Recording Signals with an X-Y Recorder
One of the limitations of an oscilloscope is the relatively small size of the screen (generally 8 x 10 cm). If the trace is 0.3 mm thick, this gives a resolution of about 270 x 330 lines. A photographic record of the trace on the screen will be subject to the same limitations as resolution while making extra copies of Polaroid prints (the usual medium used in oscilloscope cameras) is by no means an easy matter and is relatively expensive.
The sampling oscilloscope can be considered a low-frequency oscilloscope, once the samples have been taken. As a matter of fact, a scan conversion has taken place. Since the sampling oscilloscope can enlarge signal details to a much greater extent than a normal oscilloscope, the sampling oscilloscope can be compared to a microscope.
Additionally, if we connect one of the vertical outputs and the X output of the PM 3400 oscilloscope to an X-Y recorder, we can get a recording of the signal on the screen of the CRT. On the recording, 2 cm corresponds to 1 cm on the oscilloscope screen, while the line width is about the same in both cases.
The resolution of the X-Y recording will be about twice as good as that of the Polaroid oscillogram. Furthermore, the X-Y recording has the advantage that it can be made on graph paper, thus making amplitude and time measurements much faster and more accurate.
Finally, the X-Y recordings can be reproduced with normal copying equipment. This gives excellent, cheap prints, which can be an important consideration when a large number of copies of the recording are needed, for example as educational material in schools and universities.
The Accurate Measurement of Signals
The independence of the sweep of the Philips PM 3400 oscilloscope from the time scale can be effectively used for high-accuracy measurements of hf signals. The signal in question is simply displayed on the screen and switched over to MANUAL SCAN, the spot manually adjusted to the point on the trace of interest, and the amplitude of the signal from the appropriate vertical output measured with a dc voltmeter.
The linearity of the amplitude ratio between the input and output signals was found, by measurement on several instruments, to be better than 0.2% for the middle 6 vertical divisions, measured between input and Y output. Measurements on the A and B outputs revealed linearity of better than 0.05% for all 8 divisions. The difference in performance between the Y output and the A and B outputs is caused by a network in the Y amplifier which compensates for the slight nonlinearity of the CRT found at extreme vertical deflections. The A and B output signals are taken off before this compensation network.
The accuracy of the output voltage is ± 5% and that of the V/cm control is ± 3%, a total of ± 8%. However, this error can be eliminated by introducing an adjustable attenuation of the output signal to be measured. This attenuation can be calibrated by means of an accurately known dc voltage applied to the input of the oscilloscope, for example. In this way, the total error can be restricted to about 0.1%. The signal is first recorded on an X-Y recorder in the SINGLE SCAN mode. Then, in the MANUAL SCAN mode, the spot is positioned at the point of interest and the OUTPUT signal is accurately measured. At the same time, the pen lift of the recorder can be released so an extra dot is obtained on the recording at the point in question. This procedure can then be repeated for other points on the signal.
Applications of this type are possible not only for signals with frequencies of several hundred MHz but also for signals of only a few kHz. In fact, it is possible to measure every part of a signal up to 200 µs (corresponding to the slowest time/distance setting of 20 µs/cm) after a trigger point, provided the repetition rate of the signal is about 10 Hz or more.