Restoration Notebook: Restoration of a Philco Model 38-14CB

In January 1938, new versions of Model 38-14 became available. One was in a brown Bakelite cabinet, while the other was in an ivory Plaskon cabinet. The brown Bakelite versions aren't too hard to find, but the ivory Plaskon version is seldom seen.

I really like the design of this particular series of Philco cabinets. At present I own a 38-12CBI, a 38-14CB, a TH-3, and a PT-30. Others sharing this style of cabinet include the 38-12CB, 38-14CBI, 38-15CB, 38-15CBI, and 40-90CB. There are three Philco Tropic models that use this cabinet (40-710T, 40-2710T & 41-705T). In addition, there are at least five Canadian Philco models in this cabinet style (39-3A1ACB, 39-3A1ACBI, 39-3A5ACB, 39-3A5ACBI & 39-3B4CB), and at least one postwar model (Model 200), which was sold in New Zealand and Australia around 1948.

The 38-14 is a five tube AC/DC set which receives the standard AM band as well as shortwave from 2.3 to 8.4 mc.

In 2009, I won an online auction for a Bakelite 38-14CB. Fortunately, unlike a similar Model 38-12CBI, this 38-14 was packaged correctly and arrived in one piece instead of in pieces.

The photo above shows how the 38-14CB looked before I began to do any work to it. The grille cloth is in good shape; it just needs to be repositioned and reglued. A good cleaning wouldn't hurt it, either. And that goes for the entire cabinet, not just the grille cloth.

Here is a front view of the unrestored 38-14 chassis.

Looking at the 38-14 chassis from the top. The chassis is in pretty good shape overall; it is dusty, and there is a lot of dirt/film build-up on the tubes.

Viewing the 38-14 chassis from the back, we can see some rust on the 75 tube shield. Notice that this radio only has two of its original tubes: the 6A7 on the left side and the 43 on the right. The 25Z5 rectifier is a newer replacement with a tubular glass envelope instead of the original taper-top style.

An under-chassis view of the 38-14. As you can see, I started replacing capacitors before I thought to take a "before" shot.

There is rust on the chassis mounting "feet." This set must have been stored in a damp environment at one time.

Since so many wires connect to the electrolytic capacitors, I decided it would be best to rebuild these instead of trying to connect replacements under the chassis. Besides, rebuilding old aluminum can electrolytics results in a much neater-looking restoration.

To my surprise, the original wet electrolytics still had liquid in them.

The photo above shows the original input filter capacitor. In the original Philco service bulletin for Model 38-14, it calls for a 16 uF unit. In a later production change, Philco increased the value of the input filter from 16 uF to 40 uF.

I am going to use a new 47 uF cap in this position.

In this photo, I have removed the original positive terminal connector by clipping the positive electrode off and then uncrimping the terminal from the aluminum electrode. The negative terminal is laying at lower right, and the cardboard insulator is at center right.

I cut the electrolytic can open with a Dremel tool - very carefully, since the liquid was still inside.

This is how the original electrolytic looked.

In the photo above, notice how the black cover has split at the top. This was caused by leaking electrolyte.

This picture shows the new 47 uF cap that will go inside the can. At the top of the photo, the thick white wire is a piece of solid 12 gauge wire. Part of this wire will become the new positive "electrode."

The positive lead of the new cap is soldered to a piece of the 12 gauge wire.

Notice in the photo above that I have drilled a small hole in the base of the electrolytic can. The negative lead of the new cap will pass through this hole.

A smaller wire is soldered to the negative lead of the new electrolytic.

Enough layers of heat shrink tubing and/or tape are added to the 12 gauge wire to provide a snug fit for the positive "electrode" in the base of the aluminum electrolytic can. The small wire has been run through the hole in the base. The positive electrode is held in place with some hot glue. The can is ready to be put back together.

The can is glued back together with epoxy. I intentionally left the joint flimsy, so it can be easily broken when replacement of the electrolytic becomes necessary.

I also glued the black covering back down at the top where it had split and was trying to peel away.

Except for the hole in the base of the can for the negative lead, the electrolytic looks much like an original.

The 38-14 chassis, upside down, can be seen in the background. Notice how I have cleaned all of the tubes.

The newly rebuilt electrolytic can be seen at center left in the photo above.

Now, it was time to rebuild the output electrolytic capacitor.

The output filter was bolted to the chassis. It is of a smaller diameter than the input filter. I had never rebuilt one of these smaller electrolytics before.

I was very lucky here: The black cardboard cover merely slipped off this electrolytic can. Most of these cardboard covers are securely glued to the cans, and are often difficult or impossible to remove. This cover obviously had been glued originally (you can see the glue residue on the aluminum can), but the glue had completely let go.

The value of the original was stamped on the aluminum can: 20 uF, 150 volts.

I drilled a hole in the base of the can for the negative lead of the new electrolytic which will be put into this old electrolytic can.

As before, I cut the can open with a Dremel tool. The innards can be seen above.

A new 22 uF cap will be used inside the output filter can, along with another piece of 12 gauge solid wire.

The positive lead of the new 22 uF cap is soldered on to the 12 gauge wire. This new positive "electrode" is then covered with three layers of heat shrink tubing. A smaller wire is also soldered to the negative lead of the new cap.

I decided to bolt the base of the filter can back onto the chassis, and then proceed to rebuild the can.

As before, enough layers of heat shrink tubing were placed over the new positive electrode to make it fit snugly in the hole it passes through in the base of the can. The negative lead was slipped through the small hole drilled for it, and the assembly is held in the base with hot glue.

The output filter can was glued back together, then the masking tape holding the parts of the can together was removed. Then, the black cardboard cover was slipped back over it.

I then soldered the positive terminal to the new positive electrode, then reattached all of the wires which connect to the output filter.

I attached an AC cord and a new cloth-covered wire for the antenna lead.

The photo above shows how the 38-14 looks with all new paper and electrolytic capacitors.

I only replaced two of the resistors.

The set did not work correctly upon initial power-up. I was able to tune in AM 740 by only holding on to the antenna lead, but it buzzed loudly and the volume control was very erratic.

Simply cleaning the volume control solved the problem; the buzzing disappeared and the volume control now works smoothly.

The set was easy to align. It is extremely sensitive; picking up local and area stations with just its short (1-1/2 foot) antenna lead. I can just imagine what it will do when it is connected to a longwire antenna!

I fixed the dial lamp socket using the method in this site's Tech Section.

Unfortunately, the speaker - which was a non-original replacement - sounded terrible; the voice coil was rubbing against the pole piece, causing severe distortion in the sound.

In my reserve of parts chassis, I discovered the remains of a 37-602 with a good speaker. This was a cheap radio I bought a long time ago; it had a very poor cabinet which was not salvageable, and was disposed of.

Notice that this chassis has no tuning capacitor. This is how the chassis looked when I pulled it from its cabinet.

With the original 37-602 audio output transformer, this unit sits 5-1/4 inches above the chassis.

Unfortunately, the 37-602 speaker will not fit in its original configuration; its audio output transformer sits too high to clear the inner top of the 38-14CB cabinet (as indicated by the red arrow). At 5-1/4 inches, it would be jammed against the top of the cabinet. And this is not acceptable.

I apologize for the blurry photo. I took this picture with my left hand while I held the tape measure with my right hand. You see, I am right-handed.

I could have used the 37-602 speaker by removing the tall audio output transformer and substituting the original 38-14 audio output tranny. However, the field coil and bracket of the 602 speaker took up too much real estate on the 38-14 chassis.

Then I remembered I had a Philco 47-204 that was in pretty bad shape...

This Philco 47-204 could be restored, but the cabinet had holes drilled in the side. Plus, the chassis is rough and very dirty; it looked as if it had been a home for mice at one time. Fortunately, they did not bother the speaker which appears to still be in excellent shape.

The 47-204 uses a permanent magnet speaker. It is a little smaller than the original 38-14 speaker, but (a) it was handy and (b) would take up much less space on the 38-14 chassis.

The 38-14 uses an electrodynamic speaker. However, the speaker field in the 38-14 is not needed in the set's power supply; it was only used to energize the speaker so that it will work. Since the replacement speaker has a permanent magnet, there is no need for a field coil in the 38-14, so we can do without it.

I needed a speaker that had some sort of mounting bracket, so it could be easily mounted to the 38-14 chassis. As you can see here (in spite of the blurry photo), the 47-204's speaker has the mounting bracket which I wanted.

In the photo above, I have removed the non-original speaker from the 38-14 chassis. As you can see, there is not a lot of room here for a speaker unless it is rather small and takes up very little room.

I drilled two holes in the chassis so that the 47-204 speaker could be bolted onto the 38-14 chassis in the same way it was in the 47-204 chassis.

Here, you can see the 47-204 speaker has been bolted onto the 38-14 chassis from underneath.

I had to move the filter choke in order to drill the holes and attach the speaker mounting bolts.

In the photo above, the 47-204 speaker is now hooked up to the 38-14 chassis. It sounds great!

With the non-original electrodynamic speaker that was in this radio, the set still had some residual hum. However, with the 47-204 speaker, the 38-14 is completely hum-free.

Top view of the 38-14 chassis with the replacement speaker. The set was playing as I took this picture.

And now, with the speaker having been successfully replaced, the 38-14 is put back together and left playing on the bench for awhile.

I still need to pull the chassis back out to address the rust issues on the chassis mounting feet. I will also need to clean and polish the Bakelite cabinet and knobs, clean and re-glue the grille cloth, and then put it all back together. But for now, we can consider this radio to be "finished."

Important Modification of the Power Switch Wiring in Model 38-14

The power switches on most transformerless (AC/DC) radios were wired so that the neutral side of the line was switched. This is not desirable from a safety standpoint. The hot side of the line is what needs to be switched, so that the hot side of the AC line is disconnected from the radio when it is turned off.

In order to completely accomplish this, we not only need to move the switch to the hot side of the line, but we also need to use a polarized line cord. This is the type of cord with one blade wider than the other.

In the original 38-14 schematic diagram, you can see that the power switch is in series with the B- line of the radio.

An under-chassis view of the 38-14. This shows the original factory wiring of the switch and the power line.

The red lines show how the AC line cord connects inside the radio chassis. The neutral lead goes to the first terminal on the left of the long terminal strip in the center of the chassis. A wire goes from here to the power switch. Another wire runs from the other terminal of the power switch to pin 1 of the 75 tube.

The hot lead goes to pin 5 of the 25Z5 tube.

Notice also the blue line, which goes from pin 5 of the 25Z5 tube to the ballast resistor (36).

In the partial schematic above, I have changed the location of the power switch from the neutral side of the line to the hot side.

The pictorial above does a better job at illustrating what needs to be done.

As in the earlier pictorial, the red lines indicate the AC line. The neutral side of the line now connects directly to pin 1 of the 75 tube.

The hot side of the AC line is now connected to the first terminal on the left of the long terminal strip in the center of the chassis. A wire from here goes to the power switch, as before. However, the other side of the switch now connects to ballast resistor (36) as shown above.

The blue line remains at the same terminal of ballast resistor (36), and the other end is still connected to pin 5 of the 25Z5 tube as before.

A final note. Be sure that the ribbed wire in the line cord connects to B-. The ribbed wire is the neutral side of the line. The smooth wire in the line cord is the hot side of the line.

If your line cord does not have a molded-on plug, be sure you attach a polarized AC plug to it! And remember, the ribbed wire connects to the wider prong on the plug. The smooth wire connects to the narrow prong on the plug.