Use of digital oscilloscope and MIPI-DSI signal measurement

Preface

Digital oscilloscopes are mainly used for time domain waveform testing and measuring the changes in voltage/current over time. MIPI-DSI is a standard interface protocol developed by the MIPI Alliance for display devices. Here I record some of my summaries on learning how to use digital oscilloscopes and MIPI-DSI signal testing.

1. The main indicators of oscilloscope

The work of a digital oscilloscope can be divided into the following parts: amplifying and attenuating the signal collected by the test pen, ADC converting the signal to digital, storing the converted data in the cache, and reconstructing and displaying the signal. The front-end amplification and attenuation circuit determines the bandwidth of the oscilloscope, the analog-to-digital conversion circuit determines the sampling rate of the oscilloscope, and the cache determines the storage depth of the oscilloscope. The following describes these three indicators separately.

1. Oscilloscope bandwidth

Signal transmission in the circuit will be affected by capacitance/inductance, which is illustrated by the frequency domain response of the amplifier circuit. As the input signal frequency increases, the amplifier gain decreases in the high frequency region, and its bandwidth is defined as -3DB (20Log (A1/A0)) corresponding to the frequency. The bandwidth of the oscilloscope is defined in the same way. When the test signal frequency approaches or exceeds the oscilloscope bandwidth, it is reflected in the waveform as the attenuation of the signal amplitude. The bandwidth of a general oscilloscope is expressed as 1GHz, 4GHz, etc.

2. Oscilloscope sampling rate

The measured analog signal is continuous in time and amplitude. After being processed by the front-end analog circuit, the ADC circuit performs A/D conversion on the signal. The signal is reconstructed and displayed after digital processing by the back-end module. The frequency of the sampling clock during A/D conversion is the sampling rate of the oscilloscope, such as 20GSa/s (sampling 20G points per second). The quality of the waveform reconstruction by the back-end circuit depends on the sampling rate and sampling accuracy of the ADC (the number of bits of the ADC). According to the Nyquist sampling theorem, if the sampling rate is twice the bandwidth of the measured signal, the waveform can be reconstructed and restored. Otherwise, the waveform will overlap and cannot be accurately restored. This should also be observed in actual testing. The ADC bit number of oscilloscopes on the market is generally 8 bits. Without considering other factors that affect the measurement error during the ADC conversion process, there is 1/256 of the full scale. It is more accurate to use a multimeter when accurate DC measurement is required.

3. Oscilloscope sampling depth

Due to the limitation of the processing speed of the back-end circuit, the data converted by the ADC cannot be processed in real time and then the waveform can be reconstructed and displayed on the screen. Instead, the data converted by the ADC is first stored in the cache, and the back-end processing circuit reads the data from the cache. The depth of this cache is called the oscilloscope sampling depth, such as 100M (a maximum of 100M sampling data can be stored). The length of the waveform that the oscilloscope can continuously sample at one time = sampling depth/sampling rate. In the actual use of the oscilloscope, this situation may be encountered. After the sampling depth is fixed, the sampling time of the horizontal axis is increased or decreased, and the sampling rate will also change. Generally, the oscilloscope will automatically adjust the sampling time to improve the user experience. At this time, it is necessary to pay attention to whether the sampling rate meets the requirements.

2. Introduction of Oscilloscope Probe

The parasitic capacitance and inductance in the probe transmission line will affect the high-frequency signal. The larger the bandwidth of the probe, the smaller the parasitic capacitance and inductance are required, the more complex the process is, and the more expensive it will be. Currently, the probe can be active or passive according to the implementation method:

Active probe: built-in amplifier circuit, requires oscilloscope power supply, such as differential probe and current probe;

Passive probe: only passive components, such as high-impedance passive probe;

The probe and the input interface circuit of the oscilloscope together constitute the detection system. The oscilloscope can be set to different gears to match different probes to meet different measurement requirements:

AC position: For AC signal testing:

DC1MΩ: used for low frequency (generally below 500MHz) measurement;

DC50Ω: For high frequency signal measurement

1. Passive high impedance probe

Passive high-impedance probes are generally used as general-purpose probes with the widest application scenarios. They can be used to measure power-on timing, high-frequency signals within the bandwidth, etc. The following are their typical parameters:

Probe input impedance: 10MΩ

Probe measurement bandwidth: 500MHz

Probe attenuation ratio: 10:1

When using a high-impedance passive probe, the oscilloscope setting gear should be selected as DC1MΩ. In addition, the probe attenuation ratio refers to the attenuation of the detected signal to 1/10 of the original signal after passing through the probe. The oscilloscope can set the corresponding magnification factor so that the signal amplitude displayed on the screen is consistent with the measured signal. Generally, a probe with an attenuation ratio of 100:1 is suitable for high-voltage measurement, and a probe with an attenuation ratio of 1:1 is suitable for ripple measurement.

2. Active differential probe

The differential probe contains a differential amplifier that can subtract and amplify the differential signal and is generally used to measure high-speed differential signals.

3. Active current probe

The current probe is based on the Hall effect. The magnetic field around the power line outputs a voltage through the probe Hall sensor. The voltage value is proportional to the current. The typical probe conversion coefficient is 0.1V/A.

The minimum detection capability of a general probe is a few mA. When more precise measurement is required, the wire to be tested can be wrapped around the probe several times to detect smaller current signals.

When there is no current probe but only a high-resistance passive probe but you want to test the current signal, a feasible method is to connect a resistor in series with the line to be tested and determine the change in current by measuring the voltage difference across the resistor.

3. Oscilloscope capture signal

1. Oscilloscope triggering mode

Edge trigger

<>Rising edge trigger

<>Falling edge trigger

Pulse width trigger: You can capture specific waveform behavior by setting the pulse width.

2. Oscilloscope trigger mode

Auto: After setting the trigger condition, if the trigger condition is met, the waveform will be triggered and displayed. If the trigger condition is not met for more than a period of time (usually tens of ms), the oscilloscope will automatically generate a trigger and capture the waveform display. It is generally used to capture continuous periodic signals.

Normal: The oscilloscope is triggered strictly according to the trigger conditions. If the conditions are not met, the oscilloscope will not trigger. If the new trigger conditions are met after a trigger, the waveform will be updated. Otherwise, the waveform will not be updated.

Single: It will only be triggered once according to the trigger condition. It is generally used to capture the power-on and power-off timing, etc.

MIPI-DSI signal measurement

1. LP signal measurement

The LP transmission frequency is generally 10MHz, and the oscilloscope sampling rate is generally set to be greater than or equal to 50MS/s. The higher the sampling rate, the smaller the waveform distortion.

Sampling rate: 20GS/s

Trigger condition: edge trigger (generally, the read operation identifies the ID during the power-on stage, and can be triggered by the rising edge of the power supply +D0N/D0P)

Trigger mode: Single

Escape Mode: A special mode in LP mode, including LPDT, ULPS, Trigger, etc. The following is a reading process description

Read: Read register 0x04

BTA: Exchange D0N, D0P control to the Slave end

Read Response: Return 3Byte parameter 0x40,0

BTA: MIPI protocol control switching process

<>The MIPI speeds of the master and slave must match. Both the master and slave are between 6MHz and 10MHz. Generally, the Master can set the speed to match the Slave. Otherwise, there may be a risk of read failure.

<>The following processes and time intervals must be strictly followed. Generally, the Tta-sure+Tta-get time is measured to ensure that the switching time is sufficient. The time requirements are provided by the slave/supplier end;

TLPX=80ns

Tta-sured+Tta-get=620ns

2. MIPI LP->HS signal measurement

When switching from LP to HS mode, the slave will generally give a time requirement, and the time can be set according to the specification provided by the supplier;

Sampling rate: 20GS/s

Trigger condition: edge trigger

Trigger mode: Auto/Normal

**- CLK Lane:

Tclk-prepare:

Tclk-prepare+zero:

Tclk-trail:

- DATA Lane:

Ths-prepare:

Ths-preapare+zero:

Ths-trail:

Summarize

That’s all. You can follow a similar process to measure other signals. Adjust the sampling rate and sampling depth according to the length of the signal time to be displayed, and select the trigger condition and trigger type according to the type of signal to be captured.