Paid answers to questions about lm5117

electricor

Paid answers to questions about lm5117 [Copy link]

 

Input 16v, output 5v, current reaches 3A, 5V is sufficient, 3A is not reached,

whuer

I have used this chip before, not to mention 3A, 30A is no problem. Your way is totally unacceptable, you need to make the following modifications: First, there is an iron sheet on the belly of the chip, this sheet should be in contact with a large area of the ground to dissipate heat, otherwise the chip will overheat and be protected. I suggest you simply draw a PCB and send it to the factory for 30 yuan, 10CM long and 10CM wide

whuer

The inductor is changed to a shielded inductor

whuer

MOS is changed to MOS with very small RDSON, and can be connected in parallel

electricor

whuer posted on 2018-8-20 06:00 I have used this chip before, not to mention 3A, 30A is no problem, your way is totally unacceptable, you need to make the following modifications: First, there is iron sheet on the belly of the chip, ...

Now I don't have time to draw the PCB, so I want to ask how to modify it in the case of welding

electricor

whuer posted on 2018-8-20 06:00 I have used this chip before, not to mention 3A, 30A is no problem. Your way is totally unacceptable, you need to make the following modifications: First, there is iron sheet on the belly of the chip, ...

but there is a base under the chip, there is no way to connect the inner iron sheet to the ground

whuer

electricor Published on 2018-8-20 08:09 But there is a base under the chip, and there is no way to connect the inner iron sheet to the ground

This is the problem. The chip cannot dissipate heat, so it has thermal protection, so the current cannot go up

PowerAnts

whuer posted on 2018-8-20 19:51 This is the problem. The chip cannot dissipate heat, so it has thermal protection, so the current cannot go up

There are two external MOS, and the chip seems to be just a controller, so how can it overheat?

whuer

PowerAnts posted on 2018-8-20 20:43 There are two external MOS, and the chip seems to be just a controller, how can it overheat?

Haha, according to what you said, why did TI design the bottom of the chip to be iron? Haha, are you bored?

electricor

whuer posted on 2018-8-20 19:51 This is the problem. The chip cannot dissipate heat, so it has thermal protection, so the current cannot go up

Is it okay to connect the PGND and AGND pins on the chip and all the ground connections in the entire circuit to this large piece of solder?

qwqwqw2088

electricor Published on 2018-8-20 21:34 Is it okay to do this, connect the PGND and AGND pins on the chip, and all the ground connections in the entire circuit to this large piece of solder here

The feedback line is too long, which is a problem. Other reasons are difficult to find. It is not good to pile up tin like this. I still suspect that there is a problem with the soldering of the perforated board. In addition, the original poster’s post said that it is not simulation. Can you compare the waveform of SW during simulation with the waveform measured by the actual oscilloscope?

PowerAnts

whuer posted on 2018-8-20 20:51 Haha, according to what you said, why did TI design the bottom of the chip to be iron sheet? Haha, are you bored?

The EP pin certainly has the function of heat dissipation, and its function is to reduce the thermal resistance by about 10 degrees/W. From the operating voltage limit of LM5117 and the operating voltage of the original poster, it can be seen that the problem here is definitely not because the chip is overheated. The original poster has been unwilling to provide a waveform diagram. . .