LM5116: When the LM5116 is designed, the output temperature of the two MOS tubes is high, and the SW has large and small waves.

Hello Hao 

Would you please let us know more information ? I need 

• Temperature (Ambient, MOSFET, Inductor, IC) 
• Test condition (VIN, VOUT, IOUT, FSW) 

Also, please capture the large and small waves and send them to me. 

-EL

1.The temperature is mainly concentrated on the copper sheet near the MOS tube, and the temperature is about 100 degrees

2.The size of the wave is shown below

3.How to eliminate the spike shock in the above picture

When I output 24v5a fully loaded, the temperature reached 130 degrees Celsius, is this temperature abnormal? In addition, what is the highest degree of 24V5A fully loaded when you test it on the demo board?

thank you

I modified the circuit into 12V5A output for testing, and the frequency was modified to 150KHZ (the frequency reduction was to reduce the temperature of the PCB board, and the above 24V5A output was also modified), and the relevant peripheral parameters were also modified. The frequency changes to 78KHZ at 12V5A output, and the frequency is normal at 48W. I tried to change the Rramp(768K) and Cramp(680pf) values and the result is that the frequency is now 268KHz at 12V5A output,Not sure what caused the problem.

Hello Hao

I guess you are seeing DCM operation / subharmonic oscillation. Please try to measure the inductor current together with SW node voltage. 

-EL 

During my use, when the load loaded 120W, the chip could not work normally once the power was turned on, and the power indicator at the load side was off

Hello Hao 

Please send me some waveforms for my investigation, and especially check the temperatures of external components when the output is in regulation.

-EL

1.During the operation of LM5116 chip, I found my input voltage waveform as shown in the figure

The burr waveform above is 48V power input, and the frequency of the spike at this time is consistent with the frequency set by my LM5116 chip, and the spike also increases with the increase of the output load. How can this spike be eliminated?

2.Can the 17 pin of the LM5116 chip be suspended? At present, the 17 pin of my chip is converted from 48V DCDC chip to 12V to supply power to LM5116 chip. Similarly, when the load of LM5116 chip increases, the DC12V also has the same spike. When the load is full, the spike above 12V is too large, causing the LM5116 chip to burn. Is it feasible that I do not want to use the 17-pin of the LM5116 chip at present?

When the load is full, the chip will not work, and the SW has no waveform. I set the chip operating frequency to 150KZ or so and the inductance is 47UH. What causes this?

Hello Hao

How can this spike be eliminated ==> It is not possible to totally eliminate the switching noise. It can be reduced by adding a gate resistor. 

When the load is full, the spike above 12V is too large, causing the LM5116 chip to burn ==> Would you please explain me the reason why you think the 12V power supply  damages the LM5116 device ? 

Is it feasible that I do not want to use the 17-pin of the LM5116 chip at present ==> It is okay not to use VCCX. 

-EL