Selenium rectifiers: how to preserve them?

Looking at ads from old ‘Electronics’ magazines, manufacturers as Radio Receptor or Westinghouse claimed 60K to 100K hours life expectancy for selenium rectifiers. Keith Henney wrote of copper oxide rectifiers first used in railway equipment in 1924 and still operating in 1951. In my life I have just replaced three or four selenium rectifiers, servicing more than 150 radios, tape recorders or instrumentation sets, that used this kind of component; and every time the rectifier had already been burned or damaged when I bought the equipment. So, why replace these components in a preventive maintenance plan?

Selenium rectifiers were made by stacked plates of aluminum or steel, one face covered by a thin layer of selenium and by a soft-metal counterelectrode. They could be damaged by excessive humidity or by corrosive atmosphere: however, in this case, the entire set should show severe trace of rust and corrosion. Assuming that operating voltages and temperatures are normal, the other cause of failure can be an excessive current drained by the load. At high currents, the temperature rises over the safe operating values, about 75ºC, and the interelectrode layer melts at about 100ºC. In an old radio, abnormal current drain may be due to aluminum filter capacitors, when power is first switched on with no precautions, after many and many years of storage.

After years, even when stored on the shelf of the component dealer, aluminum capacitors ask for an oxide layer reforming process. The voltage on the capacitor should be raised slowly, to allow the thickness of the oxide dielectric to grow again and self-heal breaks and small holes in its surface. If the reforming is not performed, high leakage current results through the oxide and catastrophic failures may also occur. What often people ignore is that, according to Keith Henney, also selenium rectifiers show an increase of the reverse current when no AC power is applied for a long while. And also this leakage current rapidly decreases when the voltage is gradually increased to its operating value. The trick to safely operate old sets, leaving most of their original components unchanged, is to perform a short reforming cycle, starting with reduced voltage.

I always use a step-down transformer or a Variac to first apply roughly 50% of the nominal supply voltage for about 15 to 30 minutes. If nothing happens in the meanwhile, no overheat, no smoke, and all voltages stay around 50% of their nominal values, then I raise the supply voltage to its full value. I have experienced reliable operation of the power supply section after this treatment. The preforming cycle should be repeated each time the set was left non-operating for a period longer than few months.

Of course, a selenium rectifier must be replaced if its reverse leakage is still too high after the reforming cycle. Its body can be hot after few minutes and the B+ voltage is considerably lower than normal, in this case.

It would be interesting to know what is the experience of other people about.

Best regards,

Emilio

To thank the Author because you find the post helpful or well done.

Hi Emilio !

I've had to re-wire a functionning rectifier (Siemens B250C150) found in a Grundig 3397 of 1963 vintage.

The set's transformer has a setting of 220 V only and my power is at 230V. So I expected an HT+ to be a bit above the recomended value. (the heater voltage is at 6,7 V )

I could not get the 287 V prescribed into the service sheet, even without the original smoother caps (I tried new caps).

The rectifier was running really hot (I didn't measure the exact temperature but I could not put a finger on the box when the set had run for more than an hour). Last but not least it drove the power transformer above 65 °C on the outside metal core so I was afraid to fry the transformer (to get about 65° C on the outside, how much will it go in the inside of the deeper coil ? )

I've re wired the internal rectifier with four 1N4007 keeping all box, connexions and all as it was before. I've added a marking to specify the modification just in case. Then I devised that a 100 Ohm 5 W limiting resistor would put the set back in business. Now, the rectifier is, of course, cold, and the transformer does not go above 47°C.  So this one had to take a deserved retirement....

But I also have a Blaupunkt set (Arkansas 1957) and the rectifier on this one is cold, running fine, as is the transformer (does not warm at all). When I get the set, I checked the smoothers caps to find them perfect (I ran the set through a limiting bulb first) ; as are the voltages and currents ! This one, too is a Siemens one.

I also have a Grundig 4192 with a selenium rectifier with a fault. This set had seen abuse (missing tweeter, tubes in the wrong place, one loudspeaker cone destroyed so I won't discuss the state of the rectifier ;-)

So "your milleage may vary" ;-)

To thank the Author because you find the post helpful or well done.