't like it). Found some square aluminum cans with octal
plugs to use. 

Measured the 2nd oscillator and TP E402 is low, but when connecting the
scope ground, tripped the GFI. Found the chassis to be slightly AC "warm"
and was informed by several people that R-390As trip the sensitive GFIs 
we have today. Replaced the AC filter with a 3A junkbox unit which has 
an IEC 320/13 connector. Made an aluminum adapter plate (no extra 
holes, reversable). Had to relocate the bathtub capacitor a few inches to

make room, again no new holes. Shaved a little plastic off the mating AC 
cord plug to allow full insertion. Works great, no more tripped GFI, no 
more AC "warm" chassis.

The 2nd oscillator levels are -2.5V to -8V at the TP and will require
further investigation but you should hear "something". Tried all the
bands and frequencies. Looks like the 1-2Mhz band "works so-so" but
nothing else. You can barely hear the calibration signal with RF and 
AF gains advanced on the other bands. 

Time to take the unit apart because I want to add all the proactive
changes and improvements anyway. For a complex receiver, it was
easy to disassemble the modularized unit ..�even the RF deck. 
Most difficult task is to get the Bristol wrenches to bite (almost an
interference fit). Made a "long" Bristol from a section of fiber rod,
drilled a hole in it for the small end of the Bristol, and attached it
with two small cable ties �works well. Start checking the various sub
assemblies for bad capacitors, changed resistor values, etc �. can't 
find any values outside 10-15%. 

Proactive circuit changes:
  Dallas Langford's two diode SSB enhancement
  Chuck Rippel's change C-327, C-553, C-531, C-547, C-549, C-609
  Add a "no holes" adapter to use the balanced input connector for 
  higher sensitivity

Experimented around with the 2nd oscillator, changing values, better
6AK5, power it up on the bench and am not able to get -4V to -11V levels
at the E402 Test Point (six bands are above -4V). Maybe it's sluggish
crystals. Four of the crystals have one pin which is very loose. Pulled
the four sockets and replaced the pins (one came out in two pieces the
others had no spring action). Noticed that the registration of the number
tape, at the front of the deck, is "off" by a little more than 1/16" to
the left. Since the 32 switch contacts are very closely spaced, I put an
alignment mark on edge of the hole to allow proper alignment in the
future. Noted the min/max capacitance position of the trimmers since you
don't want that to be the peak. All have two peaks �OK. 

Checked all the cables by wiggling them while measuring continuity and
found one intermittent �.P-218 (one of the loose cables when I got the
unit). Turns out the center pin is not soldered to the wire. Easy to fix
but the radio is not intermittent, it's dead.

While examining the RF deck a gear fell out. Took a while to find out
where it goes. Guess what ? �.it's off the RF bandswitch shaft which is
"fixed" in the 1-2Mhz position. The gear clamp is busted and a new one
ordered from Fair Radio. Ordered one of their reproduction top and 
bottom cover sets too. 

Time to decide which mica caps are to be replaced because many of them
"fail" at 500V in the "mica" cap position on the Heathkit IT-28 capacitor
checker. On my IT-28, the eye closes at roughly 0.1uA for the
"paper/mica" cap position which is a little on the sensitive side (an
Eico capacitor checker I had, closed at 1.5uA). So here is a better way:
after determining that the voltage settings on the cap checker are
"close", insert a digital uA meter in series with the leads and read the
real leakage at the specified WV. The C327 mica (Chuck Rippel says it's 
a problem capacitor) read 2.5uA �.clearly a bad mica. I replaced it with
another mica because it's part of a tuned circuit and disk ceramics move
around too much. Most of the mica capacitors which closed or started to
close the eye on the IT-28, read less than 0.1uA at 500V ...perfectly
good. One more bad mica (also used at B+ levels) was C286 which 
measured 4.0uA leakage. That one will hose up the bias levels at the 
2nd mixer.

Received the gear clamp and re-installed the gear on the bandswitch 
shaft (glad I'm not a neurosurgeon) . Now the bandswitch is operational.
Checked cam alignment in the 7+000 position, looks "OK". 

A little light lubrication since you can access everything while it's
apart, some contact cleaner, and put the unit back together.

Had to make a tool for installing the Osc deck "third screw" if
you-forgot-to-install-it-before-mounting-the-front-panel. Powered it up
�.Nothing !! Not even the Broadcast band. Found that the Osc deck tube
was not lit. Power at the Osc deck plug is OK, turns out the RF choke in
the tube filament lead opened up. Fixed it and tried again �.Works !!
Tuning was very stiff �found the dial lock was not installed into the
front panel detent, putting pressure on the disk.

A spot check of alignment looks good so we'll use it a while and get
familiar with what else needs looking into. The tuning is still "too
stiff"
and I think the source is the VFO. The S-Meter acts "very damped" and 
the AGC does not respond as it should for the three positions.

Listened to some 20m SSB and a few European SW stations using a 20ft
piece of wire on the floor. Nice Receiver, but I guess you guys knew
that.