Internal trigger-and-hold circuits are used by oscilloscopes to trigger the sweep circuit reliably at lower frequencies. At high frequencies, however, other methods are generally necessary. Oscilloscope sweep circuits usually synchronize to high-frequency signals using direct subharmonic locking of the sweep multivibrator. The fastest oscilloscopes use resonant-tunnelingdiode countdown oscillators. Such methods have been used with a 50-GHz Tektronix CSA803 waveform analyzer to display sinewaves as high as 110 GHz [1].
When plugged into the scope’s external trigger input, the circuit in Figure 1 can provide reliable, low-jitter triggering for both older and modern oscilloscopes. The output is in the form of fast step-pulses, as seen in trace A of Figure 2, which occur on every 256th RF cycle. This constitutes a subharmonic frequency which is accurately phaselocked to the RF signal.
The fast step output from this circuit was used to externally trigger the Tektronix 515A oscilloscope (circa 1960) for trace B. This trace shows a clear 40-MHz sinewave with hundreds of traces superimposed, even though the oscilloscope’s bandwidth is only 15 MHz and its risetime is 23 ns. Lower frequencies are equally clear using this circuit. Trace C shows the blurry display resulting from the 40-MHz signal using the scope’s internal trigger setting, with hundreds of traces superimposed. These are provided by optimal tuning of the HF SYNC circuit within the Tek 515A, used without the benefit of the Figure 1 circuit.
Signals well above the oscilloscope bandwidth, such as the signal in trace B, suffer severe attenuation and phase problems. Yet, they’re still useful for displaying narrow-band signals. Accurate timebase calibration with several cycles per division is easy with this crisp the slow startup of this nominally 50-ns/div sweep.
According to the graticule divisions, the step pulse of trace A would seem to indicate a 10-ns risetime signal. But when measured against the sinewave calibration signal, its rising edge covers about one RF cycle. This shows that its actual risetime is about 25 ns. Such timebase behavior is not uncommon in older vintage oscilloscopes. Often, the most linear region is found later in the sweep.
For the oscilloscope in Figure 2, pulses and waveforms would best be delayed by a suitable delay cable or delayed-trigger technique. Using these methods would cause pulses and waveforms to appear at least 300 ns into the trace, for proper display at 50 ns/div. The first 300-ns section of the sweep would be shifted offscreen and ignored, leaving the most linear part onscreen.
Reference:
- E. Ozbay and D.M. Bloom, “110-GHz Monolithic Resonant-Tunneling-Diode Trigger Circuit,” IEEE Electron Device Letters, Vol. 12 (9), p. 480-482, September 1991.
Reprints
