A simple 2-tone-generator for testing the linearity of SSB transmitters

This device is as simple as useful. With 2 transistors and an oscilloscope it is fairly easy to test the linearity of your SSB transmitter. The 2-tone-method is a standarized test method that can be used to determine the maximum output and the amplification characteristics of a sideband transmitter.

The method involves 2 audio tones simultaneously applied to the microphone input of the DUT (device under test). The two frequencies must not be harmonically related and must both fall within the audio passband of the transmitter. Also they should have a significant spacing in the audio spectrum. 850 Hz and 2200 Hz are a good choice, for example. If you own a spectrum analyzer with an appropriate resolution the Third Order Intermodulation products (IM3) can also be made visible. I’ll refer to that in a later page, today we want to talk about the amplitude side only.

First, here’s the circuit:

Wien bridge oszillator as two-tone-test gegenrator
Wien bridge oszillator used as a two-tone-test generator for checking SSB transmitters’ linearity

The heart of the device are two oscillators producing 2 sine waves which generate a new waveform that is formed by superposition of the 2 single signals. It resembles an amplitude modulated carrier and should look like this when displayed with an oscilloscope:

SSB 2-tone test signal: Transmitter correctly adjusted
SSB 2-tone test signal: Transmitter correctly adjusted

The “balance” of the tone generator circuit must be carefully adjusted to regulate the amplitudes of the two audio signals in a manner that the cross section centering the waveform is as sharp as possible. Then the two audio signals are nearly 100% equal which will result in the demonstrated signal amplitude waveform.

Wrong adjustments of the transmitter can be easily observed like the following pictures show. First one is an amplifier that is overdriven:

Overdriven SSB transmitter operating close to saturation
Overdriven SSB transmitter operating close to saturation

This is a sample of the classical “flat topping”. At least one amplifier stage in the exciter is running in saturated mode thus not being able to deliver proper amplification of the signal’s higher amplitudes. The transmitter will sound very distorted. Also a high set of IM3 products will appear thus make your signal much wider than acceptable.

The following picture shows an SSB amplifier running with inadequate bias setting:

Wrong bias in SSB exciter, increase of quiescent current strongly recommended!
Wrong bias in SSB exciter, increase of quiescent current strongly recommended!

The bias in this case is much too low so that lower voltage areas of the amplitude are not amplified sufficiently. Bias setting must be increased to ensure proper amplification.

To measure the peak output of your transceiver you can either use the amplitude reading of your oscilloscope. Divide the peak-to-peak voltage by 2.81 (two times the square root of 2) to convert p.-p- voltage to effective voltage, square it and the divide the result by your 50 ohm resistive load.

Veff. = Vpp. / 2.81 (I)

P = Veff.² / R (II)

Or you use a VTVM (valve tube voltmeter) with an RF probe that gives you an RMS reading of the transmitter’s output voltage.

Keep in mind: The power you’ll get is half of the peak output power of your transmitter!

If you managed calculus, then it’s quite clear why. 😉

73 de Peter!

18 thoughts to “A simple 2-tone-generator for testing the linearity of SSB transmitters”

  1. Does not say what size coax to use on this project,and what is good for inside of radios.

  2. Hi, I was wondering, since capacitor type is not specified other than the electrolytics, if ceramic capacitors could be used for the 6.8 and 22 nF ones? Thank you.

    1. Hello Jon,

      yes, you can use ceramic caps without any problems. Frequency stability is no problem due to the low audio frequencies generated.

      73 de Peter

      1. Hi Peter,

        Thanks for responding so fast. That definitely tackles my junk box only having ceramics, but brings up another issue:

        I only have the 22 nF ceramics and none of the 6.8 nF. I assume I could change the frequency by changing some of the resistors. I have 6 of the 22 nF, what resistors would I change and to what value to get the frequency that would have been produced with the 6.8 nF?

        Thanks again!

  3. So, how would I go about changing the frequency of one of the oscillators to something like 1 Khz instead?

    1. Hi, that is a matter of try and check. If you increase C and R the frequency decreases and vice versa. But there is no 100% linear function. I suggest trying it on a breadboard and then get the best values out of it.

      vy 73 de Peter

  4. Good day Peter, i would like to ask where do i connect the output signal, directly to the mike or through a empty microphone plug, please be patient as i am new to electronics , i do have a 20mhz Chinese pc oscilloscope a wattmeter and digital voltmeter ,as the finals are burnt in my radio ,i’m waiting for arrival ,so in the mean time i could build this circuit ,because i don’t have a 2 tone signal generator ,thankyou for sharing this project i greatly appreciate it, greetings from nico Italy

    1. Hi Nicolo, you can connect the output of the generator to a microphone plug for the radio you want to test. I receommend using a series resistor of 2.2k (for decoupling the radio from the generator) and a parallel capacitor 0,01uF (10nF) to GND to avoid rf stray energy intruding into the radio. That works fine for my homemade radios and it should also work with commercial ones.

      vy 73 de Peter

Leave a Reply

Your email address will not be published. Required fields are marked *