Boost DCDC for constant current LED driving, inductor heating problem

S3S4S5S6

Boost DCDC for constant current LED driving, inductor heating problem [Copy link]

 

The chip is XL6006. According to the typical application circuit design, we want to boost 24V to 36V, output current 1A, load current more than 900mA. After the circuit design is completed, the inductor heats up quite a lot. The inductor rating is 150uH. The inductor parameters are as shown below.

The switching frequency of XL6006 is 180K

The design principle refers to the official design guidance

XL60XX系列升压恒流产品设计指南.pdf (622.82 KB, downloads: 8)

2024-6-17 20:47 上传

点击文件名下载附件

tagetage

message TEXTAREA textarea

You described your problem very clearly, and I have read your information. I have a small question.

1. The inductor of XL6006 is usually 47uF, why is yours 150uF?

2. The manufacturer of the inductor is fine, it is a big manufacturer, but is the type of inductor correct? ?

I'll send you two PDFs with the temperature rise data of the inductor corresponding to this chip. See if it helps you. You said your inductor is hot, how many degrees is it? Have you measured it?

针对投光灯和泛光灯的双节锂电池供电方案.pdf (288.93 KB, downloads: 3)

2024-6-17 21:50 上传

点击文件名下载附件

XL6006 演示板.pdf (135.02 KB, downloads: 1)

2024-6-17 21:50 上传

点击文件名下载附件

qwqwqw2088

Try changing the inductor

The larger the inductance, the larger the inductor current ripple, which increases the loss and heat

S3S4S5S6

tagetage posted on 2024-6-17 21:51 message TEXTAREA textarea You described the problem very clearly, and I also read your information. I have a little doubt. 1. XL6006 is generally equipped with...

(1) The selection of inductance is based on the calculation method provided by it, as shown in the figure below

According to this calculation, the minimum value of the input voltage is 22V, so the calculated inductance is about 100uH. Considering that the error of the general inductance is 20%, and the inductance value will decrease with the increase of current in practice, the inductance is taken as 150uH.

(2) The power inductor selected is 150uH with a rated current of 3A and an accuracy of 20%

S3S4S5S6

qwqwqw2088 posted on 2024-6-18 08:31 Try changing the inductor. A larger inductor will result in a larger inductor current ripple, which will increase losses and heat up.

Should the inductor be reduced? Isn't the heating of the inductor related to the internal resistance of the inductor? The internal resistance of the inductor I chose is 130mΩ

S3S4S5S6

qwqwqw2088 posted on 2024-6-18 08:31 Try changing the inductor. A larger inductor will result in a larger inductor current ripple, which will increase losses and heat up.

I followed the calculation on the 4th floor to get the minimum value of the inductance, and then increased the margin appropriately.

tagetage

I think your inductor power is too small and the inductor reactance is too large. The internal resistance of the inductor is 130 milliohms, which is still too large.

tagetage

I checked and the type of inductor you chose should be fine.

If it doesn't work for you, try changing to a larger one with a higher current, smaller inductive reactance and smaller internal resistance.

Yijiayuan YSPI one-piece molded inductor (flat coil one-piece molded high current high-quality power inductor/
one-piece molded chip shielded high-power and high-current resistant chip inductor):

YSPI series one-piece molded inductors (power inductors, high current inductors, automotive grade inductors)

YSPI Series Power Inductors

Main technical indicators of integrated molded inductor

1. Electrical characteristics:

Inductance range: 0.1uH~100uH (tolerance ±J/K/L/M)

Rated current range: 0.2A~80A (IDC)

Dielectric strength between coil and alloy powder: ≥300V (DC/1mA)

Temperature rise: ≤40℃

2. Mechanical properties

The results of impact, drop, vibration, centrifugal and pressure resistance tests show that they are better than traditional chip power inductors.

3. Environmental characteristics:

Storage temperature: -55℃+125℃

Working temperature: -22℃+85℃

Product life: ≥60000H

One-piece molded inductor Features:

Surface Mount

Low DC resistance and high current

Excellent saturation current, low DCR, low loss

Inductors for filtering, PFC boost, DC-DC converter

tagetage

If you have any news, please reply here, what happened? How did you solve it?

S3S4S5S6

tagetage posted on 2024-6-19 12:39 If you have any news, please reply here. What happened? How did you solve it?

According to what qwqwqw2088 said, and combined with your suggestion of choosing a low-impedance inductor, the heating is normal now, and it feels about 40 to 50 degrees by hand, so it shouldn't be a big problem.

Another disadvantage of this chip is that if there is no load connected, the output voltage will continue to increase until it reaches the 60V limit inside the chip.

tagetage

S3S4S5S6 Published on 2024-6-19 14:24 According to what qwqwqw2088 said, and combined with your mention of choosing a low-impedance inductor, the heat is now normal, and it feels about 40 to 50 degrees by hand, so it should be...

It's good that the original poster solved the problem. It seems that the main problem is that the inductor's inductive reactance is too large, and the temperature rise is caused by the large internal resistance.

Another bad thing about this chip is that if there is no load connected, the output voltage will continue to rise until it reaches the 60V limit inside the chip. This is not because the chip is bad, but because the output is a constant current output. If there is no load, it will continue to rise to the overvoltage protection level. This is a normal chip designed like this, and it is normal as long as the chip is not damaged without a load.

S3S4S5S6

tagetage posted on 2024-6-19 16:39 It would be great if the OP could solve the problem. It seems that the main problem is the high inductance and high internal resistance which cause the temperature to rise. In addition, this chip has a bad...

(1) After remaking the board and testing it, the inductor howled for the first few seconds after power-on. During this process, the waveform on the SW pin of the XL6006 chip jittered violently. As time went by, the waveform gradually became clearer and finally reached a stable state. The waveform measured with an oscilloscope was the yellow waveform in the figure below. Why did the inductor howl? Is it because during this process, the PWM duty cycle has been constantly adjusted, causing the inductor to howl?

(2) Then measure the voltage waveform of the sampling resistor on the current sense resistor. The blue waveform in the figure above has a voltage of 220mV, but there will be large glitches in the switching cycle. How to improve this glitches? The figure below is an amplification of the blue overshoot waveform.

S3S4S5S6

tagetage posted on 2024-6-19 16:39 It would be great if the OP could solve the problem. It seems that the main problem is the high inductance and high internal resistance which cause the temperature to rise. In addition, this chip has a bad...

video

Waveform changes from power on to stability

tagetage

After remaking the board and testing it, there was an inductor howling for the first few seconds after power-on. ----------- There was a howling in the first few seconds, but it stopped after a few seconds??? Do you have a load? If so, is there any load change???

Then measure the voltage waveform of the sampling resistor on the current-sense resistor-----------Not all places on the circuit board can be directly observed with an oscilloscope, because the input resistance of your oscilloscope is not large enough. Direct measurement in some places will cause the circuit working state to shift, and the sampling resistor is one of them. If the input resistance of your oscilloscope is not high enough, when the probe points to the sampling resistor, the oscilloscope is equivalent to a resistor connected in parallel to the sampling resistor, so the duty cycle of the chip PWM changes.

tagetage

http://www.xlsemi.com/datasheet/XL6006-CN.pdf

http://www.xlsemi.com/demo/XL6006-DEMO.pdf

Have you read these two documents carefully? What kind of inductor do you use?

S3S4S5S6

tagetage posted on 2024-7-3 18:28 After remaking the board and testing it, there was a few seconds of inductor howling at the beginning of power-on. ----------- There was a howling in the first few seconds, but after a few seconds, there was no howling? ? ...

First connect the load, then power on. The waveform during the whole process is as shown in the video on the 13th floor

The probe impedance of the oscilloscope is 1M, and connecting it in parallel with 220mΩ has no effect on the resistance value.

tagetage

There was a howling sound in the first few seconds, but it stopped after a few seconds. My first impression of this phenomenon was that it was overloaded. What is the maximum output current of your circuit and how much current it uses when it is working?

If the oscilloscope does not measure the sampling resistor, will the power-on effect be the same? ? ?

Did you measure the voltage waveform at the input of the chip when it is powered on?

S3S4S5S6

tagetage posted on 2024-7-4 10:00 There was a howl in the first few seconds, and then there was no howl after a few seconds-------This phenomenon gave me the first feeling that it was overloaded. What is the maximum output of your circuit...

(1) The maximum output current of this circuit is within 1.1A, the sampling resistor is 220mΩ, the load is 24 parallel LED strings, each string current is 40mA, the total current is 960mA, the chip boost target voltage is 36V, and the chip input voltage is 24V

(2) "If the oscilloscope does not measure the sampling resistor, will the power-on effect be the same? " The effect is the same

(3) "Did you measure the voltage waveform at the input of the chip when it was powered on?" I haven't measured it, but I adjusted the voltage regulator to 24V output with unlimited current. I connected the load to this board, and then connected the 24V of the voltage regulator. After turning on the voltage regulator, the voltage regulator dashboard showed a voltage of more than 9V and a current of more than 4A. The inductor began to howl, and the switching waveform on the chip SW pin began to adjust. When the SW waveform stabilized, the inductor stopped howling (in fact, the inductor stopped howling before SW stabilized. The total length of the video on the 13th floor is 40s, and the inductor howled for about 20s). At this time, the voltage regulator dashboard showed a voltage of 24V and a current of 1.53A. The process from 9V/4A to 24V/1.5A is an instantaneous jump. When the chip SW pin is stable, that is the moment when the value on the voltage regulator dashboard jumps. No other voltage appears on the voltage regulator dashboard in the middle.

tagetage

The voltage on the power supply dashboard shows more than 9V and more than 4A, and the inductor starts to howl---------this is why I ask you. The input is 9V, and the inductor is overloaded, so why doesn't it howl? ? ? ? Don't use a power supply when testing, use the 24V power supply you actually use to test.

tagetage

If your voltage regulator cannot directly power up to 24V, it means that the power is insufficient, or the transient response is insufficient.