Philco made this model for the 1931 season and it was Philco's
first superheterodyne. It has eleven tubes, four 24's, four 27's, two
45's in the pushpull output and an 80 rectifier. The cabinet design
is by Norman bel Geddes (1893-1958) one of the great industrial
designers of the twentieth century. (bel Geddes also designed the IBM
THINK sign among thousands of other items!)
This Philco 112
was an accidental Ebay win. I bid the starting price plus a dollar
and no other bids were forthcoming. The seller dissasembled the radio
and packed it in three well padded boxes. It arrived undamaged! The
cabinet was in a large box with the speaker packed inside in another
box. The base was removed and packed in another box. The chassis was
in another. All were double boxed. UPS would have had to REALLY try
to break this radio!
The chassis electronic restoration was
fairly straight forward. I restuffed all nine bakelite blocks, the
two metal capacitor cans, the tone control and the main filter
capacitor box. The bakelite blocks can be difficult to restuff. I
used a modified version of Chuck
Schwark's procedure. I did not unsolder or cut any wires. The
tiny wires that connect the internal components to the lugs were
broken by twisting a drill just slightly larger than the rivet hole
by hand. Then the hole was cleaned with a drill that would pass
through the hole. The hold down screw was removed, the block tilted
away from the chassis and a hair dryer used to heat the block for a
couple of minutes. The smaller drill was then used to push the tar
block out of the bakelite block in one piece. The new caps and
resistors were then placed in the block and the leads soldered to the
lugs. It takes about 10-15 minutes per block using this method.
The
filter choke was open so I rewound it using some number 32 magnet
wire. The wire I had was larger than the wire originally used so the
number of turns that could be wound was less. This made the
inductance and resistance smaller but it works fine since the
original 2uF filter caps were replaced with 10uF 450 volt
electrolytics.
I then applied power and started the
troubleshooting. Audio was very weak but some strong local stations
could be heard. Bias, B+ and filament voltages all were reasonable.
Using the scope, I determined that there was a good IF signal into
the last 24 plate circuit but almost nothing on the grid of the
detector. I powered down and did a little checking with the ohmmeter
and found the secondary of the last IF transformer was open. The IF
transformer could be repaired I suppose, but instead I chose to
couple the primary signal to the detector via a 100 pf cap and a 150K
ohm resistor. That brought the audio up nicely.
I left the
radio playing while working on the cabinet. Suddenly the audio became
distorted and weak. A quick check with the voltmeter indicated
something very wrong in the output stage.The grids of the 45's were
100 volts positive! They should be 45 volts negaive. I unplugged the
chassis and did a few checks with the ohmmeter and determined that
the connection to the center tap of the interstage audio transformer
was open. The transformer had continuity end to end so it was still
delivering a signal. However the minus 45 volt bias is applied to the
center tap. That left the 45 grids floating and they had assumed a
potential somewhere between cathode a plate voltage. As a quick
repair, I connected a 100K ohm resistor from each 45 grid to the
minus 45 volt bias terminal and tried it. The radio sounded great and
all voltages were normal. Rather than risk destroying the delicate
interstage transformer, I chose to leave the resistors. A schematic,
with the changes highlighted is here.
Here is a very complete explanation of the change by Don Black
as posted on the Antique Radio Forum.
"This will work
and the center of the winding will be at the same potential as the
resistor center tap provided there's no significant grid current. It
forms a bridge circuit with the two halves of the secondary as one
arm and the two resistors as the other. They will remain at the same
voltage provided nothing else unbalances them. If you're confused by
the apparent voltage division, consider that there's twice the
voltage across the full winding as each side so the voltage divides
equally over the winding and also the resistors. Just another way of
looking at the bridge balance. The resistors will load the
transformer somewhat and reduce the stage gain but they will also
damp the windings reducing peaks and extending the bass response.
It's similar to reducing the plate resistance of the driver tube as
far as damping is concerned.
Don Black."
I used
builders nylon twine to restring the dial cord and a faucet spring to
replace the rusted out originals. The cord must be very tight to
prevent the weight of the capacitor rotor from detuning the radio. I
found this out the hard way. I originally had used some springs that
were too weak and the tuning was very 'iffy'. I even considered using
car wheel weights to balance the rotor! There were two springs
installed originally. I used one strong faucet spring that had been
intended to be in compression.
The final step in the chassis
restoration was alignment. The alignment trimmer capacitors are
accessable through holes in the top of the chassis. I made an
alignment tool from polybutylene plumbing tubing by heating a 1/4
inch Allen wrench and forcing it into the tubing. It made a good
insulated adjustment tool. There is B+ voltage on the trimmer nuts!
Don't try adjustment with a metal tool!
The cabinet base was in several pieces and had to be
reassembled, glued, sanded, stripped and refinished. I bolted it back
on to the cabinet and stood the cabinet vertical for the first time.
It took some time to decide what course to take in the restoration.
One thing was sure. The joints needed to be reglued. The old hide
glue had lost much of it's 'cling'. I removed and reglued all of the
gussets and forced glue into all loose joints. After several stages
of glueing and clamping over the course of about a week, the cabinet
was solid once again. I added two additional gussets at the top rear
since that is a classic weak point in most cabinets. The top can pull
off easier than you think!
I chose to strip the top because of
flower pot water damage. After striping, the top was sanded with 200,
320, 400, 600 and finally 1500 grit sandpaper. I then applied many (I
did not count them.) coats of clear lacquer sanding between coats
until smooth then sprayed one light coat of Mohawk Perfect brown
toned lacquer. I did pretty much the same to the cabinet front but
chose to just sand the original finish on the sides and respray with
toned lacquer. The grain pattern on the front was almost completely
hidden until stripped. I can't believe that the cabinet manufacturer
would hide a beautiful grain like that but I suppose that was the way
they did it. The factory lacquer could have darkened or maybe there
was just a buildup of seventy-five years of wax and dirt. After the
last lacquer coat was applied all surfaces were sanded with 400, 600
and 1500 grit wet sandpaper. The last step was to apply a red
mahogany stain to all surfaces. The stain filled all of the tiny
scratches left by the 1500 grit paper and left a very smooth finish
that almost appears three dimensional. After the stain was throughly
dry I used clear coat auto polishing compound for the final buffing.
The grill cloth is original. I thought it was a replacement
until Ron Ramirez (Mr.
Philco) saw a photo and told me it was original. It was faded but
undamaged so I reused it. To clean the grill cloth I used Arm and
Hammer powdered upholstery cleaner and a vacumm cleaner. A restorer
from South America (don't know which country and lost the URL) had a
writeup on his website that discussed using lacquer on a grill cloth
after dyeing it. I was afraid to wet the cloth or even pull it off
the backing so skipped the dye step. I have had cloths come apart too
many times. I sprayed the cloth with clear medium gloss lacquer and,
after drying, it tightened on the backer and was strong and stiff yet
still was porous so sound could pass freely. The lacquer should last
a bit longer than starch and also not attract vermin.
Update 04-21-2005: I gave this radio away.
