When working on old valve radios, get into the habit of using a 240V isolation transformer at all times. These sets were all manufactured before OH&S became fashionable, and the exposed 240VAC wiring can be deadly. While an isolation transformer won't protect you under all circumstances, it will protect you from the majority of situations - ie where you are likely to come between the active terminal and an earthed object.
I also put a new three-wire power cord on all sets I restore, and make sure that the earth is connected to the chassis of the set. (This is not possible with some AC/DC sets where one side of the power input goes directly to chassis).
When removing the top cap connections to the older valves, don't just grab and pull up as you'll often end up with the top cap of the valve as well. Carefully prise open the clip slightly before pulling off and you won't have to worry about trying to resolder and re-glue the top cap back to the valve - assuming that there is a long enough piece of wire sticking out of the valve on which to solder.
There are a number of ways to check if the oscillator section of the mixer valve is actually oscillating. Three of these are:
- Use a high impedance meter (preferably a VTVM with the series multiplier resistor fitted in the tip of the probe so that the capacitance load on the oscillator is minimised), measure the DC volts on the oscillator grid. This is pin 5 on mixers such as 6A8, 6J8, ECH33 etc. If the oscillator circuit is working, a negative voltage between neg 5v and neg 10v should be measured. It is negative because the grid is driven positive on the positive peaks of the oscillator signal and draws grid current. This grid current flows to earth through the grid to cathode resistor (usually 50k) and produces a negative voltage with respect to earth.
- Measure the volts at the oscillator anode (pin 6 on 6A8, 6J8, ECH33 etc) and then short out the oscillator section of the tuning gang. The volts should drop, eg. from +150 to +100, if the oscillator was working. The volts drop because the negative volts on the oscillator grid (see above) will disappear when the oscillator stops. This causes the oscillator section of the valve to draw more current which drops more voltage across the series resistor from the HT supply and thus lowers the oscillator anode voltage.
- Set the tuning on the radio under test to around the centre of the broadcast band ie approx 1000Khz. Hold a small portable transistor radio, tuned to approx. 1455Khz, near the oscillator coil assembly. Sweep the tuning of the set under test backwards and forwards across the centre of the broadcast band. If the set is oscillating, you should hear the transistor radio background noise reduce as the oscillator signal picked up by the portable radio drives its AGC voltage up and the IF gain down.
Always measure the cathode and grid DC volts on the output valve to ensure that it is properly biased. The valve data books will give the correct bias volts. The grid must be negative with respect to the cathode, and the two most common measures to obtain this bias voltage are:
- The use of a resistor(s) between the CT of the HT winding and earth. This will produce a negative voltage with respect to earth, and this is usually fed into the grid circuit via a 500K resistor.
- The use of a resistor between the cathode and earth of the output valve. This will produce a positive voltage at the cathode. This resistor is usually bypassed with a 25µF capacitor if not bypassed the cathode resistor introduces negative feedback into the output stage.
When removing the top cap connections to the older valves, don't just grab and pull up as you'll often end up with the top cap of the valve as well. Carefully prise open the clip slightly before pulling it off and you won't have to worry about trying to resolder and re-glue the top cap back to the valve assuming that there is a long enough piece of wire sticking out of the valve on which to solder.
Don't just plug in and turn on those untested indirectly heated rectifiers such as a 6X5GT otherwise you are likely to smell a frying back bias resistor and power transformer. Always first test to ensure that there is no cathode to heater short. As one side of the heater circuit is usually earthed, a cathode to heater short will put a direct short circuit on the rectified HT output.
If you are having problems with crackles and pops and/or varying volume, suspect the two mica capacitors, usually approximately 100pf, around the detector valve. One is usually used to feed the IF signal from the anode of the final IF stage to the AGC detector diode, while the other is used as an IF filter at the base of the secondary of the final IF coil.
On the more modern sets fitted with AGC circuitry, always measure the AGC line to ensure that the AGC circuit is working correctly. The AGC line should be approximately zero volts or slightly negative with no signal, and rise to between neg 5 and neg 10 volts with a strong signal.
To repair small tears and holes in speaker cones, buy some craft glue and black sewing interface material from a craft shop. The material is very cheap and reasonably close to the texture and grey colour of most speaker cones.
If you have determined that a winding on an aerial oscillator or IF coil is open circuit, carefully remove the coil from the set after first drawing a small diagram showing how it is connected. Then remove the coil from the can, and using a powerful magnifying class, inspect the connections from the coil to the terminal posts and also the outer layer of the coil. I have found some where the coil wire was poorly soldered to the post and had come adrift, and a few others of the older, unwaxed type, where corrosion, often visible as a small blue/green spot, has eaten through the wire.
If you don't bulk replace all capacitors but rather just test them for leakage and replace as necessary, you should also check the actual capacity of the capacitors with a capacitance meter. I have seen quite a few that measure, as good when tested for leakage, but in fact have been open circuit and have no actual capacity.
When there is a need to align the IF stages of older superhet sets, don't assume that the IF frequency is 455Khz. Always check the circuit as the actual IF frequency used by the many manufacturers varied greatly, both higher and lower than 455Khz, before it was standardised at 455Khz in the late 1940s.