Atwater Kent Model 84 Restoration

The Atwater Kent 84 (Early Version) from about 1932 is a 6-tube AC non-AVC Superhet with separate oscillator and mixer tubes. It receives only the broadcast band.

The schematic for the model 84 is on-line at Nostalgia Air. Another document is also needed for parts placement and for restuffing the three metal cased bypass capacitors. That document shows how the replacement capacitors are to be connected. http://www.nostalgiaair.org/PagesByModel/493/M0001493.pdf

The radio had seen minimal servicing in the past - most of the original parts were still in place. I decided to try to reverse all previous repairs to the extent possible. However this vintage of AK receivers is difficult to restore and maintain originality due to the unique cast end resistors and metal can bypass capacitors used.

My antique radio restoration logs

Condition As Found

This radio was purchased at the 2011 Charlotte AWA Radio Conference. The cabinet was in good original condition. The knobs and grille cloth were original and in good condition. The grille insert was obviously missing (the pattern can be seen on the grille cloth). However, there was no sign on the grille surround that the insert was ever present! But it must have been removed (likely were broken) and the attachment points smoothed out and somehow refinished. The repair is invisible. I always avoid knowingly purchasing a radio that has been restored by a collector, as many collectors take shortcuts such as removing the original capacitors and filters. In this case, the original filter capacitors and all tube shields were still in place. The only evidence of repair was some tape on the grid cap leads (no new wiring obvious). The AC plug was broken, (possibly preventing someone from "testing" it before sales). The chassis had a few rust pock marks, but still looked good.

Previous Repairs

All the tubes were ST or shouldered types. The originals would likely have all been globe types (the sockets were labeled 280, 224 etc).
Several grid cap leads had been insulated using tape. The lead from the volume control to IF amplifier grid cap had been spliced to a new piece of wire, and the existing portion insulated with electrical tape. There was evidence that the shield on the 2nd IF amplifier transformer had been removed and replaced in order to insulate the grid cap lead. Tape was also used to insulate fraying in the speaker cable near the plug.
The power cord was original, but had been shortened.
New filter capacitors had been installed under the chassis. Thankfully, the originals had been left in place, and terminal strips attached to the original lugs. The new capacitors were then installed on the terminal strips. This technique results in a totally reversible repair! I only wish that more collectors would use this method.

Survey

My usual restoration procedure is to first make a complete survey of the condition of all components. The survey results guide my restoration strategy. I never apply power to a radio before restoration, even through a "dim bulb tester" or variac "to see if it works". If major and unique components are defective or missing and cannot be restored or replaced, I may elect to sell the radio rather than restore it. I always assume that all paper and electrolytic capacitors are leaky and thus should be replaced (I always "restuff" the original containers if possible). Any mica capacitors are assumed OK until testing proves otherwise.

The AC power switch was bad (it measured a high resistance) - dirty and/or oxidized contacts were likely. It responded to a shot of Big Bath cleaner followed by repeated cycling and then worked correctly.
The speaker field was OK.
The speaker cone was perfect.
The output transformer was OK.
The power transformer checked out OK. My test procedure is to first remove all tubes and apply about 10 volts through a fused variac and analog wattmeter. I then check if the high voltage winding is balanced across the center tap. An unbalance of more than about 1 volt indicates possible shorted turns. I then raise the voltage gradually up to the specified line voltage (110 volts) while monitoring the wattmeter. At full voltage I check the voltage of all windings, and again check the high voltage balance. Unloaded, a good transformer will draw less than 10 watts and each half of the high voltage should be equal within a few volts.
All RF coils and the IF transformers were OK. The coils used in this radio are quite unique in construction, and difficult if not impossible to repair. If any of them had been defective, I would have abandoned the restoration and sold the set for parts.
All the RF chokes (there are several used) were OK.
All the tubes tested good (but all should have been globe types)
Some wiring in the radio was rubber covered and the insulation was falling off. All of this wiring would have to be replaced. The speaker cable had tape on wires near the plug, indicating fraying. There was a break in the pilot lamp socket leads. The insulation sleeving on the filament leads from the power transformer was brittle and falling apart, with numerous breaks.
One of the wire wound bleeder resistors measured a high and changing resistance. These resistors have a unique appearance and cannot easily be repaired or reproduced.
Several cast end fixed resistors were out of tolerance.
The tuning dial drive rubber was functional, but did slip somewhat. Replacements are available from Adams Manufacturing, or in the plumbing parts area of large hardware stores! Others have used a slice of rubber hose from an auto parts store.
The volume control measured close to the correct resistance, but was likely going to be noisy in operation. When measured using my old Knight VTVM while rotating the shaft, there was some jumping of the pointer.
The original power cord had been cut short, and the AC plug was broken (not sure if it was original).
All adjustment trimmers were still sealed with wax as originally shipped from the factory.

Repairs

Before starting repairs I made BEFORE photos of the chassis bottom. I use these photos to ensure that replacement parts and wiring are placed as close as possible to their original positions. Some radios are subject to problems (such as oscillation) if wiring is re-routed or lead dress is not the same as the original. The Riders Manual pages for this radio fortunately had parts placement diagrams. And unlike original AK documentation, parts values had been written in on the schematic. The original documentation only contained AK part numbers - no values. AK resistor color codes are unique and very non-standard. Servicing this radio is very difficult. The resistors are either large wire wound units or cast end dogbone type held in insulated clamps in groups of 3-6. All the bypass capacitors were in 3 metal cans which were riveted to the chassis! Each can held 3 or 4 capacitors.

All tubes and tube shields were removed. All non-original parts were removed. The tuning capacitor and dial assembly was removed for cleaning.

The top and sides of the chassis was cleaned with GoJo hand cleaner and 00 steel wool. Since this process may leave small steel wool fragments that can cause problems later, I follow up with a thorough vacuuming and go over everything with a small magnet and masking tape to pick up any stray fragments. I continue to use steel wool as I have yet to find a substitute that does as good a job removing the "gunk".

The tuning capacitor was cleaned using a Brillo pad (external surfaces), soap, water, and toothbrushes. It was then dried using a heat gun. I did not attempt to disassemble the unit in order not to disturb the adjustments and alignment. The rubber on the tuning shaft was replaced using a slice of 5/16" (inside diameter) rubber hose purchased at an auto supply store. Although slightly smaller than the original, it worked OK and did not slip. The hose was a very tight fit over the knurled shaft. This caused the outside diameter to increase enough to engage the tuning capacitor drive mechanism. In other restorations I have used the correct part available from Adams Manufacturing.

Insulation breaks pilot lamp wiring were repaired using shrink tubing. It was found that the speaker cable was long enough to allow the plug to be removed, the frayed sections cut off and stripped, and the plug reinstalled. The original shortened power cord was retained and an old style AC plug was installed. The filament leads from the power transformer were re-insulated using spaghetti tubing. The leads were unsoldered from the tube sockets, the existing insulation removed by crushing, new insulation slipped on, and the leads reattached.

Volume Control

The volume control potentiometer showed signs of noisy operation. The control was large and unique in appearance, and I wished to keep it original, if possible. The resistance element is simply thin cardboard with a thin carbon layer. This type of control CANNOT be cleaned - the element is destroyed if contact cleaner is used. The wiper does not directly contact the element. Rather, there is a metal ring that is pressed against the strip as the control is rotated. I brushed off the element using a small clean acid brush. Then I cleaned the under side of the contact strip using lacquer thinner on a bent pipe cleaner, without contacting the resistance element. This improved the operation somewhat. I would have to wait until the radio was assembled and tested to see if operation was satisfactory. A thin layer of grease on the metal strip greatly improved smoothness of operation.

Capacitors

The original power supply filter capacitors were removed and restuffed. The original capacitors were both 8mfd at 475 volts. I used 10mfd at 450 volts to restuff them. I would recommend 500 or 600 volt capacitors if available. 10mfd/500 or 600 volts and 8mfd/600 volts are (at this writing) available from http://www.justradios.com/capacitors.html in Canada. One capacitor had a cardboard insulating sleeve. This sleeve was glued to the capacitor and could not be slipped off. It was split down one side and carefully removed (the material is VERY fragile and falls apart easily). After re-stuffing, the sleeve was reattached using service cement. The split is visible, but was oriented so as not to be that obvious.

The capacitor cans were chucked in my Unimat lathe and their cases scored about 1" from the bottom. The cuts were then completed using a hobby razor saw and the edges cleaned up using an Exacto knife. The capacitor cans are nickel plated copper, and are hard - not the usual aluminum. The original contents were then removed and the capacitor case cleaned inside and out. The original positive element along with the stud is one piece of aluminum and is difficult to remove. Diagonal cutters were used to remove as much of the positive element as possible. The positive lead of the replacement capacitor was extended using buss wire, insulated using spaghetti tubing, and routed through a small hole drilled along side the existing stud. The lead was then secured under the original positive terminal lug (which is held to the stud using a nut and lock washer).

The negative lead of the new capacitor was extended, insulated, and then routed though a small hole drilled into the capacitor base near the outside and secured under the square nut securing the capacitor in its clamp, thus making contact with the chassis. The two halves of the case were then joined together using a plumbing 3/4" PVC pipe coupling wrapped with masking tape, and epoxied to each case half. The masking tape is needed since the couplings are slightly too small in diameter. It also should make it possible to disassemble the capacitor in the future should that be necessary.

All of the bypass and coupling capacitors were in three metal cans each with three or four capacitors inside. These cans were riveted to the chassis, and their bottom covers were attached with solder. They don't make it easy! Obviously, this radio was not designed with service in mind. First, all the leads had to be identified and careful notes made as to which wire went to which terminal. Then the rivets had to be drilled out and the capacitor removed from the chassis. The bottom cover was then removed using a large 80 watt soldering iron. I start on one end and remove as much solder as I can from about 1" of the cover by simply liquefying it and then slinging the melted solder into an old waste can. Then while heating the end and the solder is liquid, I pry up one side with a small screwdriver. Once started, the remainder of the solder is heated and the cover removed. I do NOT reinstall the cover after re-stuffing with modern capacitors! The contents can then be removed and the case cleaned of wax by heating with the large soldering iron. The component attachment points were then cleaned of solder and any wiring remnants. The new capacitors were then mounted to the original attachment points and re-soldered. For replacement capacitor connections see http://www.nostalgiaair.org/PagesByModel/493/M0001493.pdf. I used 630 volt axial film capacitors for restuffing except in cases where a smaller part is needed, only because that's what I stock. Radial film capacitors are much smaller than axial, and also cheaper (and may be all that's available in the future). In this application they are not visible!

For Bypass Capacitor #3, I used four 0.1mfd 630 volt axial film caps. The ends of all four connect to the four lugs, and the other end of all four were soldered to the case.

For Bypass Capacitor #2:
For C5 I used 0.047mfd/630 volts (axial film)
For C6 I used a 0.0015mfd/630 axial film capacitor and went through my parts bin and measured the values, picking the one closest to 0.00145 as measured on my HP 973 DVM which measures capacitance. You can likely get away with any 0.0015mfd film or mica capacitor, or use 1% mica capacitors in parallel to get to the actual value (0.001 + 400pf plus 50pf etc.)
For C7 I used a 0.47mfd 100 volt radial film capacitor (available space is limited). This is a low voltage capacitor and you may have to shop around to find a 0.33mfd or 0.47mfd 50 or 100 volt capacitor small enough to fit inside the case.

For Bypass Capacitor #1:
For C1 and C3, I used 0.1mfd/630 volt axial film capacitors.
For C2, 0.033mfd/630 volts
For C4, 0.22mfd/400 volts (630 if it will fit!).

Any common ground leads were soldered to the inside of the cans. The cans were then reattached to the chassis using 8-32 round head slotted screws, nuts and lock washers salvaged from a scrapped AK55 parts chassis.

Resistors

One of the unique cast-end resistors, the Second Detector Screen Resistor (100K) was 30% high. The rest were either in tolerance (+/-15%) or in the low 20's. I decided to leave these in place until the radio could be tested. They would likely work OK. The screen resistor would have to be replaced.

I first took a photo of the original resistor and measurements of the diameter and length (critical, since the resistor is retained by a clamp).
The ceramic tube was then crushed as close as possible to the cast metal ends.
The remnants of the tube were then removed from the cast ends. This is a difficult process. I used various tools such as nail sets and ice picks to shatter the ceramic and then pick out small pieces one at a time.
Once the cast metal ends were clean, I center drilled a small hole in each end to clear the leads of the replacement resistor.
The original diameter of the ceramic tube was about 1/4", so I decided to use an ordinary piece of 1/4" wooden dowel to replace it. The dowel was first center drilled on a lathe to accept the 1/2 watt carbon composition replacement resistor. While still chucked in the lathe, the dowel was sanded and painted flat white using hobby paint, then cut to length.
The resistor was placed inside the wooden dowel, with its leads protruding through the holes drilled in the cast metal ends. The wooden dowel was then attached to the metal ends using epoxy. The resistor leads were soldered to the metal ends at the point where original wiring leads were attached to the resistor.
The red and blue value stripes were then painted to match the original resistor using hobby paint. The result is shown in the photo below.

The wire-wound bleeder resistor #2 originally measured about 12K ohms - the marked value was 6K. Bleeder resistor #1, also 6K, was OK. I connected a variable DC power supply to the combination of #1 and #2 (they are in series) and ran the voltage up to about 100 volts. Then I noticed that the voltage at the junction of the two was correct (about 1/2 the input voltage)! And after this exercise, resistor #2 measured the correct value. At this point I suspected a bad connection between the resistance wire and the terminal. But the voltage was stable and did not change, even when the resistance wire and terminal were wiggled. So at this point, I decided to leave it alone pending the outcome of testing. This resistor would have been impossible to replace with anything similar to its original appearance.

Cabinet

The cabinet needed a good vacuuming inside and then cleaning on the outside with GoJo and 00 steel wool.

Testing and Alignment

Once the radio chassis was reassembled and the tubes installed, power was brought up slowly using a variac. AC power consumption was monitored using a watt meter, and a DVM monitored the B+. The radio came to life immediately and worked, but the volume was very low. I noticed that if I touched the grid cap of the 1st detector, the volume increased greatly. The schematic does NOT agree with the wiring around the three RF transformers as found. As found, the grid lead to the 24 first detector was connected to the lead from the top of the #3 RF Transformer to the tuning capacitor. The schematic shows the grid lead connected to the junction of the #2 and #3 RF transformers (and other 80 series receivers are drawn the same way). The RF wiring in the vicinity was rubber covered and falling apart. When I replaced it, I originally noted that the wiring was different from the schematic. So I used the schematic wiring (that area had been disturbed in a previous repair to insulate a grid cap lead with tape.) When I returned the wiring to the as found condition, the set worked great. But I never figured out why, or if there was indeed an error in the schematic. The "double spot" trimmer was still sealed with wax, so that could not explain the difference in performance.

All of the trimmer adjustments were sealed with wax, and the first IF transformer adjustments are hidden under an aluminum cover with no adjustment holes. The only accessible trimmers were the oscillator and 1st detector grid trimmers, and the 1st detector plate trimmer (which peaks the 1st IF transformer primary). I did not mess with the oscillator trimmer, as the dial settings were very close. I checked the other trimmers, but there was no improvement, so these were left alone. I did NOT adjust the first IF trimmers. The radio is quite sensitive for what it is - basically equivalent to an AA5 superhet in performance. The the sound was somewhat distorted. All the voltages looked reasonable. I did go back and trim up some of the resistors in the 2nd detector and audio sections, since I had left unchanged some resistors that were 20-23% high. When these resistors were trimmed to the specified values using a resistor substitution box, no improvement resulted.

The volume control was OK - there was no scratchiness, and the operation was smooth. The screen voltages were stable, and the two bleeder resistors measured very close to the specified 6K ohms. So no further action was taken.