Atwater Kent 112 Chassis - update
Written by Bryce Ringwood   

The Atwater Kent Model 112 is a classic set dating from 1934. The model presented here was not in the standard cabinet, but had been installed in an art deco radiogram cabinet, which is at present being restored. It should look stunning, when complete.

This model features variable selectivity, AGC, a DX/Local switch and the other standard controls. The variable selectivity control is linked to the tone control, so that sharp selectivity coincides with loss of high frequencies. Frequency coverage is Medium wave, Long wave and two shortwave ranges 4-10 MHz and 10-18MHz

The audio amplifier section uses 6 valves - a 56 amplifier followed by two 56 triodes in push-pull to drive two 2A3 valves in push-pull. The set appears to work undistorted on one output valve (the other was faulty), from which I deduce the output valves are biased in class A linear.  The bias voltage of -35 or so volts is derived from the loudspeaker field coil in series with the HT transformer centre tap in an "ingenious" arrangement. There is no negative feedback in the output stage, as would be expected from a design of this vintage, as the negative feedback amplifier was yet  to be invented by Harold Stephen Black (US patent 2,102,671, issued in 1937) Nowadays there are anti-negative feedback rumblings, and this design simply doesn't require it. Interstage coupling is via transformers. The 2A3 is an output triode, still manufactured today and used in bespoke amplifiers where users want an arguably almost distortion-free output.

The detector and first AF stage is a 2B7 diode pentode, which also provides AGC voltage.

The remaining active valves are all 58s, comprising RF, Oscillator, Mixer and 2 intermediate frequency amplifiers. These are all perched on a separate upper chassis that is completely inaccessible for service unless all the wires connecting it to the lower chassis are unsoldered - there are many. There are no "tie-points" to allow fault diagnosis, so any hint of problems with the RF/IF section means that this chassis has to be removed and every component tested before it is replaced. It is absolutely essential that the wire colours of the original  be retained, if wiring has to be replaced.

The power supply rectifier is a 5Z3. The wiring to the smoothing choke (and possibly the mains transformer) can easily be replaced, as there are small tags inside the choke (and presumably transformer casing.Restored slow motion drive

The set has a shadowgraph tuning indicator. This is simply a vane which casts a shadow on to a screen. The vane is attached to a meter movement. When the vane is "edge on" the shadow is narrow (strong signal) and when it is normal to the light beam, the shadow is widest (weak signal). These were replaced by "magic-eye" tubes from about 1935 onwards with the introduction of the 6E5.

Any set of this age must be expected to need a great deal of work (even if, by some chance, it switches on and plays - please don't be tempted to do this.) By this time, much of the insulation will have failed, incorrect valves may have been substituted, fuses will have been replaced with things other than fuses and components will have changed their values. In other words, the chances of getting by with a simple repair are not good and by now the set can be expected to need a comparatively expensive restoration.

Atwater Kent 112 Chassis

I repaired the tuning drive first. The old rubber friction drives were hard and simply fell off. In some cases they were replaced with rubber drive belts glued to the rim (see photo). I rebuilt the small friction drive from "Wonda Fix" - a flexible epoxy and then machined it to the correct diameter. These parts were set aside, because the front panel and top sub-chassis had to be removed in order to rewire the IF transformers.Removal of the RF chassis is a little daunting because of the large number of wires connecting it to the main chassis. All the front panel and most of the sub-chassis wiring was replaced. Each IF transformer had to be rewired. I cut the existing wiring as close to the coil as I dared, then soldered new wire in place of the old, being very careful to use the correct colour wire. I tested each transformer to ensure it had the correct resistance before replacing it. The grid wire leading from two of the transformers has a spiral of ground wire wrapped round it. I used heat-shrink tubing to hold the ends in place. Next, each component in the deck had to be tested. I used a component tester, but this doesn't give an indication of whether the capacitors will withstand the working voltage. I didn't wan't to use an insulation tester on them, as it might have caused damage. I replaced the RF deck and rewired it to the main chassis.

I removed the incorrect "80" rectifier from the power supply and replaced it with a 5Z3 - correct type. I then gently applied power, and everything seemed to go well - all of a sudden, the 5Z3 began to glow blue and the power was removed sooner than immediately. There was clearly no hope for the electrolytic capacitors, which I replaced. I mounted the new (minute) capacitors on a tag strip supported by the old electrolytics.

I removed the power supply choke and rewired it with modern vinyl wire, the same diameter as the original rubber covered wire.

I next replaced the speaker field coil temporarily with a huge 500 Ohm resistor, so I could work on the set while the speaker was being repaired.

Before I replaced the front panel, I tested the set to see if anything else needed to be done. This time the 5Z3 did not glow blue and and the set came to life.


The "electric gramophone" was an early Garrard RC1 autochanger. The mechanism was based on a "wind up gramophone" with an electric motor replacing the spring. Speed control was achieved by means of a centrifugal governor. 

Lid open showing turntable. The piece of wire now connects to a switch. 


The pick-up head was a moving iron design requiring gramophone needles. 

All the wiring had perished,so the first step was to replace it as far as possible, lubricate all the parts and at least try to get the turntable to work. 

That done - the next thing to tackle was the pick-up head. Before modern plastics were in general use, there was a material called gutta percha. It is a tough rubbery substance, but in time it becomes hard and crumbly. This is the material used to hold the magnet in place in the pick-up head. It was also used to support the needle holder. Unfortunately, removal and reassembly is a very fiddly job and in the process I broke one of the leads to the pick-up coil. Even more unfortunately, it was the inner lead, so the coil had to be rewound. I removed 10000 turns of thinner-than-hair-fine wire from the bobbin, but could only get 6000 back on before the bobbin was full. I reassembled it anyway and hoped for the best. In place of gutta percha, I used modern silicone sealant. Perhaps what I would lose on the turns count, I would gain in freer movement of the modern compound.

I confess, my heart sank when I saw my client's new collection of 78 rpm records. As it turned out, I needn't have worried. The old scratchy recordings came to life and we bathed in nostalgia and sounds from a time long before even I was born.


Joomla template by a4joomla