AT Super 660 WK3 radio (1951)

The AT660 has a reasonably large speaker and delivers a good sound. It has 5 frequency ranges: LW, MW and 3 SW ranges. On the dial, both frequency and wavelength are indicated. Most Western European radios of the time indicated only wavelength. The radio is built as an average superheterodyne using the following set of valves, all made by RFT (East German valve manufacturer).

• ECH11 frequency changer
• EBF11 IF amplifier and detector
• EF11 as AF pre-amplifier
• EL11 as output amplifier
• EM11 tuning indicator
• AZ11 rectifier

All valves have a “G8A” base which is why I bought this radio. These are practically only seen in German radios. The “G8A” is a base which enforces a correct insertion in the socket, helps to find the right orientation in dark places and holds the valve tightly in place. It can be seen as an improvement of the Octal base. A base that has advantages, but is expensive. After WW-2 better valve bases were developed that cheaper to produce. So I found it quite odd to see a radio from 1951 that uses G8A valves all over.

I bought this radio from a friend. The case looked nice and stylish, but a bit scratched and a few small pieces of veneer had come off. My friend told me the rectifier had a broken heater, so I thought it would be an easy repair.

After I replaced the rectifier and switched the radio on, I noticed some sparks in the new rectifier. I switched off immediately and tested the electrolytics. These had almost entirely lost their capacitance and had a considerable leakage current. These electrolytics are encased in oiled paper cylinders, sealed at both ends with pitch. The pitch was cracked and some white powder was leaking through the cracks. Because the old electrolytics were so large, it was easy to fit new capacitors in. I used 450V types because the originals were 450 V too. I thought this was overdone, but later on I found out it wasn't.

After replacing the electrolytics, I tried the radio. No sound at all. When I checked the output transformer, it turned out to have an interrupted primary winding. I looked at the schematic and saw a possible explanation: the terminals for an external loudspeaker on this radio are connected to the primary of the output transformer, not to the secondary. This means things go wrong if you connect a low-impedance speaker or accidentally short one of the loudspeaker terminals to ground. I found a suitable replacement transformer, though I was not sure of the impedance. It turned out to work, but the sound was distorted and the supply voltage was low. After 5 minutes, one of the HF filter capacitors across the HV winding overheated and started sizzling. I turned quickly off the radio and hoped the power transformer would not be damaged. For the moment I decided to do without the HF filter capacitors and tried the radio without them.

I measured the current through the output valve, which was about 100&nsp;mA. This meant he vale had ample emission, which is good, and the coupling capacitor from the AF pre-amplifier was leaky, which is bad. I measured only 2 MΩ resistance. I checked a number of other capacitors, and these were very leaky, too. Actually, testing with my HV power supply set to 250 V, the total of the leakage currents was 3 mA. So I decided to replace most of the paper capacitors. Fortunately, these RFT capacitors are relatively easy to fix invisibly: they are encased in a glass tube, sealed with pitch. I heated the pitch with a soldering iron, pull the old capacitor out and fit a polyesther capacitor in. Because all of the paper capacitors were 500 V types, I used 500 or 630 V polyesther capacitors.

The loudspeaker is a type with a field coil instead of a permanent magnet. This field coil is also used as the power supply choke. I temporarily replaced the choke with a spare choke to be able to test the chassis outside the case without having the the clumsy loudspeaker and baffle on my workbench. When I turned on the radio I was surprised to see the supply voltage rise to a staggering 480 V, to drop to 320 V after warming up, which was still quite high. I realised this was not the intended plate voltage, the loudspeaker would cause a 100 V voltage drop because its field coil has a resistance of 1500 Ω. That is much more than the resistance of the choke I was using. Now I understood why all the capacitors had such a high voltage rating! The power transformer delivers a higher voltage to compensate for the voltage drop across the loudspeaker's field coil. All the capacitors must be able to withstand the 20 second voltage surge at startup, that's why they are 450V and 500V types.

The radio was playing again, though not very impressively. Om MW, I received a few stations in the 500 kHz range, and hardly any at higher frequencies. I decided to realign the coils.

In that process I found out another defect: the 4 trimming capacitors that are used to align the MW and SW frequency ranges, had lost their capacity. These are silver plated ceramic capacitors. The connection between the silver plating and the connecion lug seemed to be interrupted. This explained why reception on the high end of the MW band was so weak. I was able to fix the two trimmers on SW by using conductive paint. The two trimmers on MW could not be fixed. I added two air trimmers (they can be seen on the photograph of the underside view). After this, the radio played very well again on MW and resonably on SW.

The general condition of the valves is quite good actually, as can be seen on the photograph of the EM11 tuning indicator, which is still very bright.