Hallicrafters Web Site--Restoration Projects

Bottom of Chassis After Initial Cleaning, Before Rebuild

As received, the receiver appeared to have all the parts. The bandspread bezel and "glass" were loose, the two tube shields were loose inside the case, and the overall case and chassis were very dirty.

Upon inspection of the under chassis, I found a kludge capacitor in parallel with the first section of the twistlock electrolytic, and there were quite a few spider webs. There were also lots of old paper capacitors, probably leaky, old style body/end/dot resistors, and a piano tuning felt wedge jammed in the BFO coil.
Neither of the tuning dials worked due to the age of the dial cord, which though intact, was stiff and brittle. Since there is no removable panel, it appears the complete tuning assembly must be removed to restring the dial.

No attempt was made to apply power. It is a certainty that the twistlock can is bad, so this will have to be rebuilt or replaced. Not sure yet how to approach cleaning up this beast – maybe remove the tubes and tuning assembly, and start from there? The tuning capacitors, shafts and dial plates will need cleaning, and it will need new dial cords.

To start, the kludged add-on capacitor was removed, and most of the spider webs cleaned out. A schematic was available on the BAMA website. This schematic was nearly the same as my unit except for the tube complement. On my receiver, a 6K7 and a 6Q7 are used instead of the 6SK7 and 6SQ7 shown in the 19R. I have since discovered there were two runs of S-19R, and mine is the earlier run. I figured a logical place to start would be with the dial cords. Since the whole assembly was quite dirty, and had obviously had several "quickie" fixes applied, (like tape around the shaft) it looked easiest to remove the whole works. This is accomplished by unsoldering two wires from the tuning capacitor on the bottom of the chassis, one ground braid on the top side, then removing the main dial, knobs and the mounting nuts holding the tuning assembly. Two of the mounting studs have ground lugs attached. These have two nuts, one holding the tuning assembly, and one holding the lug. (Remember this when you re-assemble.) Once this is all loose, you need to loosen the setscrew on the bandspread dial and push it back slightly on the shaft, so it will clear the top of the case. The entire tuning mechanism can then be removed and serviced.

Once out of the chassis, I made a sketch of the dial cord stringing, such as it was. The old cord was removed, and the tuning assembly frame was stripped down to bare metal by removing the shafts and tuning capacitor. This frame was scrubbed with kitchen cleanser, rinsed and given a light coat of deoxit. The shafts were cleaned and polished, then trued up slightly. The tuning capacitor was sprayed with WD-40, scrubbed carefully with a brush, then rinsed and blown out with an air hose. A bit of deoxit was sprayed on the rotor wipers, and a drop of silicone grease was applied to the two bearings. The tuning shafts were then re-installed in the metal bracket, using a small amount of silicone grease in the bearings. The shorter, single diameter shaft goes in the top bracket, and the other two (which are identical) go in the bottom. The tuning capacitor was remounted and checked to insure its shaft was parallel with all the other shafts.

Next job was the restringing. For the cord I used my standard 45lb Dacron fishing line. I started with the main tuning, since it is strung behind the bandspread cord. Initially, I strung it exactly like the original, with both ends of the cord tied to the spring. After trying it out this way, I felt there was not quite enough tension, because one end wraps a full turn around the tuning capacitor pulley, while the other end only goes about a quarter turn around the pulley. The spring cannot take up the slack on the short end of the cord because of the friction of the long end around the pulley. So, I undid the thing, and re-strung it starting with a tied loop attached to the spring lug on the pulley, wrapping around the pulley one and a half turns, then around the tuning shaft two turns, then back to the tuning condenser pulley, through the hole in the pulley then tied to the spring with as little slack as possible. After gluing the knot, the other end of the spring was attached to the spring lug, and the whole thing worked much better.

The bandspread cord was done in a similar manner. I was a little surprised that the cord in the original stringing did not do a complete wrap around the pulley on the bandspread dial. I tried stringing it with a full wrap but it seemed cumbersome and I wasn’t sure I could move the dial back far enough to get the assembly back in the cabinet. So, I stuck with the original stringing, except for using a tied loop to the spring lug on the long end. This appeared to be the way Hallicrafters did it originally, and if it worked for them it should work for me. I figured the worst that could happen would be that the bandspread dial could slip, but it could be easily corrected by opening the top of the case and turning it slightly. On this receiver, the bandspread dial is not marked in frequency, but with just a 0-100 logging scale.

The tuning assembly was now ready to reinstall in the main chassis, but first I needed to do some cleanup. I removed the remaining knobs, and the old twistlock can. Due to the fading of the old wire colors, and my own partial color blindness, (I can’t tell a 100 ohm resistor from a 1meg without an ohmmeter.) I marked the removed wires with stick-on numbers, and made notes as to which section of the capacitor was connected to which number wire. I started the cleaning with a mild dishwashing soap and water, scrubbing the chassis with a brush and being careful not to get any water near the speaker, transformer or IF cans. There was a lot of "gook" around the area where the old twistlock had been mounted; indicating it had been leaking for some time. This stuff required a little more aggressive cleaner to remove, but Ajax seemed to work. The corroded spots on the top of the chassis and on one side of the chassis now became very visible. The front panel, however, looked reasonably good. I decided to address the cosmetic issues as needed during the re-assembly phase. After drying the chassis, I had a close look at the speaker. The cone looked quite dry and brittle, and I made the mistake of touching it, only to have my finger go right through. Oh well, live and learn. Time to remove the speaker and have it reconed. The speaker is held to the panel by four ornamental screws, rubber grommets and nuts. Six wires pass from the speaker and output transformer through the chassis. These were marked as described above, and then the entire speaker assembly was removed. All rubber grommets were either turned to goo or brittle. Fortunately, these are still readily available. I saved samples of the sizes I needed and tossed the rest.

Next I installed the new twistlock, and decided to replace all the old paper and wax capacitors. Most of these were no problem, with the exception of C4 and C5. One end of C4 connects to a terminal on the front deck of the bandswitch, which was virtually inaccessible. The wiring is tight here and it did not appear feasible to remove or rotate the bandswitch. I elected to cut the bandswitch lead close to the capacitor, strip off a bit of the sleeving, slip on a piece of heatshrink tubing, then splice and solder the lead from the new capacitor to the stub. After this, the heatshrink was slipped over the splice, and a bit of heat applied. One end of C5 goes to a very crowded ground lug, the same one that carries a ground braid through the chassis to the tuning capacitor. Even the original assemblers couldn’t get all the leads through the hole in this lug, but the real problem is the other end of C5, which goes to the bottom terminal of an antenna coil. Again, this connection is nearly inaccessible. Though this capacitor could be relocated, it’s placement made sense RF wise, and I didn’t want to second-guess the original designers. After some study I elected to temporarily remove one wire from the adjacent coil. This allows access to the lower coil terminal, providing you have a small pencil-tip iron. Capacitor C22, which goes from one side of the AC line to the chassis, was replaced with a capacitor certified by UL/CSA for this type of service. A one lug terminal strip was mounted using one of the transformer mounting screws so that a fuse could be added in series with the hot lead from the line cord. The old line cord was replaced with a somewhat more rugged polarized cord. I checked the values of the old dog bone resistors, and amazingly, didn’t find any that were off by more than 20%, so I left them alone.

While the speaker was out for reconing, I did some touch up painting on the chassis and side panels using flat black Krylon spray. The front panel looked pretty good after a light rubdown with WD 40. There were some white paint specs here and there, which I touched up by spraying some Krylon on a Q-tip and dabbing it on. (Someone once said since these spots are so common on old radios, people must have used them in place of a tarp when painting.) I cut out a new bandspread "glass" from thin plastic, and scribed a line down the center. I tried inking this line, but decided I liked the scribe better. I then installed new grommets in the tuning assembly holes, loosened the setscrew on the bandspread dial and moved it back a bit, then remounted the tuning assembly and reconnected the wires. Be sure to use plenty of heat on the shield braid connection, and don’t forget the solder lugs that go on two of the mounting studs. After the tuning assembly is mounted, move the bandspread dial forward and tighten the setscrew. NOTE: Failure to move the bandspread dial back during tuning assembly removal and re-installation may result in cracking the plastic dial. Also, beware of the screw head holding the bandspread bezel, it can scratch the dial.

The headphone jack is intended for high impedance headphones, being fed by C15 from the output tube plate. If you want to use low impedance phones, you will have to install a new headphone jack with an isolated SPDT switch, and do some rewiring. This is best done when you are reinstalling the speaker, as some connections between the output transformer and the speaker have to be rearranged.

Speaker Assembly Before Reconing.

When the speaker arrived, I installed new grommets, mounted it back in the case, and reconnected all the wires. The knobs were cleaned up and reinstalled. A bit of white crayon restored the white dots on the audio gain and BFO pitch knobs. I used Naval Jelly and fine steel wool on the main dial, with results that were fair. (Rock Sea is supposed to be working on a new dial for the S-19.) The plastic pointer and spacer washers for the main dial were reinstalled. The tubes were unwrapped and tested. Only the rectifier, (an 80) was reinstalled for now.

The receiver was plugged into a variac and the B+ voltage monitored. The variac was turned up to the point where the rectifier became hot enough to do its stuff. Everything was carefully watched for signs of spitzensparken and fusenblowen. So far, so good. With 70 volts AC from the variac, the B+ read about 130V. With 120V input, the B+ was reading almost 300V. The rest of the tubes were then plugged in, at which point I noted I had a metal 6K8 instead of a glass one. The converter tube shield was therefore not required, which is why it was loose when I first inspected the receiver. The shield IS required for the 6K7 IF amplifier.

I was now ready for the first test. I turned the radio on with the variac set to 100V. After the set warmed up with still no smoke, I turned up the volume and touched a pencil to the top of the audio gain control. Some hum was heard, indicating the amplifier and speaker were working. I then switched to Band 1, jumpered A2 and GND on the antenna terminal strip, and connected about a 10ft. piece of wire to A1. On tuning, several AM stations were heard and the sound quality was pretty good. Turning on the BFO switch, I was able to adjust for zero beat with no problem. Under these conditions, with the variac turned up to 120V, the B+ from the rectifier read 255V. No specific voltage is called out on the schematics I had, but the output of the optional vibrator supply was shown as 260V, so I figured I was close. I played the radio all day at 120V and all seemed to be in order. The transformer laminations read 58 degrees C after 7 hours. A bit on the warm side, but not what I would consider hot. It’s a personal decision whether to run this receiver through a voltage reducer, setting the input voltage at 110V, which was probably its design center.

While "burning in" the radio, I cleaned, straightened and repainted the top and bottom cover panels. The "feet" on the bottom were replaced after the paint dried. Apparently, there was some kind of paper legend once attached to the bottom, which may have identified the trimmers, but this was missing on my particular set. It may be that the purpose of this was more to cover the access holes than inform, I really don’t know. The trimmer information is included in the schematic and alignment instructions anyway.

Satisfied that nothing was going to burn up, I proceeded to align the set according to the instructions. This set uses "padder" capacitors to adjust the low end tracking on bands one and two, rather than adjustable inductors as used on later radios. High end tracking is set by trimmers as usual. There are no padders for bands three and four, so low end tracking is set by design, and hopefully will be pretty close. During alignment, keep the signal generator output as low as possible, and set the bandspread at MINIMUM capacitance. Mine tuned right up with little difficulty, although the tracking on band four was not perfect. I adjusted trimmer CH for best compromise around the 10 meter area. The final alignment should be done with the bottom cover in place, as it influences the inductors.

After alignment, the performance of the radio was surprising, considering this model was made about 1938, and had all but one of the original tubes. Buttoning it up, I took it upstairs and hooked it to the attic dipole. The performance on bands 1 and 2 was just about the same as my SX-117, though not with as much selectivity. It was possible to copy sideband on 75 meters with good intelligibility. Tuning sideband with just a BFO is a bit tricky, but not impossible. Performance on band 3 was definitely poorer than the 117, but still not too shabby. Performance on band 4 was slightly better than poor. It could be that a new 6K8 would help, but since this receiver is to be used mainly for SWL, I wasn’t too concerned. To me, it was amazing that it did as well as it did for a 1938 radio with no RF stage. The finished product didn’t look like it just came out of the box, but it sure looked a lot better than it did when I started, and was quite presentable. The weak spot is the main dial. If a replacement scale becomes available, I’ll probably buy it.

It is my hope that these notes assist you in your restoration of a similar receiver. There is a lot of history in these early receivers. Who knows what sounds came out of its speaker? The celebrations on VE day? Truman’s speech when he fired General McArthur? Cold war propaganda from Radio Moscow? I know I’ll think about this every time I play this old Hallicrafters.