DuMont Model RA-103 Television/Radio (1948)

Here is a DuMont model RA-103 television from 1948, a 12-inch Teleset in the Meadowbrook console cabinet. DuMont prided itself on quality and had prices to match. This TV cost $525 when new, which equates to thousands of 2014 dollars. It includes FM radio and a "magic eye" tuner. With a generous 10-inch speaker, this set has very nice audio for a 1940s television.

The moment I saw this classic set, I decided to give it a full restoration. This involved a lot of work spread out over several months.

RA-103 Design

DuMont offered the RA-103 television in five cabinet styles. Mine is the Meadowbrook. Click on these thumbnails to see Meadowbrook ads from the New Yorker and National Geographic:

The Meadowbrook is described as "a traditionally designed cabinet in mahogany finish and selected mahogany veneers. Unsurpassed television on a 72 square-inch screen. FM radio."

Here's an ad from a New York City newspaper. It lists dozens of DuMont dealers in the NYC area, including prestigious locations such as Steinway & Sons in Manhattan:

Here is my RA-103 Meadowbrook after restoration.

Other RA-103 cabinets include the Sutton, a blonde console shown in the newspaper ad, and the Chatham, a "doghouse" shaped tabletop favored by collectors due to its compact size. The Savoy and Colony consoles added a phonograph and AM radio. All of the RA-103 DuMonts receive FM radio.

My restored chassis appears below.

You can download the complete Riders and Sams service manuals for this set at the Early Television Foundation schematic archive. I got a free copy of the Sams Photofact manual from my local library. You can also buy it from Sams (folder 90-3). Both manuals are worth using, since each contains some information not available in the other. This article uses the Riders part numbers.

The RA-103 uses 25 tubes:

Like my 1946 RCA 630TS and others of the time, this DuMont uses a "split sound" audio system. You can read more about that design choice in the 630TS article.

The earliest Riders schematic shows a type 6U5 eye tube, but my set and others use a 6AL7. It's unusual to see a magic eye tuning indicator on a television. I have often joked that a TV already has a fantastic magic eye—that big thing called a picture tube!

The eye tube can be used for both radio and TV tuning. The function selector gives you the option to turn off the eye for television viewing.

Most indicator tubes have a round green eye with a dark "pie slice" that opens or closes as you tune in a station. The 6AL7 has two rectangular targets. One of them is always about 3/4 full and marks the ideal tuning point. The other grows or shrinks as you tune.

When the green targets match, you have tuned the station exactly. The middle photo shows the tuner centered on a station. The others show the eye when the tuner is slightly off the mark, either above or below the ideal.

Including FM radio was no marketing gimmick, nor did it add extra components or complexity to the TV. It arose naturally from their choice of tuner: the famous Mallory InducTuner, which DuMont dubbed the Inputuner. We'll learn more about the InducTuner later in this article.

Another notable feature is the time delay in the power supply. This involves an electromechanical relay, a 6AL5 tube, and a few other components. In the words of the DuMont service manual:

The relay circuit has been designed so that the relay is energized approximately ten seconds after the power is applied to the television receiver. In this way all capacitors and other components are protected from the high surge voltage which otherwise would occur before the tubes heated up and started to draw plate current.

Only DuMont would make such an investment to improve the television's longevity. Most 1940s manufacturers simply ignored the surge issue. Some later TVs, such as my 1958 Philco Predicta or 1961 CTC-11A, get a soft startup thanks to an inexpensive thermistor in the AC power line.

The RA-103 does not have an automatic gain control (AGC) circuit, a topic that we'll investigate later.

First Look

I bought my DuMont locally for $40. Here it is after unloading on the front porch. The cabinet is not in great condition. It is missing a little molding piece in the upper left. The top has warped and come loose, and the sides have warped and delaminated.

The chassis is complete and looks decent. Strapped atop the high-voltage cage is a slapdash repair—a large box capacitor wired to substitute for a filter capacitor in the power supply. The power transformer is visible at lower right. It is massive, almost six inches tall, needed to accommodate the two 5U4G rectifier tubes. DuMonts of the 1940s are known for excellent design and robust construction.

Carrying the chassis indoors (uff da!), I got a first view of its underside. Nothing scary here. A few capacitors have been replaced over the decades. The component layout is roomy, promising a carefree restoration.

The white material in this 5U4G rectifier suggests that the tube lost vacuum when it failed. Not to worry, in this deal I also got a box containing 39 good spare tubes, including two 5U4Gs. I love goodie boxes! The tubes alone are worth considerably more than I paid for the whole shebang. I also got four pages from the service manual, one of which is seen in this background.

The seller told me that all of the knobs had been lost, but I found this one at the bottom of the box of tubes.

To complete this project, I'll need four more knobs like this, plus a large tuner knob.

Cleaning Up

This weary traveler needs a bath and a shave. The usual drill is to wipe off dust and grime, clean the pins and sockets for every tube (testing each tube as we go), and use a little DeOxit to clean user controls such as volume. You can read more about those basics in the article First Steps In Restoration.

Note the eye tube lying on top of the tuner box. Remove this by reaching into the cabinet before you slide out the chassis. Otherwise, you may bust some of those pretty colored cables.

Another minor issue is evident in the previous photo. See how the base for the picture tube is hanging a bit crooked. It has come loose from the CRT neck, fairly common when old glue dries out. I'll resecure the base later with a few drops of cyanoacrylate (super glue). If I leave the base loose, that risks breaking the fragile wires leading into the glass neck.

Here I have turned the set and removed the cover from the high voltage cage. I see a dab of surface rust, but nothing seriously amiss.

Looking under the little phenolic board that holds the 1B3GT rectifier tube, I spied a small resistor that looks like it has gotten very hot. I make a note to replace it.

Here's an odd sight. DuMont put this little permanent magnet in the back of the high-voltage cage, right behind the 6BG6 horizontal output tube.

No doubt it is there to suppress Barkhausen oscillations, which appear as a dark vertical line in the left side of the screen. One of my old TV service books mentions using a spare ion trap magnet on a 6BG6 to effect the same cure. (My Hallicrafters T-67 restoration article shows a screen with typical Barkhausen's interference.)

The picture tube's high-voltage lead needs help. Where the cap enters the tube, someone has wound a couple of licks of electrical tape around the lead. With a gentle jiggle, I can feel loose wires inside. "Replace CRT lead," I write on my to-do list.

Twenty-two of the TV's twenty-five tubes tested OK. I replaced the three weaklings from the box of spares that I got with the television.

The 12JP4 picture tube looked very strong on my Sencore CR70 tester. On its bell is a sticker from a rebuild shop with the type number. A properly rebuilt tube can last as long as a factory original, so this one may have lots of life left. It looked bright, indeed, when I got a picture later on.

Clean Tuners Are Happy Tuners

The TV's tuner was immovably stuck with prehistoric lubricant, so I disassembled it for cleaning. I was eager to get a peek inside, anyway.

Removing the tuner demands care. You need to unsolder several leads under the chassis, as well as the cable leading from the antenna terminal to the tuner box. Then you remove five screws under the chassis and carefully pull the leads out.

This and most other service can be done with the chassis standing on its side, with the high voltage cage resting on the workbench. The service manual tells you to put the other (power transformer) side down, but that makes no sense to me. The chassis is perfectly stable this way, and the other way tends to crush the speaker cable coming out that side of the chassis.

You must also remove the dial lamp by sliding it off its mounting bracket. The lead to my lamp happened to break off when I moved it, so I left it in the following photo. The tuner is on the workbench, ready for service.

Next, we see the tuner with its hat off. The design is simple and elegant. Three rotating coils take you through the VHF television and FM radio frequencies. The workmanship is admirable. The coils are mounted on a ceramic shaft and the tuner frame is cast, rather than sheet metal.

Removing the second shield exposes the rest of the tuner. Inside are trimmers, little coils, and so on. You don't need to remove this for cleaning and lubrication; I was simply curious. Don't mess with any of these things or you'll need to realign the tuner.

I could see that I'd need to remove the dial, so I took a photo to record its position, which will be important on reassembly. As I later found, the tuner had stopped all the way clockwise. Notice how the two thirteens are aligned. You can also see how the dark inner dial is visible through the clear outer dial.

The two dials are geared in about a 10:1 ratio. It takes nine complete turns of the outer dial to make almost one turn of the inner dial, which stops before completing a full rotation. This TV has no fine tuner as such, but the outer dial is a fine tuning aid. Its markings include the FM radio frequencies and TV channels 7-13, occupying the dial's full rotation.

The inner dial shows every frequency covered by the tuner. Channels 1-6 are spread farther apart, while the FM frequencies and channels 7-3 are much closer together. This non-orthogonal spacing reflects the way this tuner works, not the way that TV channels are allocated in the ether. Every TV channel is six megahertz wide, whether it's channel One or Thirteen.

My outer dial has a radial crack. I doubt it will bother me that much, but I made a high-resolution scan of the dial while it was out, in case I want to make a reproduction some day.

The next photo shows an intriguing gizmo. Look at the circular plates with little ears, lined up in a stepped or spiral pattern. I couldn't tell what these did just by staring at them, and since everything was stuck solid, I couldn't find out by turning the tuner and watching.

Other DuMont owners told me that these plates create end stops for the tuner, which has to make nine revolutions to traverse the VHF band. Each plate can rotate freely, but the ears catch one another, leaving and interleaving as you go from end to end of the outer dial. There is one eared plate for every time the outer dial rotates. I hope they gave a bonus to whoever invented this cunning scheme!

Now I have removed the dials to expose the gear train. The outer dial is secured on the shaft with little setscrews. When you loosen it, the inner dial (with its gear) is also freed to slide off the shaft.

The old dried grease looks awful but cleans up easily with Q-tips and lacquer thinner. I used wooden toothpicks to dislodge bits of it from crannies.

After cleaning the surface mess, the tuner was still stuck securely. I dabbed penetrating oil into all of the bearings and let it marinate overnight. Do not use force to unstick this tuner. You don't want to damage anything, especially the ceramic coil shaft, which could be expensive to replace.

Once the mechanism was free, I carefully relubricated everything. Use only enough lubricant to let things move smoothly. Excess grease or oil will attract dirt.

Each of the three main coils has a wiper and contact pair on each end. Clean these, like other electrical contacts, by dipping a Q-tip in DeOxit and gently rubbing. Follow that treatment with electronic lubricant, again sparingly applied with a Q-tip.

The next photo shows the gear train and dials when I'm almost finished. You can see one of the setscrews on the outer dial's collar. You don't need to remove the inner gears unless something has been bent or broken.

Before putting the dials back on, I turned the tuner all the way clockwise, then aligned the dials as in the earlier photo I had taken. The tuner is ready to reinstall.

Shining up the outside doesn't make the tuner work any better, but it was my little reward for working a long time on the hidden innards. Now it works well and looks good, too.

InducTuner Variants

This tuner lists several U.S. patents on its cover. The patents, such as 2377790, define underlying technology but don't show this exact mechanism.

One of my old TV service books describes the InducTuner with this diagram.

Here, the contacts are on the bottom of the tuner frame rather than the ends of each coil, and the contacts move on "trolleys." I don't know whether this version was actually produced, but it certainly differs from mine. If you read 1940s patents for things such as variable inductors, you'll see many similar gizmos, often quite complex.

DuMont made a later InducTuner with more upright spirals. Here's a photo from the same book.

Mallory holds U.S. Patent 2666906 on this so-called spiral tuner. Here is one of the drawings.

Click the image below to see a brief video of the spiral tuner used in my DuMont RA-113. Refer to that article for more photos and information.

Continuous Tuner Pros and Cons

The InducTuner was a Mallory product, so any manufacturer could have chosen it. Other companies didn't rush to embrace the costly continuous tuner, however, or the integral FM radio that it enabled.

When a TV and radio were desired, it was simpler to bolt an existing radio chassis into a cabinet and operate it separately. Most TV/radio/phono consoles were designed in this way. My Scott 800B "combo" is an example. An advantage of this setup is that the radio can cover AM standard broadcast and shortwave frequencies in addition to FM.

FM radio stations are closer together than television stations. This tuner has a constant rotation, however, so you must move the knob slowly and carefully in the FM portion, compared to its TV portion. The clear outer dial helps you do this, as noted before. I believe that later DuMont continuous tuners had extra gizmos to give finer tuning control within the FM radio band.

It's appealing, in a nerdy engineering way, how the InducTuner puts the actual frequencies in your face. The average person has no idea that FM radio sits in the middle of the analog TV band.

On the downside, this tuner is less user-friendly than modern ones that simply click from one channel to the next. It also lacks a fine tuning control, which soon became standard.

A different type of continuous TV tuner is found in my Scott 800B and Pilot TV-37 sets. They use ganged air variable capacitors like those found in communications radios.

The Pilot's tuner can be seen at the left in this underchassis view. It is desperately cheap, as befits a budget TV. The tuner uses pulleys and strings, common in radios but unheard of in televisions. The tuning dial shaft is second from the top, on the left.

The Pilot does not receive FM radio. A switch toggles between low TV channels and high ones, bypassing the FM frequencies.

Below is the tuner from the mystery television in my Scott 800B combo. The capacitor vanes are continuously variable, but the tuner has mechanical detents to stop you at the stations. A separate control is provided for fine tuning.

After years of inquiry, I still have no idea who made this television. It certainly wasn't Scott, whose sets are well known and do not include this critter. Scott contracted with other manufacturers to supply TVs in some of these combos, so I'm guessing that was the case here. If anyone has a clue about this set, drop me a line. It has been restored but needs some final touches that I don't want to attempt without a schematic.

The Scott TV tuner does not enable FM radio reception, either. Internal switching confines it to television frequencies. Anyhow, who would use a TV tuner for FM when in the same cabinet you have one of the finest postwar FM radios that money could buy?

Okay, so a few TVs could receive FM radio "for free," as it were, but could any radios receive sound from TV channels without extra circuitry?

The answer is Yes, but not many. My Hallicrafters SX-42 boatanchor has continuous FM coverage from 27-110 megahertz. This allows it to receive TV audio up through channel 6, which ends at 88 megahertz. Channel 7 begins at 174 megahertz, so the SX-42 would be deaf to the higher TV stations. All of which is academic, now that analog TV broadcasting has ceased, unless you happen to operate an in-home TV transmitter, as I do.

So, What Kind of Tuners Did the Others Use?

Most television manufacturers used non-continuous tuners of two main types: the robust (and expensive) turret tuner or the less costly wafer type. Here is a turret tuner:

Turret tuners are used in several of my vintage TVs, such as the RCA CT-100, a color set, or the black and white Admiral 24C15.

Here is the wafer tuner used in my RCA CTC-11 color television.

You can see more of that sort of tuner in my RCA CTC-11H article.

A Strange Intruder

Back to the RA-103, the next phase is usually the most time-consuming part of any restoration, replacing capacitors. I always start with the filter capacitors in the power supply, so the set can be powered up as soon as it's safe.

In this case, the filters involved a peculiar box capacitor. It was mounted on the high voltage cage with "lazy man" wires leading all the way under the chassis. I suppose it was the only thing that the tinkerer had on hand, and far too big to fit anywhere underneath.

Here's how the strange intruder was connected. In this photo, the red and blue leads from the box cap have been detached and coiled up. Two can capacitors for the power supply are visible from their bottoms: C241 (above) and C262 (below). I have also included the power supply schematic.

Look how things are wired. The box capacitor is connected between the positive lead of C262 and the negative lead of C241. The other leads for those capacitors are disconnected, removing the original caps from the circuit. Also missing are resistors R316 and R317, formerly connected in parallel with the original caps to equalize voltage across them. In place of those four components is the single 10 mfd capacitor.

You wouldn't normally replace large capacitors with one that's so much smaller. C241 and C262 are connected in series, making 40 mfd. The replacement is still only 1/4 of the design value. Whether this actually worked is academic, since I'll put everything back as the DuMont designers intended.

Restuffing a Can Electrolytic Capacitor

Let's start with capacitor C241. There isn't much room under the chassis for a new electrolytic, so I'll "restuff" it by putting a new capacitor in the old can. Here is the can and the tools we'll use to start.

In this type of can, the negative terminals are on a ring around the outer edge. They are connected to the metal can. Positive terminals are always located toward the middle. Although a can may contain up to four capacitors, this has only one, hence only one positive terminal.

The black cardboard sleeve is an insulator. In some cans like this, the negative terminals slip through slots in the chassis and are twisted and soldered. This makes a physical and electrical connection to "chassis ground." In this power supply, the negative leads of the capacitors are not connected to chassis ground, so it's necessary to insulate the can from the chassis. Don't discard this cardboard sleeve or your TV won't work!

There are different ways to get the insides out of the old can. I often simply cut the can near the base (see Replacing Capacitors for more capacitor info). This time, I decided to uncrimp the can and use the original terminals to preserve a little more authenticity.

Before we do that, let's soften the adhesive with a heat gun and pull off the sleeve. The tarry adhesive cleans up easily with mineral spirits and paper towels.

The next step is to carefully uncrimp the edge of the can so that you can remove the terminal ring and other goodies from inside. You may have to use the point of a knife blade to get it started. Work slowly and bend the edge upward very gradually. The can metal is soft and easily torn.

I usually go around the can about three times before the edge is fully straightened. I use a small pliers in the later stages.

The metal terminal ring pulls out when the edge is straight. Below that is a phenolic disc that holds the positive terminal(s). Underneath that disc is a circle of rubbery insulation and under that is a second phenolic disc.

With patience, you can carefully pull and pry out these layers. You will normally break the thin strips of metal that connect the terminal ring and positive terminal(s) to the capacitor innards. Don't worry about that; you'll be hooking up wires to your new capacitor(s) later.

The next photo shows the metal terminal ring, two phenolic discs, rubbery insulator disc, a little strip of metal that formerly made the capacitor's negative connection to the terminal, and a big wood screw that we'll use to pull out the capacitor innards.

If you save all three discs, they'll fit snugly into the can's collar when you reassemble everything.

Fire up your heat gun and pull away! If the innards don't come out easily, heat the can some more.

In the next photo, I have partially unrolled the layers of foil and paper that form the capacitor body. The powdery junk is dried electrolyte. In a new or usable capacitor, this would be a paste. Some people try to "re-form" crummy old electrolytics by gradually applying increasing voltage, but this photo shows why such efforts are usually doomed. I throw this mess into the garbage where it belongs.

The last steps are easy. I connect the new capacitor leads to the terminals: positive to the center one and negative to an outer one. In this case, I simply ran the new wires through existing slots in the discs. If necessary, in a multi-section electrolytic, you can drill tiny holes for the new wires. You don't need to glue the discs together; they'll be held in by pressure after you re-crimp the can's edge.

I insulate the lead that runs along the capacitor body and secure it with three turns of tape. If you like, you can fill the empty space in the can by winding some plastic or thin cardboard around the new cap. That can't do any harm, but it's not really necessary. The new cap weighs almost nothing and shouldn't be bouncing around inside the can unless you drop the TV off a building.

Slowly fold the edge of the can back down. I use the edge of a heavy screwdriver. In the next photo, I have just begun the process.

You can do a neat job if you're patient. Don't worry if your first effort doesn't look perfect. The important thing is to firmly secure your new assembly inside the can.

The last photo shows the restuffed cap ready to install in the chassis. A couple of drops of glue inside the sleeve will keep it in place. In this TV, the can protrudes halfway through the chassis and is secured by a round metal clamp on the upperside.

Some people put a little label on the capacitor so that future restorers don't tear it out, mistaking it for a bad original. You can also attach a note inside the restored set.

High Voltage Capacitors

The RA-103 uses a couple of paper capacitors rated at 1,000 volts. They are C275 and C276, connected to either side of the horizontal linearity control L219 (see schematic, Page 2). This photo shows their replacements, the two pinkish components near the center.

Needless to say, it's important to use new capacitors of the correct voltage rating. If either one fails, it may affect more than linearity.

Maiden Voyage

The first time you power up an old TV is always exciting, yet nerve wracking. Will it go up in sparks? Will it make a sound? Will it do anything at all?

After replacing the most critical electrolytic capacitors, I gave it a try, using a metered variac to increase power under a watchful eye. I started out with the selector turned to Radio.

The first gratifying sound was a faint click from the relay, indicating that the time delay circuit was functional. Moments later, I heard FM radio and saw the green magic eye light up. Audio quality was good and the tuning was right on the money.

Not bad, but what about the TV? Here's what I saw when I selected Television.

Not the greatest picture, but the TV is still full of rotten old capacitors. Under the circumstances, even a blank raster is a pretty positive sign. I powered down, brewed fresh coffee, and started recapping.

First Picture

That task took me a few hours. Since it was now safe to play the TV, at least for short periods, I tried it out after every few replacements, watching as everything slowly came together. First a picture appeared. It grew brighter and clearer. The already-good audio improved even more. The picture stopped rolling uncontrollably. Eventually, we got a solid horizontal lock. Voltage checks along the way confirmed that the TV was gradually regaining overall health.

With a stable picture and juicy audio, I can move into finishing mode. I'll get out my pattern generator to make adjustments such as linearity and centering. I'll also use an oscilloscope to peek at a couple of waveforms, although judging by this image, they can't be too far off.

Improving the Focus

Looking back at the previous photo, you can see that the picture isn't very crisp. Turning the focus control all the way clockwise didn't make it any better.

That reminded me of an odd sight I had noticed earlier. See how the big doughnut around the picture tube neck is slanted?

That's the focus coil, and it normally wouldn't be that far off-kilter. The photo includes another item of interest—the ion trap. It is the smaller gray metal collar, located downward in this photo, between the focus coil and the black tube base.

When I first powered up this set, there was no hint of a picture, only vague swarming shapes on a very dim screen. It eventually occurred to me to adjust the ion trap. In the position shown above, it produced a nice, bright picture.

Having to turn a control all the way to one end often indicates that something is fishy. The voltage at the focus control was more than adequate, so possibly the cockeyed focus coil was to blame. In the next photo, I have loosened the coil on its mount, to replace its wrecked rubber grommets and try repositioning it. The ion trap has been removed and can be seen below. It is simply a magnet with a springy collar.

Calling on the friendly folks in the Videokarma television forum for advice, the consensus seemed to be that, when the tube was rebuilt, they installed the type of gun that requires an ion trap, although a 12JP4 tube wouldn't normally have one.

The DuMont service manual stated that the focus coil should be located 1/8 inch behind the deflection yoke, perpendicular to the picture tube neck. Leaving the ion trap off for the moment, I positioned the coil that way and tried the TV.

The picture was extremely dim, visible only in a dark room. The next photos were taken with the focus coil in two positions. First, the coil is perpendicular. Note the dark upper crescent in the upper right. The electron beams are missing that part of the CRT.

In the second photo, I have moved the coil off-kilter as it was originally. The image fills the screen, but it's still unacceptably dim and the focus is bad.

This confirmed that the ion trap was necessary. I reinstalled it and tried again, with the focus coil perpendicular as DuMont intended. With the ion trap in its original position, the picture was completely dark. Rotating it 180 degrees on the CRT neck brought back a bright picture, but with awful focus. However, the focus control was still all the way clockwise. When I turned it back toward the center, I found a nice adjustment point near the midrange, as you'd expect with a properly functioning control.

Problem solved!

Why did the ion trap and focus coil work right this way? An ion trap can be adjusted in three ways. I'm guessing that I inadvertently flipped the trap 180 degrees when reinstalling it, reversing the polarity. Some traps have a little arrow pointing to the front of the CRT, although this one doesn't. In any case, now it produces a bright image. The focus coil is aligned normally and the focus control works as it should, too.

Life Without AGC

Every early TV is a little different when it comes to picture adjustments. The RA-103 has no AGC (automatic gain control), as noted earlier, so you may find yourself adjusting brightness and contrast more than with other sets.

The lack of AGC is somewhat unusual. My RCA 630TS has a manual gain control, slightly more convenient than this set. My T-100, made a few years later, has AGC, as do all modern receivers.

I had been using this TV for a while before I noticed this section in the manual.

Normal Operation

With the service selector switch turned to the Television position in which the pilot light is on, turn the audio volume control to the right about half of its range, thus turning on the receiver. Turn the Brightness control almost completely counter-clockwise and turn the Contrast completely counter-clockwise. Approximately ten seconds after the power is turned on, a "click" will be heard indicating that the surge protection relay is energized.

Subsequently, a raster will appear on the picture tube. Adjust the Brightness control for a moderate brightness, below the point at which the raster size increases due to excessive drain on the high voltage power supply. Adjust the focus control for greatest clarity of lines at the center of the raster.

Turn the Brightness control counterclockwise until the raster just becomes invisible. Turn the illuminated tuning dial to a television broadcast station by adjusting the tuning eye indicator until both green bars are of equal size. Turn the Contrast control to the right until the proper contrast is obtained.

Once things are approximately set, you won't need to repeat this entire drill every time you turn on the TV. However, life without AGC means that the receiver can't compensate for signals of different strength.

When the RA-103 is in radio mode, some FM stations will be much louder than others. In a modern set, the AGC reduces the signal strength for stronger stations, so that everything sounds about equal as you move across the dial. For this radio, you must turn the volume down or up as needed. The same is true for television reception, and you may need to tweak picture controls as well as volume.

Early TVs were designed for a world with mostly-weak broadcast signals and they are sometimes too sensitive for their own good. The DVD player that I used to make these screen shots has a strong signal that can overload the TV, leading to extremes such as "blooming," in which the image grows much too bright and loses focus. You may find such differences between a DVD player and a VHS player, or even between one DVD player and another. Output devices are not all identical.

The difference is marked when I switch from a DVD player to my in-home broadcast system, which by its nature produces a weaker signal. The DuMont actually seems happier with this over-the-air input. I can obtain a great picture with less fussing, once the antenna is correctly positioned.

Width Adjustment

Width remained on my to-do list. The picture overscanned the CRT edges and the width adjuster had no effect. If I couldn't reduce it before putting the TV back in its cabinet, the picture mask would cut off a bit more on each side. Some degree of overscan is common on TVs with round picture tubes, and the average person wouldn't notice this, but it annoys me when a control doesn't work at all.

Preliminary checks didn't reveal an obvious cause. The width coil and flyback transformer checked out OK, as did the voltages on the 6BG6 horizontal output tube and the high voltage to the picture tube anode.

Referring to Page 2 of the schematic, T204 is the flyback coil and L220 is the width adjuster coil. Assuming reasonable voltages on the relevant tubes, there aren't too many components that could be at fault. I had previously replaced paper capacitors C275 (.035) and C276 (.05), both rated for 1,000 volts.

I hadn't checked capacitor C287, in part because it's a reliable mica and also because it's rated for 1,500 volts and I don't have that kind of capacitor lying around. A fellow DuMont owner advised me to check it, however, since it's wired to the width adjuster.

These photos shows the flyback coil before and after I replaced C287. The two black insulated leads at upper right go to the width adjuster. Don't use too much heat when working around coils like this. The wires coming out of the flyback are delicate!

Replacing that capacitor brought the width adjuster to life. Now I can bring in the image far enough to see its vertical edges on the CRT. Not ideal, but better than it was.

Falling Off the Horizontal Wagon

Hoping this project was finished, one evening I brought the chassis into our family room and sat down to watch a movie on DVD. Everything looked peachy until about an hour had passed. Then the TV suddenly lost horizontal sync. Back to the workshop!

This horizontal problem differed from others I had seen before on this TV. With careful adjustment of the horizontal hold, you could get a nice upright picture, not a mess of slanted lines. The picture refused to lock into place, however. Left alone for a moment, it would slowly slide sideways in one direction or the other, eventually scrolling faster and faster.

This photo shows another TV with a slightly different problem, but it gives you a general idea what the picture looked like.

What could this mean? Since we could get a reasonable upright image rather than a zillion slanty lines, it meant the horizontal oscillator was able to run at the correct frequency (15,750 cycles per second). But it wandered higher or lower, unable to lock.

A television gets a 15,750 Hertz signal from two sources. One comes from the TV's horizontal oscillator tube, which ideally runs at that speed. The second comes from the horizontal sync pulse embedded in every television broadcast signal.

Locking these two is the job of a tube called the "sync discriminator" in this TV. It compares the two signals, and if the TV's oscillator doesn't match the broadcast's frequency, the discriminator nudges it back into line, so to speak.

I checked the voltages on the 6AL5 discriminator tube (V214 in the schematic) and immediately noticed big trouble. Pins 1, 5, and 7 should have very low negative voltages: -1.5, -1.5, and -1.8, respectively. At those pins I measured positive 287, 287, and 145 volts. No wonder the discriminator didn't work!

Where could the high voltage be sneaking in? In the voltage chart, I noticed that pin 8 of V213, the sync clipper tube, had 300 volts. Between that and the discriminator tube was a tiny 47pf ceramic capacitor, C247.

Replacing that capacitor solved the problem. Without the bad cap leaking high voltage into the discriminator tube, its voltages returned to normal. The horizontal instantly locked and stayed put.

The next photo shows the little tubular culprit, held in front of the new maroon cap that replaced it. While I was in the neighborhood, I also replaced three associated 470K resistors.

Ceramic caps, like micas, are pretty reliable, but they do fail on occasion. High frequencies, as well as high voltages, are hard on components, and some horizontal circuits have little margin for error. In this case, one leaky cap was all it took to make the TV completely unwatchable.

Note, too, that this failure occurred only during an extended "real world" trial. During restoration, I had played the TV for many hours in the aggregate, but usually just long enough to check something. A marginal component may hang on for a while, yet give up the ghost when exposed to fresh voltages and the heat of normal operation. Your restoration isn't truly complete until the set performs reliably for long periods, as it was designed to do.

We're Halfway Home, Kids!

After many hours of labor, the electronic restoration is complete.

I couldn't resist putting the chassis back in the cabinet to try it out. Here's the picture at this stage. My auto-everything camera doesn't make it as sharp as in actual viewing, but you get the idea.

The DuMont is very enjoyable to watch, and the most stable of all my vintage TVs.

Cabinet Restoration

The television performed beautifully now, but its cabinet still looked as rough as the day when I brought it home.

The television was stored in a damp garage for years, and problems were immediately evident.

Surveying the Damage

The top had warped badly, eventually pulling itself loose from the top. The veneer is still attached to the top panel, but I'm not sure whether the panel can be straightened. Perhaps I'll need to replace it. The veneer is scratched but salvageable.

The sides have warped and delaminated. The next photo shows how the inner veneer, as well as the outer, came loose from the solid panel. At the top of the photo you can see how far the side panel warped away from the cabinet structure.

The front finish is cracked and weathered, with a number of nicks and scuffs. The lacquer is completely scraped off in a few spots near the knobs, although the decals themselves have mostly been spared. Reproduction decals for this set are not commercially available. Fearing that I might have to make my own repros, I took a series of close-up photos to serve as a starting point for the decal art.

The grille cloth, fortunately, is undamaged, with no stains or rips.

The RA-103 has six wooden knobs: five small ones and a big one for the tuner. I started out with only one, but while restoring the electronics, I was able to get a set of them from a fellow collector (thanks, Chuck!). The knobs were almost devoid of finish, but that's easy to remedy. The second photo shows some of the knobs and decals, as well as damage in the control area.

A small piece of molding near the upper left corner fell out years ago and was lost. The corresponding piece on the right side was loose enough to pull out when I examined it. The first photo shows the empty space on the left, where a replacement will be needed. The second photo shows the loose molding piece and the space on the right where it came from. Two new pieces are actually needed: a mirror of the right piece that came out and a shorter "return" piece to go around the corner of the cabinet.

Time to Call in a Professional

I have refinished a number of wooden cabinets, but I don't have a woodshop or the skills needed to deal with these structural problems. I brought the cabinet to a Seattle refinisher, Michael Mueller.

After Michael examined the cabinet, we agreed on a price and he promised to do what he could for that amount. This would include fixing the structural problems and improving the finish. He wouldn't know until later whether the old finish could be successfully reamalgamated or might require more drastic treatment

Before delivering the cabinet, I removed the knobs, eye tube bracket, bezels for the dial and magic eye, safety glass, and the CRT mask that holds the glass in the cabinet.

The grille cloth is glued to a board which comes out after removing ten screws. Here, the board is leaning against the detached top panel. The cloth is a nice match for the mahogany cabinet color. It's somewhat faded where it was exposed to light, but I have seen many old cloths with worse fading.

A couple of days after I delivered the cabinet, Michael sent me some photos of the work in progress. Here is our friend, the top panel, undergoing some serious rehab.

That looked a little scary to me, but the straightening was successful, as you'll see later. The next photo shows a side panel in process. The old veneer was removed from both sides and reglued.

While the glue was curing, Michael began to try reamalgamating the front panel finish. Reamalgamation softens an old lacquer finish so that it can be respread. This eliminates surface defects while preserving the original color and luster. The end result looks like the original finish because it is the original finish.

I had read about reamalgamation and tried it a few times in the past, failing spectacularly. In my fumbling hands, the finish ended up one of two ways: mostly stripped, or a mess of uneven streaks. Michael remarked that the technique demands practice, as does French polishing.

In the next photo, while the glue on one side is curing, Michael has started reamalgamating the finish on the top and right side. It's only partly done, but you can see that the cracking and uneven coloration is going away.

The project is nearing completion. The absent molding has been fabricated and installed. The warping and delamination has been repaired. The top panel is back in place, and the finish is mostly restored. After more detail work, I'll be able to bring the cabinet home.

Welcome Home, Cabinet

After delivering the cabinet, Michael mentioned that it wouldn't be a bad idea to apply a coat of finish to the interior. This would help prevent one side of a panel from drying out (or absorbing moisture) more than the opposite side.

In the next photo, I have applied walnut stain to the interior. It's time to reinstall the cabinet parts I had removed before refinishing.

The speaker and cabinet hardware are back in place. As soon as my son shows up to help, we'll slide the chassis into the cabinet.

With everything in place, I can take some final photos. Here's a rear view, much as it looked when the TV left the factory.

Remember that corner with the missing molding? Quite an improvement!

It's hard to believe this is a 62-year old television.

Final Thoughts

This was one of my most satisfying restorations. It's a treat to work on a set of this caliber. My RA-103 is definitely a keeper—a TV that I'll use and enjoy for a long time.

How long did this take? I bought this TV on October 5, 2009 and reassembled it today, February 22, 2010. Four and a half months is not unusual for me, since I work on these projects sporadically, with frequent time-outs to putter with other radios and TVs or write articles for this website.

If you're looking for a vintage television project, I'd recommend any 1940s DuMont. They're among the best-designed TVs of the time and their build quality is second to none. There's nothing particularly wrong with post-1950 DuMonts, but I think the company lost focus when Allen DuMont made his ill-starred attempt to create a national broadcasting network. DuMonts of the 1950s were respectable, but they no longer stood head and shoulders over the competition as in the 1940s.