"R-390A Overhaul Steps..."
Well, after spending months slowly going thru my two R390A's, one is
finished. Below is an outline of the steps that I took during my
overhaul. This was probably one of the more evolved R390A "overhauls"
done in the South. (grin) The other, will take longer, I'm probably
going to replace all of the bushings in the RF deck amoung other
things that I didn't do with this one.
The victim: I started with a cherry 1967 EAC contract model that was
the "lowest mileage" R390A I've ever seen.
(insert Dr Frankenstein laugh here) Jerk all of the modules out of it and rip it's gizzard out and scatter and toss the parts around!
- All of the original modules, meters, covers, etc. were still on it.
- In addition, all of the tubes in it had date codes within a two or three month period of each other
- Even with clean gears, there wasn't even a hint of a wear
pattern in any of the gears and all of the aluminum finish in the
tracks on the RF deck was still intact.
- The green paint on all of
the module hold down screws was even 100%.
- I'd be surprised if this
thing saw more than a few hours operation after the burn in period.
- There's no diode load hole in the front panel or adjustment hole in
the top dust cover for the meter adjustment.
I tried, but I managed to not loose any of
the parts and didn't even have any extra ones left over. (grin)
Power supply module:
- Tested the dial lamps, checked the value of all of the
resistors, the diode, the 2 capacitors, the meters, tested the selenium
rectifier, and the antenna relay and inspected the contacts in the
- Verified function of the main power micro switch, it's placement,
and measured the resistance of it.
- Measured the resistance of all of
the front panel switches and potentiometers, and very lightly lubed
the shafts of each, checked the ovens switch, verified the values of
the fuses, checked their resistance, replaced one of the fuse holders
that I didn't like the look of with a NOS one, and replaced the rear
panel IF connector, the center terminal was missing.
- Checked the line
filter, and checked the tightness on all of the screws holding the
whole damn mess together.
- I removed all of the knobs and inspected and
lightly lubed the set screws.
- Also verified that the index washers were
installed in the two big knobs that prevent the clamps from turning.
- The next step ate up a lot of time. I removed all of the hoods of the
chassis connectors to inspect and then measured the resistance of EVERY
damn wire in the chassis.
- Then I installed a NOS military 3 wire 8 foot
rubber 16 gauge power cord with a molded plug. Nice and flexible SJ.
The original strain clamp for the cord was still there. First one I've
ever seen. :-)
PTO: Why bother? Chunked it in the R390A parts pile and installed an
Army rebuilt Cosmos that I've been sitting on for about ten years,
sealed in the box, to replace the Cosmos that was in there. It
turned out that the endpoint was out a little less than 2 KHz, and
linear within a little less than 200 Hz across the spectrum. I don't
know what the specs on it where when it left their hands in Feb. of
1984 but it sure aged well. :-) I guess that I'll let it run a few
weeks and then adjust the endpoint. I did power up the oven and
verified that the thermostat worked, measured the resistance of the
transformer, and tested the tube. I like the Cosmos PTO's. That blue
label sure is pretty, bubba!
- visual inspection,
- resistance readings of the transformer windings and wiring,
- inspected the solder connections,
- verified 115 volt setting,
- stuffed two new 26Z5W's in it.
- Checked all of the screws and nuts for tightness.
On a side note, I probably use a bit more complicated method than
most people do when fitting a PTO to a receiver. If you're going to do
it right, it might as well be done right the first time.
Crystal Oscillator Module:
- Both halves
of the oldham coupler should be perfectly parallel to each other and
the centerline of both shafts should be perfectly in line with each
other on both planes.
- I only spent about an hour adjusting the position
and height of the PTO in the chassis, but spent several hours measuring
the components and setting up the fixtures to measure other aspects of
- I first measured the run out of each half of the
couplings while installed on their shafts. The one on the PTO was true
within .001. The one on the KHz shaft of the RF deck was machined
improperly. he rib was .003 off to one side and wasn't square with the
bore either. I tossed it and pulled a few others out of spares. It
took several before I found one that was square with the bore and only
had a little more than 0.001 run out.
- The next step was to check the
center section. The width of the two groves seems pretty consistent,
but I was curious if the two groves were machined at exactly 90 degree
angles to each other. The groves were of a dimension that I didn't
have any key stock for so I used two 12" long pieces of 1/4" ground
steel rod. I centered the two pieces of rod, lengthwise, one in each
of the two grooves, clamped the three pieces together. Then by measuring
and comparing the distances between the four rod ends, I could determine
the exact angle that the grooves were from each other. This part of the
process was a wasted effort, the coupling center piece from the EAC was
90 degrees like it should be and so where the ones in spares that I
- When I finally assembled the receiver and physically aligned
the PTO to the chassis, mechanically and electrically, the dial indicator
measured a total movement in the center section of the oldham coupling
of .003 when the KHz knob is turned. Close enough! Put that spring on!
Oh, I used a little dab of Penzoil wheel bearing grease to lube the
coupling. It's red and contrasts, in a pleasing fashion, the blue label
of the Cosmos PTO. :-)
- Tested the tube, and tightened the screws
holding the tube socket to the chassis, they were loose.
resistor values, transformer windings and crystals. Bad 10 MHz
crystal, throw one in it from spares. Most are still on the money,
the few that are "off" are well within 1 KHz or maybe a shade more.
I decided against spending ~250 dollars on new ones. (grin)
the resistance of all of the wiring and switch contacts and tested
all of the fixed capacitors and spun all of the trimmers a couple
- Powered up the oven and verified function of the thermostat.
Also, "timed" the two switch bodies. They were "off" a bit.
when thru and re tested everything on it's underside just to make
sure. I figured that anything that was a pain to remove, I double
- One of the original plug in electrolytic caps showed
signs of leaking, tossed it in the trash, installed another one from
spares. Reformed both, leakage at 50 volts over the rated voltage was
less than 1 ma. per section after reforming. I fused them and ran
them for a couple of weeks at full rated voltage on one of the HV
supplies, they didn't explode and leakage declined even further. Good
enough, bubba! (yeah, I know, Doc, but the power factors were good,
I even checked that. :-)
- Ripped all of the paper capacitors out of
the module, and tossed them in the trash. Installed two new .022 400V
orange drops in the location that Chuck likes, and NOS Vitamin Q's in
the other locations. I have the orange drops on hand and could have
used them thru out but didn't like the way they sit on the circuit
board. I did use an Orange Drops to replace the one on the chassis
under the circuit board. All of the new caps were tested for leakage
at their rated voltage and tested to verify their value before
- Checked all of the resistors for value, replaced a couple.
Tested the mica cap, no problem there.
- Tested all of the tubes, they
all passed but tossed the 0A2 and stuffed a new 6626 in it's place. I
don't trust used 0A2's, had some weird problems with them.
- Tested the
relay and measured the resistance of the wiring, the chokes, and the
transformers. I left the 800 cps filter alone. Probably not a whole
hell of a lot of R390A's out there that still have their original
6AK6's. All of the tubes are original except the rectifiers and the
Whew! The last one, the RF deck: After removing it, the first step was
to take it apart.
- Tested the tubes and the 3TF7.
- Measured the resistance of
the wiring, the transformers, the switch contacts, and the resistors.
I replaced more than a half a dozen resistors that were out of spec.
- Checked the capacitors and resistors inside the IF transformer cans,
that could be tested. Some could not be tested in circuit.
- I tested
the big above chassis oil filled capacitor for leakage and value.
- Tested all of the mica capacitors for leakage and value and then
ripped all 18 or so of the axial lead paper capacitors out of the
module and tested them just for kicks. EVERY "brown beauty of death"
tubular capacitor that was in it leaked like hell and a good percentage
had microscopic cracks in the bodies within maybe .020 of the seams
and paralleling them. Most of these caps leaked at voltages below 50
volts when tested. Only one of the metal can axial capacitors leaked
when tested. I replaced all 18 of the capacitors with Orange Drops.
For the .1 and the .033 values I used 400VDC rated ones and for the
.01 values, I used 600VDC rated ones. The reason that I didn't use
6600V rated ones thru out was their size. It was a pain in the ass to
the fit the ones that I used in there properly. If I'd have used the
600V ones everywhere, I'd have had to move the locations of some of
the capacitors and a bunch of them would have had excessively long
leads. I didn't think that this was such a hot idea in the IF section
and figured that the best placement of the parts was in the original
- Let's see, other than checking all of the screws and
nuts, I think that was it for the IF deck other than lightly
lubricating the shaft extensions where they passed thru the front of
the IF module chassis.
- I didn't test the mechanical filters.
- I tested
the blocking cap before I tossed it and it had tested good even at
100 volts over it's 300V rating.
I fired it up and let it cook a while in Standby mode, at 7+000. None
of the magic smoke escaped so I switched over and set the PTO to 2455
KHz and tightened the clamp. I stuck a VTVM lead into the unbalanced
antenna connector and cranked it down to WWL on 870 and let it run more
than a day before I did the first alignment. I always like to align a
receiver twice. I go thru it and then when I'm finished, I start all
- I removed all of the tubes and tested them, the crystal
oven and tested it, all of slug racks and springs, four of which (for
the variable IF slug racks) were really weak, so I installed NOS ones
in that location when I put everything back together. The geometry for
those 4 springs suck, they're stretched a lot more than any other
- I removed all of the RF coil assemblies and measured the
resistance of all of the windings and checked what capacitors I could.
The bridge wouldn't work on some, so I kept track of those in case I
had some weird assed problem when I tried to align it later.
- I disassembled the gear train and tossed all of the parts, except
for the counter, in a coffee can and sprayed a mess of gunk in there
and let them brew. They weren't really dirty, but the original lube
had mostly evaporated and what was left was stiff as hell and I don't
really find the gear train much of a mechanical challenge so I ripped
- About the only thing I didn't take apart was the 6 camshafts
and the antenna trimmer can.
- I worked a few drops of penetrant into the
bearings of the cam shafts and kept lubing and wiping them until only
clean oil would come out.
- Oh, I used 10W30 Mobil 1 synthetic oil for the
RF deck except the detent where I used Penzoil wheel bearing grease.
Two of the cams appear to have been stamped, I guess, with cracked dies,
leaving a couple of sharp burrs on the surface that the rollers ride on.
I stoned these down while maintaining the original cam profile. :-)
- When you take the split gears apart, tie them together, with a bit of
soft wire in the orientation that they were originally assembled with.
I suspect that the halves were matched.
- While all of the stuff soaked, I replace the three paper capacitors,
with Orange Drops, and replaced close to ten resistors that were out of
spec, checked all of the other capacitors and found a cracked .005 1KV
- And yes, I measured the resistance of all of the wiring
and of the band switch. ;-)
- I found an odd thing. One of the tube
sockets only had one screw holding it to the chassis. When I attempted
to install a screw there, it turned out that the little "C" shaped
piece of metal that curves all of the way around one side of the
socket had an unthreaded hole in it for the screw. I'm surprised that
an inspector didn't catch this at the factory. I tapped the hole and
- Most of the gear clamps were either viably cracked or
showed cracks when dye checked. I guess that they must have been
over tightened when it was built. I replaced all of them with NOS
clamps to be safe.
- I found that several of the roller retainers had been over staked on
on a couple of the slug racks. This prevented the rollers from turning.
In addition, a few of then ends were not square and had to be
straightened. Burrs and gouges on the end surfaces had to be stoned
down and polished.
- The fit and finish of mechanical portion of this
EAC RF deck didn't impress me at all. The old Collins decks were much
more finely finished mechanically.
- I wiped each of the RF cores out with a pair of damp Q-tips, wiped
the slugs off, and eye balled them. The Collins part numbers on all
of the RF slugs are all the same EXCEPT the six variable IF slugs.
They are different from the RF slugs. So, they aren't interchangeable.
- I assembled the RF deck and mechanically aligned it and put the
receiver back together. For what it's worth, the repeatability of
the RF slug racks averages about .001, the repeatability of the variable
IF slug racks averages .004 on one and .005 on the other. I suspect that
this could be improved upon by relocating the location of the attachment
point on the chassis of those four springs. This would require either
shorter springs or possible just creating spring "wells" that extent
slightly below the chassis so that standard RF deck rack springs could
I've been playing with it for about a day and a half since the alignment.
This is the most sensitive receiver I've ever owned. It kicks ass. I did
a few sensitivity tests using my URM-25F. I questioned the results so I
dug out the URM-25D and tried them again. REAL close.
I started with a receiver that hadn't been abused and tried to do the
best job that I could going thru it. I wanted something that I didn't
have to screw around with every few weeks. Something that wouldn't wake
me up at night with a burst of light like a Romulan disrupter (I've had
R390A's do that before). Something that would sit there and run for
month after month and need nothing but tube and dial lamps like my
R-1051B's. Hopefully, this will do that. Many of the of the steps that
I took, were "over kill", but I had fun doing it and learned a few
The numbers you ask? Lets just say that they're as good as the best
sensitivity levels that I've ever seen posted or in print on the R390A.
Numbers, that up until now, I always had my doubts about.