We used multiple models of LOTO oscilloscopes, with bandwidths ranging from 20M to 100M , to measure passive and active crystal oscillators . We tested the waveforms of 12M, 15M , 24M, 48M, and 80M crystal oscillators. We also analyzed why we often measure crystal oscillator waveforms that are not square waves? Why sometimes we can't measure crystal oscillator waveforms? Why is it unnecessary to waste oscilloscope bandwidth to meet the needs of measuring crystal oscillators ?
We usually encounter the following crystal oscillators:
Active crystal oscillators are usually 4 -pin SMD, while passive crystal oscillators are usually 2-pin direct plugs. Active crystal oscillators are usually 3.3V and 5V , and the frequency can be made relatively high. Passive crystal oscillators do not require power supply, but generally need to be connected to two very small capacitors to help start oscillation.
We specialize in welding 3.3V active crystal oscillators with 12M, 24M, 48M, and 80M frequencies for testing:
When measuring the crystal oscillator waveform, you need to use an oscilloscope with a bandwidth higher than the crystal oscillator frequency.
Bandwidth list of LOTO oscilloscope models
|
model
|
Bandwidth(Hz)
|
Maximum sampling rate (SPS)
|
|
OSC482
|
20M
|
50M
|
|
OSC802
|
25M
|
80M
|
|
OSCA02E
|
35M/60M
|
100M/200M
|
|
OSC2002
|
50M
|
1G
|
|
OSCH02
|
100M
|
1G
|
We start with the lowest frequency --- 12M active crystal oscillator. Since the bandwidth of our current lowest bandwidth oscilloscope OSC482 series is 20M , any type of LOTO oscilloscope can be used to measure the 12M crystal oscillator . Of course, the higher the bandwidth, the better.
When measuring the crystal oscillator waveform, the probe needs to be set to the X10 position, because this can minimize the equivalent input capacitance of the probe, maximize the input impedance, and reduce the impact on the crystal oscillator. At the same time, we adjust the voltage position to 0.1V per grid.
Let's take a look at the LOTO oscilloscopes of various bandwidths measuring a 12M crystal oscillator:
20M bandwidth OSC482 measures 12M crystal oscillator:
35M bandwidth OSCA02 measures 12M crystal oscillator:
OSCA02E with 60M bandwidth measures 12M crystal oscillator:
The waveform of 12M crystal oscillator under 50M bandwidth OSC2002 :
The waveform of 12M crystal oscillator under OSCH02 with 100M bandwidth :
Next, we test the 48M active crystal oscillator. OSC482 and OSCA02 cannot be tested because their bandwidth is not enough to measure the 48M waveform. We use OSCA02E for testing:
The waveform of 48M crystal oscillator under 50M bandwidth OSC2002 :
The waveform of 48M crystal oscillator under 100M bandwidth OSC2002 :
Next, we use the 100M bandwidth OSCH02 to measure the waveform of the 24M passive crystal oscillator.
Compare the waveform of the 16M crystal oscillator sent by the customer with that of the desktop oscilloscope :
From the many waveforms measured above, we can find that the crystal oscillator waveform is sometimes measured as a sine wave, sometimes as a dome-shaped sine wave, and sometimes as a bell-shaped wave close to a square wave, none of which is the ideal theoretical square wave. The rule is that the closer the bandwidth is to the measured crystal oscillator frequency, the closer the measured sine wave is; the more the bandwidth is greater than the measured crystal oscillator frequency, the closer it is to a square wave.
These waveforms are not errors, in fact, they are all correct. The reason for this phenomenon is the relationship between bandwidth and measured frequency.
As shown in the figure above, the square wave can actually be equivalently decomposed into countless sine waves with different frequencies and amplitudes. The blue one is the fundamental sine wave with the same frequency as the square wave, the green one is the harmonic of 3 times the frequency, the orange one is the harmonic of 5 times the frequency, and there are 5th harmonics, 9th harmonics ... infinite harmonics.
When a square wave is measured by an oscilloscope with a certain bandwidth, this square wave is equivalent to countless sine waves with various harmonics being measured. Harmonics above the bandwidth of the oscilloscope are severely filtered out by the oscilloscope, leaving only some harmonics below the bandwidth, so the phenomenon we just saw appears. The closer the bandwidth is to the square wave frequency, the fewer harmonics are retained, and the closer it is to a sine wave. If only the fundamental wave is retained, then it is a standard sine wave. We can see from the following figure what a square wave with different numbers of harmonics retained will look like.
Does it sound familiar? These deformed square waves have basically appeared in our actual measurements.
When measuring crystal oscillators, we may encounter a situation where passive crystal oscillators often cannot be measured. The general reason is that the crystal oscillator stops oscillating after the oscilloscope probe is connected.
Passive crystal oscillators generally need to be connected to PF- level capacitors at two pins to help oscillate. The probe of the oscilloscope has input capacitance, which is generally also PF- level. This is why we emphasized the need to use the X10 gear to measure at the beginning . On the one hand, the probe bandwidth of the X10 gear is higher. Secondly, the input equivalent capacitance of the probe is smaller. When connected to the crystal oscillator pin, it is equivalent to changing its oscillation capacitance. If the impact of this change is large enough, the crystal oscillator will stop oscillating, so it cannot be measured.
Actual test video link:
In fact, in our actual application, we basically don't need to use an oscilloscope to measure the waveform of the crystal oscillator. So we don't need to buy an oscilloscope with a very high bandwidth. If you want to know whether the crystal oscillator is oscillating normally, you can basically determine it by measuring the voltage with a multimeter. For example, for the 3.3v active crystal oscillator in the experiment above, use a multimeter to measure the DC voltage of the output pin of the crystal oscillator. If it is about 1.6v , it can basically be determined that the crystal oscillator is oscillating.
提升卡
变色卡
千斤顶
京公网安备 11010802033920号
