Museum Hours:

Saturday 10-6

Sunday 12-5

Please Become a Member of the Museum


Early Color Television

CBS Color Converter/Adapter Restoration


The Adapter

The connecting harness is missing. We have a photo we can use to build a new one. The harness has a 12 conductor shielded cable, about 6 more conductors unshielded, and a two conductor line cord, plus three thin pieces of wire in 1/8" diameter tubing. All of these are wrapped in plastic tape.

We will use the shielding from a piece of RG8 coaxial cable to make a new 12 conductor shielded cable.

At the end of the harness are three octal plug/socket assemblies which are plugged into tube sockets in the TV set, and the tubes are then plugged into the top of the assemblies. We will break some octal tubes to salvage the bases, wire the bases to the sockets and to the harness, put the sockets in the top of the octal base, then hold the sockets in place using set screws. A metal cap with an octal plug will be used for the assembly that plugs into the horizontal output tube socket.


The harness, showing the plug/socket assemblies, the HV connector, and various other connections

Cap for octal plug


The replica plug/socket assemby

The surviving portion of our harness


Start of the new harness. The shielded cable contains 10 conductors for the vertical circuits. The black wire taped to the shielding is one of 3 "low capacity" cables, which have an outside diameter of 1/8" and contain a very thin (32 ga) wire.

Here is the completed harness. From lift to right are: plug to go in 6BG6 horizontal output tube socket, low capacity wire to go on pin 6 of the 6SN7 horizontal oscillator socket, plug/socket assembly to go under 6W4 damper tube, plug/socket assembly to go under 6K6 vertical output tube, plug/socket assembly to go under 6SN7 vertical oscillator tube, and wires to connect to the yoke.

Part of the harness, showing the plug/socket assemblies for the vertical output and vertical oscillator tubes

The circuitry has been restored, with new capacitors installed in the old paper capacitor shells, and new electrolytic capacitors installed inside the old shells.

The flyback transformer and HV rectifier assemby are missing. We don't have photos of what the original assembly looked like, so we will have to improvise.

The flyback is also a problem, since it is designed to work at 29,160 Hz. The issue is apparently the retrace time, which is determined by the resonant frequency of the flyback. The resonant frequency was raised by removing much of the secondary winding of the flyback, and using a voltage doubler to get the 12 KV required for the CRT. We will install a flyback from a 1949 RCA set, and experiment with removing windings from the flyback until we get about 6 KV HV. We will then connect the adapter to the 9T-246 set, feed the set field sequential video and observe the picture, looking for evidence of horizontal retrace problems. If this works, we will add a second filament winding for the second HV rectifier, and build the voltage doubler. We will also have to experiment with the secondary winding of the flyback. The one originally used in the adapter has 4 taps, which are used in circuitry that sets the width of the picture in both color and monochrome operation (In color, the picture is reduced to about 7" diagonal on the ten inch screen. The replacement flyback has only 2 taps.

Here is the RCA flyback installed in the adapter

The voltage doubler, with the RCA flyback

I finally got the adapter to work fairly well using this flyback. There are several errors in the CBS schematic that I found out by trial and error. In the color mode, the flyback produces only 6 KV of high voltage using a single rectifier. I have now installed a voltage doubler, which produces about 9.5 KV. Here are photos off the screen:

You can see retrace problems in the picture on the left (white band) due to the flyback having too low a resonant frequency. Because we only have 9.5 KV from the doubler, we can't remove any turns from the flyback to try to solve the retrace problem. We will try the following to increase the voltage: 1) replace the coupling resistors used in the doubler with a high voltage diode. 2) replace the 1X2 rectifiers with diodes (eliminates the load of the two filaments). 3) Use a solid state tripler unit fed off the plate of the horizontal output tube 4) Try a similar flyback that is designed to be used with a voltage doubler.

If we can get sufficient high voltage to remove turns from the secondary, we will do so to see if the retrace problem is solved. If not, we will have to be satisfied with the results we have. The picture is probably acceptable as it is.

We replaced the flyback with a FLY-4, which is a Thordarson replacement for one used in 1949 RCA sets with the 16AP4 CRT. This flyback is designed to operate with a voltage doubler. The results were a dramatic improvement over the 9-T-246 flyback. The only problem was how to reduce the width of the picture without reducing the high voltage. We accomplished this by adding an inductor and resistor in series with the yoke. In the photos below you can see that the retrace artifacts are almost invisible. The only problem shows up in the crosshatch pattern, which shows that the far left side of the screen is very nonlinear. I don't think this will be very visible in most scenes.



The top of the completed adapter, with the  FLY-4 flyback, HV doubler rectifier, and vertical linearity control (far right). In front of the flyback is the inductor and resistor used to reduce the width in the color mode. At the upper left is the 11 pin socket for the converter, which houses the color wheel and control circuitry.


The Converter

Schematic diagram.

The converter, before any restortation

The converter, mounted on its sliding track, in front of the RCA 9-T-246 receiver (out of the cabinet). The adapter can be seen on the left. All we need now is the color wheel to complete the restoration

The mounting track is missing. We have made a replica using a slide assemble from a cabinet drawer.



The motor needs to be taken out, lubricated, and made to work again. After lubricating it, the motor works well.

The color wheel is missing. A replica will have to be made from the engineering drawings and photos we have. John Folsom has built a wheel, and suggested some color filter sheets that should come close to matching the ones CBS used. I will have two 1/16 inch thick 15 1/2 inch diameter clear plastic disks made by a sign company, then cut out the filter material and sandwich it in between the two disks. Then I'll seal the edge with tape.

The new wheel is complete (see below), and when I installed it, the converter worked and displayed good color pictures. The remaining problem is that the saturation of the green and blue are not good, and I'll have to experiment with different filter materials.

A CBS drawing of the wheel


The color wheel out of the other surviving converter


Filter segments, cut out of thin stock


The filter segments, laid out as they will be on the wheel


The completed wheel, with metal strips installed for the synchronizing magneto. The thin filter material is sandwiched between two 1/32 inch thick clear plastic disks


The circuitry has been restored, with paper capacitors and electrolytics stuffed with new ones. The control circuitry appears to work properly.

Photos off the screen of the converter, before installation of the color wheel. The actual picture looks quite a bit better

These are photos of the screen with the color wheel installed. Notice how weak the green is. The blue is also not fully saturated. I need to experiment with different filter materials. Also, notice the poor resolution in the lettering. The CBS system was limited to about 240 lines of horizontal resolution.

The magnifying lens is badly scratched. I used a kit to remove scratches. It consists of various grades of emery cloth and a bottle of liquid abrasive. I used the 1500 grit to remove the deep gouges, then went over all the scratches with 4000 until they all disappeared. Then I used the liquid abrasive to do the final polishing. The whole process took about 1/2 hour and the lens looks like new.