It’s hard to say what would possess someone to spend the past four years building something like this, but anyone who knows Eric can appreciate the talent that he brings to this project. Eric wanted to make sure we thanked Norm Braithwaite for all his help.
A pair of twenty-five watt amplifiers is trivial today, but in 1935, before the advent of the 6L6, it was noteworthy. But the real show-stopper was the speaker system. I have never heard another one like this in my life. The 18-inch Jensen is a wonderful piece of equipment that almost didn’t survive its initial test. After winding the bass-boost chokes last year, Eric turned up the volume and found that they were resonant at 100 Hz. The cone literally jumped out of the basket … but on Sunday, the system sounded wonderful!
Another glimpse of Allen Lein's genius appeared this year in the form of an elegantly modified Morris coil winder. Virtually unchanged since they first appeared on the market decades ago, these things have tempted technicians and repairmen with the promise of convenient replacement coils and chokes for three generations. For most of us, they've proven to be little more than frustrating and useless toys.
In Allen's capable hands, the clunky old Morris Winder is transformed into a practical tool. Aside from some carefully chosen mechanical alterations, Allen installed an electronic counter that takes the guesswork out of winding.
The sample coils that Allen showed us were indistinguishable from factory products.
Morris Coil Winder
Allen Lein's modified Morris
One of the most intriguing demos was, on the surface, quite simple. But the devil is in the details.
It all started two years ago when someone asked what to do about the increased line voltage and the effect it must be having on their old radios. Bill Fanum suggested connecting the secondary of a filament transformer out-of-phase, in series with the primary, to supply reduced line voltage. Everyone expressed a collective "huh?."
Subsequently, a couple people tried it and it works remarkably well, as long as you don't exceed the current rating of the transformer secondary. How it works is intriguing:
Filament transformer used to lower line voltage
We breadboarded the thing with voltmeters and AC ammeters and the results were surprising. Meter M1 acts just as you would expect; plug a sixty-watt bulb into the output and M1 indicates 0.5 amps. It's meter M2 that doesn't make immediate sense. With no load on the output, M2 indicated 70 ma; the idle current of the filament transformer. When we plugged-in the sixty-watt bulb, the current in M2 decreased slightly. When we replaced the sixty-watt lamp with a one hundred-watt lamp, the current through M2 increased. Bill Schmitt explained it, "With the secondary winding out-of-phase, the magnetic field created by the secondary bucks the field created by the primary. This actually diminishes the current flow in the primary ... up to a point. Continuing to increase the current in the secondary will eventually produce fields that overcome the polarity of the primary fields and cause an increase in primary current." We connected a variable load across the output outlet and were able to reduce the primary current down to a null then back up again.
The output voltage behaves just as you would want it to. With the secondary connected out-of-phase, a 6.3 volt filament transformer drops the line voltage about 6.3 volts. If you connect the secondary in-phase with the primary, the output voltage increases; so be careful, make sure the secondary is properly phased, and fuse the line input.
Finally, no workshop would be complete without a presentation by former NARC president, Matt Hyman. This year, Matt demonstrated a Teac model GF-350 LP to CD recorder. It originally retailed for about $350 and sounded so nice Sunday that somebody bought it from him.
Special thanks to Allen Lein, Bill Schmitt, Matt Hyman, Bruce Wagoner, Eric Zetterwall, Bill Fraser, Aimee Sahlsteen, John Lieberg KØFQA, and Doc Murphy KØGRM for all their help with this year's event.
Our continuing efforts to create a set of performance standards for vintage radios and chronicle the results have provided some educational moments, to say the least. I truly thought by now that we would have a set of statistics and and data on a dozen different classic radios, instead, we're still analyzing our poor, tortured RCA model 143.
Figure-skating competition and sensitivity measurements have a lot of similarities, both have rules that seem to be arbitrarily applied. Even the revered Radiotron Designer's Handbook1 is apologetically vague, "With sensitive receivers noise forms a large part (or all) of the output in sensitivity measurements, and alternative methods of measurement which specify the input required to give a minimum signal-to-noise ratio are becoming increasingly popular. One method is to specify the input required to give equal signal and noise outputs i.e. when the modulation is switched off, the output power is halved. A signal to noise ratio of 15 dB is more commonly used for communication work in England, this ratio being taken as providing satisfactory intelligibility."
Other references are equally conflicted. The 1943 edition of Radio Engineers' Handbook by Terman, page 975 and the 1948 Essentials of Radio by Slurzberg & Osterheld, page 245 say to check sensitivity by turning the volume wide open and see how many microvolts input it takes to get 500 mw audio out (!). They say that if the radio won't put out 500 mw, to use 50mw. [You better disconnect the speaker when you go up to 500mw, and use a resistive load.]
The 50 mw output is an old standard.
Here it is as part of an ad for Philco's 1929 model 95.
The ad originally appeared in the March 1930 edition of Radio Magazine.
Where Terman, Slurzberg & Osterheld came up with 500 mw is a mystery. Fifty milliwatts can be heard across the room; five-hundred could blow out your speaker. Perhaps the simplest and most sensible instructions were found in the Collins 390 manual:
Amplitude-Modulation Sensitivity Test
Use a 30 percent modulated
a. Connect the signal generator to the BALANCED
Dummy antenna for balanced input should be
the DA-121/U, part of MK-288/URM.
b. Connect the ME-30A/U in parallel with a 600-ohm 1-watt resistor between LOCAL AUDIO
terminals 6 and 7.
c. Tune the signal generator and the receiver to
the same frequency.
d. (For the AM Broadcast band) Adjust the signal generator attenuator to 4 microvolts output.
e. Adjust the LOCAL GAIN control to obtain a
ME-30A/U indication of 2.45 volts.
f. Turn the modulation off. The ME-30A/U
indication should be not more than 0.77 volt.
The R-390 manual asks for 2.45 volts rms across 600-ohms, or 10 milliwatts with the modulation on and 0.77 volts rms or 1 milliwatt with the modulation off. This is a 10 dB signal to noise ratio.
We tried a bunch of different methods. The system above is by far the simplest and most rational.
The important figure is the 10 dB signal to noise ratio.
2.45 across 600-ohms is fairly easy to read on a decent voltmeter.
With a 4-ohm load, the correct voltage for 10 milliwatts is 0.2 volts AC.
The dummy antenna supplied with the URM-25 signal generator is shown below.
As supplied, it does not have a 50-ohm termination as required by the generator.
We added R1 to provide the 50-ohm termination required by our URM-25E signal generator.
The signal-strength meter on the signal generator will only be accurate if the generator is properly terminated.
Our 1934 RCA model 143 yielded the following results at 890 kHz:
Output in Milliwatts
A proper sensitivity assessment would be taken at several points throughout the AM band.
If you were at the presentation in February,
you saw a completely different method than described here. We tried a lot of different approaches. As previously stated, Collins' method is the most rational. Modern sensitivity claims are often given at 6 dB, the minimum useful signal.
Stay tuned ... There's a LOT more to this discussion.
I'll be updating this site daily.
1. Radiotron Designer's Handbook, Langford Smith, 1952, Wireless Press, Sydney, Australia, pg. 1302