In a previous blog, “Philco Radio and the Changing of ‘State of the Art’ Technology,” I wrote how, beginning in the mid-1930s, manufacturers of AM radios used a new circuit design, most often containing five tubes, which became known as the All American Five (AA5)... “AA5 radios eliminated the big, bulky power transformers prevalent in the large consoles that were popular before the war, which greatly reduced the cost of manufacturing and owning a radio, something especially important during the Depression.” Advances in tube technology also allowed for two or more stages of the receiver to be combined into a single tube, which helped bring the tube count down to five. The drawback to the AA5 was the potential for electric shock(!) which is directly related to the story of power cords known as “curtain burners.”
We start in the mid-1920s, when all consumer radios were battery-powered - some requiring as many as three separate batteries (the "A" battery was for the tube filaments, the "B" battery for the B+ and sometimes a C battery for grid bias. That "B" battery had to be big to provide upwards of 90V to the plates.) Replacing them, especially the large "B" battery, was expensive, not to mention annoying when your favorite program is on and you discover one of the batteries had died. Some folks used lead-acid car batteries, which they could recharge, but the liquid inside was dangerous and, if it spilled, created quite a mess. The introduction of radios that plugged into an electric outlet eliminated the battery problem, but since a part of the radio circuit still required Direct Current (DC), these new radios required circuitry to convert the Alternating Current (AC) coming from the power socket into DC. As an example, here is the Philco floor model 38, which contained a typical radio receiver power supply of the time:
Anyone who has had to move grandma’s floor model radio will tell you how heavy they are. The bulk of the weight is due to the component outlined in red – that’s a big transformer with an iron core. In the schematic we see the 120VAC source coming into the transformer which has three "taps," or splits. The “tap” off the transformer at the bottom provides 6.3VAC that heats the filaments of the tubes which, in turn, heat the cathodes (cathodes produce a source of electrons, the flow of which make the receiver work.) Direct Current is created using the rectifier tube, two capacitors and inductor arranged like the Greek letter Pi, and so is called a Pi network. Designers, tasked with driving a large speaker (marked in orange in both the schematic and photo) to deliver a big, room-filling sound, used the large speaker coil as part of the power supply.
The transformer also had another function, one which relates directly to the problem of electric shock: it isolated the incoming AC from the rest of the radio. That meant that the radio could be plugged in and turned on but it was still safe to touch to chassis. The same could not be said of radios employing the All American Five design. One big reason AA5 radios are cheaper to manufacture is that the incoming AC is connected directly to the string of tube filaments. The voltage drop across the tubes equals 121 VAC (for these tubes we just add the leading number of the tube IDs, 35+50+12+12+12 which almost exactly the source voltage of 120 VAC) thus eliminating the need for a transformer.
That’s great for eliminating the threat of a hernia but there’s a price to pay because, as we see in the schematic above, one side of the incoming AC is connected directly to the chassis! Without electrical isolation from house current, an electric shock was in store for anyone who touched the chassis. Though the cabinets were made of non-conducting wood or Bakelite there were large ventilation holes in the back panel, which were needed for ventilating heat from the tubes and other components, which meant they were large enough for fingers to slip through. And I know it sounds crazy, but some of the ventilating back panels were held in place with metal screws CONNECTED TO THE CHASSIS, providing another way to get a shock.
It gets worse, because some manufacturers placed the on/off switch in the cord itself. Since radio power cords were non-polarized (meaning both prongs were the same width,) depending which direction you plugged in the outlet a radio chassis could be “hot” without ever being turned on. (The fix, employed later on, was to create an isolated ground, called a floating chassis, with a capacitor and sometimes a resistor. It worked, but sometimes the capacitor, under great electrical stress, would fail and the shock problem would return.) Despite the shock potential, millions of AA5 radios were sold from the 1930s to early 1960s.
The AA5 design I’ve described is one in which the voltage drops across the tubes pretty much equals the 120V supplied by your favorite power company. But what if that isn’t the case? Take, for example, the Emerson 409, first sold in 1933…
This early tube radio used a “Tuned Radio Frequency” circuit which, although it did not perform as well as radios employing Superheterodyne, was a still a good seller for Emerson. (One version of the radio, which featured Mickey Mouse in various poses, is much-prized today. The above, Mickey-less version, is from my collection.) The Emerson designers must have felt pretty good about designing a radio that required only four tubes, making it even cheaper to manufacture than an AA5. Their challenge was that the voltage drops across each of the four tubes was just 6.3V for a total drop of 25.2 volts, leaving about 95 volts that need to be dropped. Since all four of this radio's tubes draw 300mA, we calculate 30 watts (95 volts x .3 amps) that need to be dissipated somewhere.
Emerson designers, in their zeal to produce the lowest-cost radio (and not wanting to put a heat-generating resistor inside the cabinet) decided to use Resistor Cord.
The idea of dissipating 30 watts over six feet of cord had some merit, in theory. In practice it had the potential to be an epic and dangerous fail if owners didn’t read - or ignored - warnings not to lay the cord under carpet or roll up any excess length in a bundle. That’s where we get the term “curtain burner,” because of the potential for starting a fire - although none of the websites used in my research cited any specific instance of a fire resulting from the cord. (The optimist in me hopes that means that whatever fires were started by the cords were quickly doused.) There’s no doubt, though, that all that heat accelerates the aging of line cords, which as a rule should always be replaced as part of any antique radio restoration.
When I began my restoration of this radio I had never even heard of a “curtain burner” line cord, let alone worked on a radio that had one. I jumped on a couple of websites (thank you, Al Gore, for inventing the Internet) and started my education. Fellow hobbyists made it clear that even if I could find one, simply buying a replacement resistance line cord should not be considered. All agreed on three potential solutions: a single resistor, a diode/resistor combination, or a capacitor/resistor combination. Hobbyist Paul Stenning does a terrific job describing the advantages and disadvantages of each on his website. Paul explains that both the resistor and resistor/diode solutions generate substantial heat, as opposed to the capacitor/resistor combination that is relatively cool, making the choice obvious. However, to get the right capacitor value involves calculating reactance - and readers to my blogs know how much I want to avoid math. Paul, bless his heart, not only walks us through the math but also gives us a link to a spreadsheet that does it all for us.
Then I found Greg Farmer’s paper “Resistance Line Cord Replacement” which he wrote for the Northland Antique Radio Club Newsletter. In it, Greg also describes the math involved to derive component values for the capacitor/resistor solution. To my surprise and delight Greg used my radio - the Emerson 409 - as his example. Greg also referenced the same spreadsheet (the one that does all the math for us) that Paul shared. I downloaded it and have to say it’s a pretty neat tool which easily produces numbers for all three potential replacements, as shown below. Note how the resistor-only solution requires almost 30 watts to be dissipated and the resistor/diode combination just under 20 watts, yet the dropper capacitor solution dissipates just 3 watts:
With the values of the capacitor and two resistors calculated, my task was then to figure out where to put them. This is a small radio and there’s not a lot of room either above or inside the chassis for all the components. And even though the capacitor/resistor solution generates just 3 watts, I did not want to add another source of heat. There was also the question of authenticity, which is an ongoing discussion in the hobby. Some folks work hard to make their radios to look exactly as they did when first manufactured, not just on the outside but inside, as well. Some go so far as to “stuff” the casings of the original components with their modern replacements. To that end Greg figured out a way to “stuff” the rectifier tube casing with a small circuit that replaced the line cord. (It's on his web page, and it's pretty damn clever.)
As much as I admired Greg’s solution, I opted for what I believe is a completely unique (and much less invasive) approach, which was to use a wall wart. You may not have heard the expression “wall wart” but you know what these things are - those chunky black boxes that you plug into the wall to supply power to your electronic devices which are so named because, as PC Magazine explains, “wall warts.. are hogs that… intrude into the surrounding sockets on a power strip. Most wall-warts contain a small transformer which isolates the AC source and so have non-polarized plugs. Because there is still a “hot” side to the radio, I replaced the prongs with those from a polarized plug and wired the prongs as indicated below. A third wire was required because we still need to supply 120V AC to the plate of the rectifier tube, as shown below. (Eagle eyes will note that the calculator returned a value of 6.94, but I used an off-the-shelf 6.8µF which would have required a .15 in parallel to get 6.95µF. Greg, with whom I exchanged a few emails, suggested the tolerance was not that tight so I stuck with just the 6.8µF.
Not shown is a fuse that was added as a safety precaution. (I’m happy to report that the fuse never popped and neither the cord nor the modified wall-wart got warm.) Given that the radio uses the TRF method I was very pleased at how nice it sounds. You can judge for yourself, here’s a video I posted of the radio with its new power cord. I hope I sound this good when I’m over 85 years old.