[This article first appeared in 'Radio
Age,' the monthly newsletter of the Mid-Atlantic Antique
It is reproduced here by permission of the author. It has been edited for space. -- F.W.]
To subscribe to 'Radio Age' please click here.
(In some cases) the signal generator can serve as a wireless broadcaster. (More on this later.)
Usually when people talk about a "signal generator" they are referring to a device that generates signals above the audio frequency range. I find it helpful to have a separate audio generator whose frequency can be adjusted over the entire audible frequency range in order to test the frequency response of speakers and audio amplifiers.
This article focuses on RF generators, most of which will start at perhaps 100 kHz and go up to, say, 50 to 100 MHz in the case of very inexpensive types or up to a couple of hundred MHz in the case of better ones. Today it is possible to obtain signal generators that go way into the GHz (1,000 MHz) range, but these are not of interest to broadcast radio restorers.
This is a Heathkit Model SG-8 from the mid-1950s. It is typical of signal generators of that era that would have been purchased by someone servicing radios. It provides RF output from 160 kHz to 110 MHz, but has harmonics that can be used up to 220 MHz. It provides an audio signal output at about 400 Hertz.
RF signal generators began to be available for radio servicing about the time that more complex ac radios appeared and superheterodynes became more common. Battery radios of the early- to mid-1920s were sufficiently simple that to service them about all you needed was a voltmeter to check for proper voltages and a continuity checker to find short circuits and open circuits. By the early- to mid-1930s, most serious service shops would have had a signal generator on the workbench.
Old tube-type signal generators from the late 1930s through the 1960s that were used in radio service shops turn up at MAARC auctions pretty often. Brands like Heathkit and Eico are very common, and can usually be purchased for $10-15 or less. Other brands like RCA, Conar, Precision, Jackson, Hickok, and Knight also show up fairly often. Like the Heathkit, these are typically in utilitarian metal boxes containing a circuit with just a couple of tubes. (Although prior to WW II, companies like Supreme and Weston put their instruments in beautiful dovetailed wood cases. And Philco made some signal generators with a machined metal front panel that were sufficiently attractive to make nice display items today.)
During the past few decades, an entirely different class of signal generator has become available from companies like Hewlett-Packard and Fluke. These are precision instruments used in laboratories for more sophisticated tasks. Jammed with integrated circuits, some cost thousands of dollars when new. They feature precision digital signal synthesis, digital readout, and computer bus interfaces. Such instruments are excellent and work great for servicing radios if you happen to have one, but they are usually much more expensive and are very complicated to repair if anything goes wrong. The older simpler signal generators like the Eicos and Heathkits are easy to repair.
What are the differences between, say, a cheap Eico signal generator from 1955 and a precision Hewlett-Packard generator from the 1980s?
You may think I am trying to convince you to avoid the older, cheaper generators. Not really. If someone gives you a modern HP or Fluke precision signal generator free, or sufficiently cheap, for goodness sakes take it. It will do a splendid job. But for normal radio repairs an inexpensive Heath or Eico can do the job just fine. Let's assume that you have one of these typical old tube-type generators. How do you cope with the drift problem?
What should you look for when buying a signal generator?
Restoring that old generator that you just bought at an auction.
If you don't have a frequency meter, a generic calibration method can be used. Take a radio in good working condition and position it so that you have access to the antenna terminal or the loop antenna. If it is an ac-dc radio, BE SURE to use an isolation transformer. Don't try this without an isolation transformer or you may electrocute yourself and possibly damage your signal generator. Connect the "hot" output of the signal generator to the antenna terminal of the radio through a .01 µF capacitor and connect the ground lead from the generator to the chassis of the radio or the ground lead. If the radio has a loop antenna with no antenna terminal, remove an old loop from a junker set, prop it up close to the loop in the radio, and connect the signal generator leads to the main winding on the junker set loop. Let the signal generator warm up for about an hour before beginning the calibration check.
Jot down the exact frequencies of several local AM stations that you can readily identify. Ideally, you should try to get at least five or six stations spaced out along the dial. Switch the signal generator onto the frequency range that covers the broadcast band. Set the radio to the first station, making sure you have it tuned in to the center of the station frequency. Tune the signal generator to the approximate frequency of the station. As you tune the signal generator frequency through that of the station you will hear a whistle that changes in pitch. As you approach the station frequency from one side, the pitch will decrease. It will go through a point where the whistle disappears briefly, and then as you continue in the same direction through the station, the whistle will increase in pitch again until it gradually goes away. The point where the frequency of the whistle lowers to the point where it disappears is called the "zero beat" point.
"Zero-beat" the generator with the station and note the exact frequency of the station (not the frequency on the dial of the radio, which may be a little off) and the reading on the dial of the signal generator at the zero beat point. Plot these deviations on a piece of graph paper, connect them with a smooth curve, and save this calibration curve. This information will allow you to set your generator more accurately in the future. Here is an example of what a table of values and a calibration curve might look like from having done this exercise with a hypothetical radio and signal generator. In the example, the generator reads a bit high at the low end of the broadcast band, a bit low in the middle of the band, and a bit high again at the high end of the band. Let's say that in the future, you want to set your generator to precisely 1000 kHz. According to the curve, you should actually set the dial at 998 to get an output of 1000. (Your own generator may have a curve that looks quite different from this example, of course.)
|Station||Frequency (kHz)||Sig.Gen. Reading||Deviation|
This is what the data looks like plotted.
If you have a listing of powerful short-wave stations like the BBC in London and Radio Moscow, you can check the calibration of your generator in the Shortwave bands by using these known stations as per the process above. At the very least, you can use signals from station operated by the National Institute of Standards and Technology, which broadcasts strong signals at 5, 10, and 15 MHz.
I will not try in this article to explain how to align or troubleshoot radios with a signal generator. Those topics are well explained in most radio service textbooks.
Using your signal generator as a wireless broadcaster. It's great fun to listen to 1930s or '40s music or radio programs on your Philco cathedral. The effect is much more nostalgic than listening to modern talk radio. If you have cassette tapes of old programs or CDs of old music (which are readily available), all you need to do is take the audio output of your Walkman or CD player and use it to modulate the output of a tiny AM transmitter--a wireless broadcaster, or phono-oscillator as they were sometimes called.
So, enjoy your old signal generator both for servicing radios and for listening to old music or programs! Signal generators are cheap enough that you can afford to have one in every room of your house.
Biographical note. The author is a member
of the Mid-Atlantic Antique Radio Club and the Radio
A manager at the National Institute of Standards and Technology, he has been collecting antique radios for about 20 years.
Click here to send Brian e-mail.
You are visitor number 754 to this page this year.
Archives Home  | What's New in the Archives?  | Terms & Conditions  | Master List  | Friends of the Archives
European Radios  | Japanese Radios  | Vintage Audio  | Communications Radios  | Test Equipment
Unknown Radios  | Featured Articles  | Show Us Your Workbench!  | What's it Worth?  | Radio Attic
Links!  | Send Photos  | We're on FaceBook!
"Archives" are maintained just for fun by Steve Adams at the Radio
When cabinet material or color differences are apparent or suspected in donated photos, I have included additional photos of certain model numbers. Due to time constraints, I must rely entirely on the accuracy of information supplied by contributors with their photographs. As such, no guarantee is made that the model numbers presented here are correct.
Images may be used
for non-commercial purposes with attribution.
© 2001, 2014 the Radio Attic