FP6291 Design Issues

bluefox0919

I have developed a board power supply using FP6291. After about a month of testing, the following situation occurred. 1. Recently, I suddenly found that I2C could not be read. After checking, I found that the problem was with the power supply at the source. The input power is 5V 1A provided by micro USB 2. FP6291 inputs 5V and hopes to provide 12V. There have been no problems for a period of time. However, recently, the input is still 5V, but the output is also 5V, which causes abnormalities in the isolation circuit behind and prevents I2C from functioning normally. 3. The actual output current is only about 0.3A 4. I hope everyone can help confirm whether there is a problem with the FP6291 design? Thank you. fp6291v070.pdf (356.16 KB, downloads: 4)

2018-9-4 11:56 上传

点击文件名下载附件

Arvinˇ

Judging from the circuit schematic, there is no problem. The main thing is whether the 4.7uH inductor is suitable, whether it is saturated, and whether the DC internal resistance is too large. You can check and analyze from this aspect.

bluefox0919

Arvinˇ posted on 2018-9-4 11:35 Judging from the circuit schematic, there is no problem. The main question is whether the 4.7uH inductor is suitable, is it saturated? Is the DC internal resistance too large? ...

The inductor value is the one recommended in the document. The output current is about 0.3A, and OCP should be set at 1A. I suspect that the OVP point is too close. The upper limit in the document is 16V, but I used 12V. It is a rare situation, and it is difficult to reproduce it.

topwon

This also depends on the layout of the PCB.

topwon

In addition, it is recommended to measure the chip temperature to see if there is a problem that causes overheating protection. The package of this chip is not large, and the heat dissipation capacity is limited (especially if the PCB layout is not well treated for heat dissipation). If you use a relatively high output voltage, the output current should be reduced accordingly, so that the output power will not be too large, causing the Pd to increase and the junction temperature to be too high and protection. The thermal protection action of the chip is to turn off the internal MOS tube, so that 5V is directly added to the output through the inductor, so the output becomes 5V.

qwqwqw2088

What the above person said makes sense. Is it caused by the change of chip temperature during use? You can increase the heat dissipation measures and monitor it again. There has been no problem for a period of time. Or it may be caused by the change of load. According to the manual, the frequency of FP6291 is 1MHz. Or check whether there is potential interference in the PCB.

bluefox0919

topwon posted on 2018-9-6 09:22 In addition, it is recommended to measure the chip temperature to see if there is a problem causing overheating protection. The package of this chip is not large, and its heat dissipation capacity is limited (especially if...

1. The temperature OTP protection document seems to say that the protection will start at 150 degrees, and then the MOSFET will be turned on again after the temperature drops by 30 degrees? FP6291 will turn off the power MOSFET automatically when the internal junction temperature is over 150°C. The power MOSFET wake up when the junction temperature drops 30°C under the OTP threshold temperature. 2. I just measured it at about 35 degrees (the test environment is a bare board without a casing, in an air-conditioned 25-degree environment) 3. For the Layout part, I will compare the suggestions in the document to see if there is a big difference 4. I have lowered the output voltage to about 11.2V, a little further away from the 16V protection point, and will wait for the test results.

bluefox0919

qwqwqw2088 posted on 2018-9-6 11:24 What the above person said makes sense. Is it caused by the change of chip temperature during use? You can increase the heat dissipation measures and monitor it again because it has been used for a while and nothing happened...

1. No obvious and drastic changes in temperature have been found so far. The air conditioner will be turned off after get off work, so the temperature difference should not be so high as to affect OTP 2. Layout partly because the product itself is controlled by the off-the-shelf wireless module RASPBERRY PI, which is directly 5V USB power supply, FP6291 itself has no contact with the module, not sure if this will also have an impact.

qwqwqw2088

bluefox0919 posted on 2018-9-7 09:46 1. There is no obvious and drastic change in temperature. The air conditioner will be turned off after get off work, and the temperature difference should not be so high as to affect OTP 2...

At least three board tests are needed to judge such questions. There will be interference, not necessarily in this form. Replace FP6291 to see if the chip itself is unstable.

bluefox0919

qwqwqw2088 posted on 2018-9-7 10:08 It takes at least three boards to test to judge such doubts. Interference will occur, not necessarily in this form. Replace FP6291...

1. Currently, there are 3 boards that are hand-welded, and one of them has such a problem. I doubt whether it is a problem with the hand-welded parts. If I ask the factory to make three samples, will this situation still occur? 2. Replace FP6291. Is it because the chip itself is not stable? Does it mean to replace a new 6291?

qwqwqw2088

bluefox0919 posted on 2018-9-7 10:29 1. Currently I have hand-welded 3 boards, and one of them has this problem. I suspect it is a problem with the hand-welded parts. If I ask the factory to make three samples, is it...

"Currently I have hand-welded 3 boards, and one of them has this problem" Are the other two normal?

bluefox0919

qwqwqw2088 posted on 2018-9-7 10:35 "Currently I have 3 boards soldered by hand, one of them has this problem" Are the other two normal?

The other two are still normal

qwqwqw2088

bluefox0919 posted on 2018-9-7 11:24 The other two are still normal

This is an individual case, which means that the initial design problem is not very big. You can analyze this problematic board separately. It is recommended that you can compare and measure the difference between the good board and the problematic board if conditions permit, find out the problematic device and analyze the cause of the abnormality.

Arvinˇ

Arvinˇ posted on 2018-9-4 11:35 Judging from the circuit schematic, there is no problem. The main thing is whether the 4.7uH inductor is suitable, is it saturated? Is the DC internal resistance too large? ...

The output current is just the average value. If conditions permit, you can use a current clamp to confirm the current waveform of the inductor (or connect a small resistor in series at the inductor and use a differential probe to detect the voltage across the resistor) to see if the peak part has a tendency to rise. It may be close to the saturation point. Although it is only a possibility, it is also necessary to confirm and exclude this possibility.

bluefox0919

qwqwqw2088 posted on 2018-9-7 12:01 This is an individual case, which means the initial design problem is not very big. This problem board can be analyzed separately. If conditions permit, it is recommended to compare and measure the good board with the...

The embarrassing point at present is that I can't reproduce this situation. It is an occasional situation. The software test suddenly said that I2C could not be read. After re-powering, the functions returned to normal. I have tried to swap the three 6291s, and they are all normal now.

bluefox0919

Arvinˇ posted on 2018-9-7 12:02 The output current is just the average value. If conditions permit, you can use a current clamp to confirm the current waveform of the inductor (or connect a small resistor in series at the inductor and use a differential probe...

The current is measured by using a current meter after cutting the wire and putting it on the instrument to read the value. So 0.3~0.4A should be the maximum current consumed by my load. At present, the saturation point is set at 1A according to the file formula.

qwqwqw2088

bluefox0919 posted on 2018-9-7 16:18 The embarrassing thing at present is that I can't reproduce this situation. It is an occasional situation. The software test suddenly said that I2C could not read. After re-powering, all functions were restored...

There are still few boards for debugging and testing. For example, there are 3 boards like this. When 2 boards are normal and 1 board is abnormal, it is necessary to focus on the welding, assembly, and use of the abnormal board. Interchanging component chips is one of the methods. It is normal. It is possible that the board had welding problems such as cold soldering.

eward250

Can you find an agent to help you eliminate it? Shenzhen Yaxin is the agent of Yuanxiang.

hau30729

I would like to ask how to solve this problem later? I also encountered the same situation.
I use USB power supply, input Vin 5V and output 12V. It is normal at the beginning, but after a period of use, it will become 5V input and 12V output. I have encountered this problem on more than one chip.
Replacing the chip can solve it at first, but some will return to the same problem after a while... Some are fine just like that (or maybe there is no problem yet?)