Narrow down by specifying conditions
現在1890件がヒットしています。check
After working at a machine manufacturer for three and a half years, working on development work such as housing and mechanism design, I joined Macnica mid-career as an FAE.
Macnica has a large proportion of mid-career employees, so it's very easy to work here and you don't feel alienated at all.
Originally, I studied mechanics at university and had no knowledge of electricity at all when I joined the company, so I started over from Ohm's law and am currently studying desperately.
I had an opportunity to actually design a DC/DC converter under such circumstances, so I hoped it would be helpful for a beginner electrician like myself.
I will introduce the process.
This is the second draft. Please see the first draft here.
Now check the operation
Turn on the power to the board you created this time. The required specifications are below.
・Input voltage : 6~36V (TYP.12V)
・Output voltage : 5V
・Output current : 2A
Operation was confirmed by gradually increasing the input voltage from 6V with a load of 2A.
The output was 5V, but when I raised the input voltage to 16V, the device made a clicking sound and stopped outputting.
When I discussed this with a senior employee, he said that I shouldn't have suddenly checked with a full power load of 2A. First, the load placed on the device, such as no-load operation,
I received advice that we should start evaluating things from a small point.
At this time, I started operating the device casually to check its operation, so I did not collect any waveform data that would help me determine why it broke.
I realized that in the world of electricity, it is difficult to verify what was happening and what was broken after something broke.
That's why, as an electrician, I learned that it is important to get into the habit of always taking data such as waveforms during operation.
Checking the cause of damage
There was a high possibility that the device was broken, so I replaced the device with the conversion board.
The device was damaged when the input voltage was 16V and the output voltage was 5V/current 2A, so I suppressed the input voltage to 12V and the output voltage to 5V/current 1A and observed the waveform.
In addition to occasional irregularities in the switch (SW) cycle, the maximum spike voltage during SW reached 28V for an input voltage of 12V. The reference switch waveform in the data sheet is approximately +5V relative to the input voltage, so this is clearly too large.
If this waveform voltage is really being applied, then increasing the voltage/load current may cause a voltage exceeding the device's absolute rating of 42V to be applied. I focused on removing it.
Reference waveform of data sheet
Observed waveform ① VIN voltage ② VOUT voltage ③ FB pin voltage ④ SW pin voltage
Find the cause of SW waveform disturbances and spike noise
Why is the SW waveform distorted and why is spike noise observed?
The first thing I tried was changing the position of the input capacitor (CIN).
The data sheet states that the CIN should be placed as close to the device as possible, but due to the fact that the conversion board was interfering with it this time, it ended up being located far away.
I ported it to the device conversion board, although I can't deny that it felt a little forced.
Moving the CIN capacitor
remeasurement
Let's measure again.
What about the results? The peak voltage of the spike is 22V, which is reduced by about 6V compared to before the CIN move.
The disturbance in the SW waveform that was seen before the move disappeared and we were able to operate stably!
Observed waveform after CIN porting ① VIN voltage ② VOUT voltage ③ SW pin voltage ④ FB pin voltage
Why is it good to bring CIN close together?
Copper wires and patterns on circuits contain parasitic resistance and inductance components that can cause noise.
The input capacitor CIN has the role of removing these noises, but it becomes meaningless if it is far from the device.
The data sheet says to place them as close as possible, but I learned a lot about whether it is necessary to place them this close.
By moving CIN, the spike voltage improved somewhat, but it was still too large. If this continues, increasing the input voltage will likely violate the device's rating.
Next time, I'll dig deeper into improving spikes.
looking forward to.
Universal board edition article list
■Universal board edition
・ I made a DC/DC converter using a universal board (1)
・I tried making a DC/DC converter using a universal board (2)
・ I tried making a DC/DC converter using a universal board (3)
・ I made a DC/DC converter using a universal board (4)
■Printed circuit board edition
・ Create a power supply with your own printed circuit board! (1)
・ Create a power supply with your own printed circuit board! (2)