Five things I wish I had known before starting double-pulse measurements [Part 1: What I learned while setting up the measurement environment]

Narrow down by specifying conditions

現在2215件がヒットしています。check

Design Analog Discrete Power Supply Power Devices ON Semiconductor

Introduction

When you start evaluating SiC MOSFETs, you will almost certainly encounter the "double pulse test." (Double Pulse Test) You will come across the phrase "."
I myself
"By preparing an evaluation board and observing the waveform with an oscilloscope, the turn-on loss can be measured." (Eon) or turn-off loss (Eoff) It should be easy to measure."
That's what I was thinking.
However, once we actually started building the measurement environment, we realized that our initial assumption was completely wrong.
Before starting the measurement,
Which devices should we evaluate?
Eon and Eoff How much of a difference is there?
Which is better, ON Semiconductor's product or a competitor's product?
I was interested in things like that.

However, in reality, the power device being evaluated is different from the power device being evaluated.
What kind of probe should be used?
Why is skew adjustment necessary?
Is the current really being measured correctly?
Are the oscilloscope settings correct?
We ended up spending a lot of time on these kinds of issues related to the measurement environment.

The board used in this project was a double-pulse evaluation board that had been meticulously crafted through trial and error. This allowed me to concentrate on mastering measurement techniques. However, as I progressed with the evaluation, I increasingly encountered situations where I wondered, "Why this configuration?" and "Couldn't this be improved?" After speaking with experienced designers, I learned that the double-pulse evaluation board is built upon many trade-offs, and that the design decisions themselves—what to prioritize—are a matter of know-how.

What I learned from this experience is that
"The difficulty with double-pulse measurement lies not so much in the measurement itself, but in establishing the correct measurement environment."
That's the point.

In this article, I will introduce five points that I wish I had known first when setting up my environment, for those who are about to start double pulse measurement.
1 What is a double pulse test?
2 Why is a datasheet alone insufficient?
3 First, understand the entire measurement system.
4 Probe selection is more important than you might think.
5 If you don't know how to adjust the skew Eon/Eoff Change

If I could give this article to myself six months ago, "Read this before you watch Eon/Eoff." I would like to say that.

Figure 1. Double pulse measurement environment used in this study.

① Oscilloscope, ② Pulse generator, ③ Acrylic case with safety mechanism, ④ Evaluation board for double pulse testing.
⑤ DC link capacitor, ⑥ Emergency stop button, ⑦ High-voltage DC power supply

What is a double pulse test?

When I first started learning about double pulse testing, Eon or Eoff My understanding was limited to "a test to measure [something]".

While that's certainly not wrong, it doesn't fully convey the value of double pulse testing.

The double pulse test is a typical test method for reproducing the switching operation of power devices under arbitrary voltage and current conditions and evaluating the losses and transient responses that occur during this process.
Generally speaking,
Eon: Turn-on loss
Eoff: Turn-off loss
・dv/dt: Speed of change of drain voltage
• di/dt: Speed of change of drain current
• Qrr: Reverse recovery characteristic of the body diode
It is used to evaluate things like these.
What's important here is,
This is not a test to verify values listed in the datasheet, but rather a test that allows us to check the device's behavior under conditions close to actual usage.
for example,
- Change the DC bus voltage
- Change the load current
- Change the gate resistance
This can result in completely different waveforms even with the same device.

The double pulse test has significant value because it allows you to visualize the difference as an actual waveform.
The measurement circuit itself is relatively simple.
Basically,
・DC power supply
• Inductor load
• Device under measurement
Gate driver
It consists of [this].

The first pulse supplies the desired current to the inductor, and the subsequent pulses are used to observe the switching waveform.
This allows you to evaluate the turn-on and turn-off behavior of a device under arbitrary voltage and current conditions.

Before I started the evaluation process
"The circuit looks simple, so the measurement should be easy too."
That's what I thought.

However, in reality, rather than the circuit itself,
How to measure the waveform
How to reduce measurement errors
How correct Eon/Eoff How to get
This was the more difficult part, and this was where the real starting point for double-pulse measurement began.

Figure 2 Basic configuration of a double pulse test with the low-side device being measured.

The first pulse supplies a desired current to the load inductor, and at the end of the first pulse, the turn-off characteristics are determined.
At the start of the second pulse, the turn-on characteristics and the reverse recovery characteristics of the upper MOSFET 's body diode are evaluated.

Even a completed evaluation board reflects the designer's intentions.

The double-pulse evaluation board used this time was a highly refined board that had been meticulously crafted through trial and error.
Initially, my impression was simply, "This is a well-designed evaluation board," but as I continued the evaluation process, my perspective changed.
for example,
- We want to observe more accurate waveforms.
- We want to reduce the amount of parasitic components.
We want to support a variety of devices.
We want to secure measurement points.
We also want to maintain ease of use.
Such demands are not always compatible.
As we went through the evaluation process, there were times when I thought, "Wouldn't it be better to do it this way?"

However, after speaking with experienced designers, I learned that those improvement proposals may not have been adopted due to trade-offs with other requirements.
In designing a double pulse evaluation board, The question isn't "What's best?" but "What should be prioritized?" It is important to decide on that.

I feel that the board we used this time was also the result of a series of such design decisions.
Furthermore, some people who are just starting out with double-pulse measurement may design their own evaluation boards.
On the other hand, if you want to focus on acquiring measurement techniques and evaluating devices first, using commercially available evaluation boards is a valid option. I myself have come to appreciate their value once again this time.

The evaluation board used in this study included a gate driver and an isolated power supply, as well as a device for checking the power supply status. LED The system incorporates mechanisms to support pre-measurement checks, and each of these features reflects the designer's experience.

Figure 3: Evaluation board "ALTDPEB" for double pulse testing.

To quickly set up an evaluation environment, utilizing off-the-shelf evaluation boards suited to the evaluation objectives is a viable option. The board used in this case had LEDs to check the status of each power supply system, allowing us to confirm that the 5V, 12V, high-side, and low-side power supplies were functioning correctly before measurement. This system allowed us to focus on considering measurement conditions and waveform observation rather than on circuit setup work.

This evaluation board can be purchased from the following website.
Double Pulse Evaluation Board

Why is a datasheet alone insufficient?

What impressed me most when I started learning about double pulse testing was,
"In the datasheet Eon or Eoff My question was, "Why bother measuring it myself when the information is already available?"

In fact, the datasheets for major power devices include:
Turn-on loss (Eon)
Turn-off loss (Eoff)
- Reverse recovery characteristic (Qrr)
Switching characteristics such as these are described.

Therefore, initially, I thought that "comparing the data sheets would be sufficient."
However, as I investigated further, I came to understand that the values listed in the datasheet are merely representative values measured under specific conditions.

for example,
DC bus voltage
・Load current
Gate resistor
• PCB layout
Measurement environment
If the parameters change, the results will vary significantly even with the same device.
especially SiC MOSFET In high-speed switching devices like these, the impact cannot be ignored.

Furthermore, there is information that cannot be gleaned from the datasheet.
for example,
- Overshoot during switching
Ringing
Actual voltage and current waveforms
Behavior under application-specific conditions
And so on.
This kind of information only becomes apparent when you actually observe the waveforms.

As I proceeded with the evaluation, what I felt was,
"The double pulse test is not a test to verify the accuracy of the datasheet, but rather a test to verify how the device behaves under your specific usage conditions."
That was the situation.
This is precisely what I believe is the greatest value of the double pulse test.

First, understand the entire measurement system.

With double-pulse measurement, it's not the case that "you can get results immediately just by connecting the measuring instrument."

When I set up the environment,
Pulse generator
·oscilloscope
VGS measurement probe
VDS measurement probe
• Current probe
It was necessary to handle both simultaneously.

And in reality, rather than the device itself,
We often encountered difficulties with "combining different measuring instruments."

Especially when dealing with high-speed switching. SiC In devices, the influence of the measurement system is very significant.
Therefore, before evaluating the device,
"Understanding the entire measurement system"
I feel that this is important.

Probe selection is more important than you might think.

The first thing that surprised me when I started taking measurements was that "not all probes are the same."

for example,
• Probe for high-voltage measurement
• Optically isolated probe
• Current probe
Then their characteristics are quite different.

Also, depending on the combination with an oscilloscope,
If the settings are not configured correctly, the current and voltage values may not be recognized as correct.
I myself,
"The waveform is displayed, so it's correct."
I had assumed that,
Upon later review, it was found that the scaling and probe settings were not appropriate in some cases.

In double-pulse measurement, the probe, as well as the device, is part of the measurement system.
Understanding the characteristics of the probe is essential for obtaining accurate evaluation results.

Figure 4. An example of a measurement system in a double pulse test.

In double-pulse testing, the device is evaluated while simultaneously observing VGS, VDS, and IDS.
The diagram shows the VGS probe, VDS probe, IDS measurement CT, and load inductor.
To reiterate, in double pulse testing, the entire measurement system, not just the device itself, significantly impacts the results.
Therefore, it is important to understand the characteristics and connection methods of each probe before conducting an evaluation.

If you don't know how to adjust the skew, Eon/Eoff will change.

While learning about double-pulse measurement, I felt that skew adjustment was the most important aspect.
Enon and Eoff VDS and IDS We find the answer by multiplying the two values.
In other words,
If the time axes of the voltage waveform and the current waveform are misaligned,
The calculation result itself will change.

Even while organizing the measurement data,
If the skew between probes is not adjusted correctly, Eon/Eoff The fact that it cannot be measured accurately was repeatedly emphasized.

When I myself hear the term "double pulse test"...
Turn-on waveform
Turn-off waveform
・Loss calculation
I was only paying attention to that.

However, in reality, the preceding stage is
"Creating a measurement environment to obtain accurate waveforms"
I realized how extremely important that is.

Figure 5: Comparison of waveforms before and after skew adjustment

What is skew?

Because each probe has different internal circuits and cable lengths, the timing of waveform arrival may vary slightly even when measuring the same phenomenon.
To accurately evaluateEon and Eoff, it is necessary to correct for the time difference (skew) between probes.

lastly

When you hear "double pulse measurement," you might think... Eon or Eoff We tend to focus on the measurement results themselves, such as those mentioned above.
However, when I actually started setting up the environment,
・Measuring instrument configuration
Probe selection
・Skew adjustment
We found that these fundamental aspects have a significant impact on the results.

In the second part,
・CT current measurement
Oscilloscope settings
• Sampling rate
・ADC resolution
Points to note when calculating Eon/Eoff
I'd like to share what I learned while actually setting up the environment.

Inquiry

If you are interested in the content introduced in this article, please feel free to contact us.

Onsemi Manufacturer information Top

If you want to go back to ONSEMI Manufacturer Information Top, please click below.

Back to Manufacturer Information Top

Site Search