# [Technical master test notes series]4: Multimeter measures the transient recovery of power supply

Simply put, the transient recovery time is the length of time required for the power supply to recover to the set level after the load is applied. People who don’t design power supply generally take it for granted. No matter what kind of circuit we use in our work, we may only notice it when the performance of the power supply deteriorates to affect our work. The power supply is a bit like a person, if you ask more questions than he knows, then he may not meet your expectations.

Author: Tektronix Technology Expert Josh Brown

[Technical master test notes series]Fourth: Use a digital multimeter to measure the transient recovery time of the power supply

Simply put, the transient recovery time is the length of time required for the power supply to recover to the set level after the load is applied. People who don’t design power supply generally take it for granted. No matter what kind of circuit we use in our work, we may only notice it when the performance of the power supply deteriorates to affect our work. The power supply is a bit like a person, if you ask more questions than he knows, then he may not meet your expectations.

We use an analogy to help illustrate the importance of considering power supply pressure. If someone throws you a baseball, tell you to keep the ball at the same height when you catch it, move forward or retract a few inches (or a few centimeters), and finish it in one second. The baseball (load) is relatively light, and your hands hardly move when you grab it. (Of course the premise is that you have an accurate pitcher.)

Now imagine that someone throws you a bowling ball and imposes the same restriction on your reaction. Not only must your hands move, but your body must also move to assist, because the weight has changed. Does the ball cause the height you maintain to change relative to the target height? How big is the change? Did you fully grasp and hold the ball? If you can’t keep it within a certain height tolerance, can you stabilize the height within the allowed 1-second interval? Obviously, this challenge is much greater (unless you are an NFL forward). Similarly, if the load placed on the power supply exceeds its processing capacity, it will be difficult to meet your requirements.

So how do you know if you are not far? At this time, the use of DMM to fully characterize the characteristics of transient recovery becomes the key.
Before going into the specific details, let’s discuss the switching frequency first, because this helps determine which instrument you choose to evaluate the problematic behavior. Although both linear power supplies and switching power supplies are affected by transient recovery time indicators, switching power supplies work at a predetermined frequency and naturally have more noise and more stray signal activity. The frequency range used by switching power supplies is generally 10 kHz to 1 MHz. Because the oscilloscope has a high sampling rate and provides excellent visual tools, the oscilloscope is particularly suitable for evaluating transient recovery time, where the level trigger is used to capture the event, and the screen cursor is used for math and measurement analysis.

You might also consider a digital multimeter with high-speed sampling capability, which might already be on the workbench. Some DMM models have been able to sample up to 50 kHz many years ago, and newer models can support up to 1 MHz. With speed comes more opportunities for high-speed sampling, because its measurement sensitivity is higher than most oscilloscopes.

To illustrate how to measure the transient recovery time, we use the LM25088MH-1EVL evaluation circuit board produced by Texas Instruments, which can provide 5V voltage and up to 7A current from an input power source up to 36V. After making slight changes, we reconfigured it to switch at 25 kHz. We use two channels of Keithley 2230A-30-3 three-channel DC power supply to provide 10V input signal (maximum 6A current) to power the evaluation circuit board. We also use Keithley 2380-500-15 Electronic load to absorb a certain amount of current. In this case, we want to push the LM25088MH to its 7A limit to see how it reacts. We use DMM6500 to evaluate the voltage response of the electronic load when it absorbs a large current, and then check the current response.

To evaluate the transient recovery voltage, we connect the instrument as follows:

After setting the 2230A power supply to output 10V @ 3A on the two output channels (to ensure sufficient current for power conversion), we set the 2380 electronic load to absorb a constant current of 7A. We turn on the output of 2230A and turn off 2380 at the same time. We configure the DMM6500 as a digitized voltage, set the sampling rate to 1 MS/s, and the number of readings to 10k, and then touch the Set Up Trigger button, the waveform capture option appears, select Analog Edge from the Source Event option.

Since we expect that the voltage will drop to a certain extent when the load is applied, we set the level to 4.925 V and the slope to the falling edge. After selection, we can use the Position option to adjust the horizontal representation of the waveform on the graph. In order to start the DMM6500 to start monitoring events, we touch the CONT annunciator in the upper left corner of the screen, and a drop-down menu will appear with multiple trigger options. Choose Initiate Trigger Model.

Finally, we turn on the 2380 electronic load, apply the effective step function, and absorb 7A current. DMM6500 will capture the sampled waveform and Display it in the Graph chart column. This will show the transient response when a load is applied to the output of the LM25088MH switching power supply. Tap the DMM6500 screen to zoom in on the area of ​​interest for further evaluation. Turn on the cursor to get the signal level (y-axis information) over time (x-axis information). We can see that the switching power supply stabilizes after about 1.5 milliseconds. What’s more interesting is that there is a big sag after the load is applied, and the driving output voltage reaches 3.466 V at the lowest.

Just as important as transient recovery when a load is applied is what happens when the same load is removed. To capture this situation, we touched the Trigger label on the DMM6500 and changed the level to 5.1 V and the slope to the rising edge of Rising. You can trigger the DMM as before (using the drop-down menu option), or press the TRIGGER key on the left side of the DMM Display on the front panel of the instrument. Then turn off the 2380 electronic load, you can see the recovery waveform of the switching power supply on the DMM6500 chart.

We don’t need to enable the horizontal cursor and the vertical cursor on the display at the same time, but only use the vertical cursor, slide the information bar at the bottom of the screen, and VCursor Stats appears. This will provide basic statistics on the data points between the two vertical bars. Therefore, we do not need to enable and adjust the horizontal cursors to determine that the peak value of the switching power supply is 6.953 V when the load is removed. We can set the right-hand cursor to the place where the small sag occurs before the output becomes flat, and find that the recovery response time is about 125 milliseconds. We can also zoom in on the peak point area to better see how the voltage signal behaves.

Now we turn our attention to the current signal, using the same circuit, but the DMM6500 is connected in series with the 2380 electronic load. Pay attention to the record, because we are dealing with a large current (7 A), we need to use the 10 A current input connection provided on the back of the DMM.

The trigger option setting is similar to before, but now we trigger at the midpoint of the expected 7 A signal, and the slope is still set to the rising edge of Rising.

We trigger the DMM6500 again (through the drop-down menu), and then turn on the 2380 electronic load. The response of the DMM to the waveform event is no problem. This process was repeated when the load was removed. We also used the screen cursor to quickly analyze the data and showed that the current dropped to -2.5 A.

Although our focus has been on high-power waveform capture, we still have to pay attention to the current sensitivity advantage provided by digital multimeters. Many devices currently designed must be as energy-efficient as possible to extend battery life. This switch circuit of the LM25088MH evaluation circuit board can be integrated into a low-power wireless design. When transitioning to different states (sleep, standby, sensing, processing, etc.), the weak current absorbed by each component must be evaluated. The process of capturing low current waveform events is the same as the process listed above, but the lower range allows users to monitor low currents. For example, if we set the load to 5 mA and use the 10 mA range on the DMM6500, then we can view microampere readings.

I hope this note will give you an understanding of transient recovery measurements and experience some of Keithley’s products. Keithley has more than 500 products that can be used to provide, measure, connect, control or transmit direct current (DC) or pulsed electrical signals.