LM5149:   LM5149：LM5149 -40℃

Hi,

The first place would be to look at the Temperature Coefficient of Resistance (TCR) of the 5mOhm resistor that you are using. The resistance will change significantly based on the resistor that you have chosen. As the resistance changes, so will be the current limit set point of the circuit. 

You can also verify this by comparing/checking the voltage across the ISNS+ and the VOUT pins to check if the current limit threshold is being hit. 

Thanks,

Best Regards,

Taru

HI， Taru

It may not be for this reason. I just tried starting at 20 ° C and starting at 70 ° C with 10A load, which is OK, but at -40 ° C, 0A load does not cause any abnormality. When the load is increased, you can hear severe whistling, and when the current is increased to 10A, the voltage will jump between 1V-3V. May I ask which parameters I need to change or pay attention to in the low temperature environment, which will lead to the abnormal work of the chip

thanks，

best regards，

Hi，Tim

I added 200uf electrolytic capacitor to the output capacitance test the day before yesterday, but it did not get a good relief. Could you please help me check the parameters of quick-start? In addition, if I increase the value of Rs, the loop will be greatly improved. When I used 1mΩ, the circuit didn't even work.

thanks，

best regards，

7024.LM5149-LM25149 Quickstart Calculator_rev3.xlsm

I used a 110uf ceramic capacitor and a 200uf electrolytic capacitor, and the result was that the heavy load was severely howling.

Oh no, after my colleague tested it at a high temperature of 70 degrees and 15A, the circuit was abnormal due to the damage of the inductor, and now it is back to the origin. Even if I increase the inductor value and the larger ESR, it also becomes abnormal at low temperature. The current situation is that the output voltage between 4-9A will drop obviously at low temperature, but it does not exceed the range of PG. What is the problem? I added 1000uf solid-state capacitor, which is slightly changed, and the range of 4-9A will be reduced a little, but it is still abnormal. I have tried all methods, and those will cause abnormal output of medium-load at low temperature.Is it possible that the problem is on the PCB and why would the controller be so sensitive to the parameters of the output loop，I reduced the original 110uf output capacitor to 44uf, and the circuit did not even work at room temperature

thanks，

best regards，

Hello, Tim
After a day of trial yesterday, if I need to output 3.3V, I need to change the inductance from 2.2uh to 2.8uh (with a smaller esr) and the detection resistance to 4mΩ, which is stable at -40°, 10A. If I need to output 5V, I need to change 2.2uf to 3.3uf. Increase the output capacitance by at least 200uf, and do not use electrolytic capacitors (tantalum capacitors and solid-state capacitors are acceptable), so that it is stable at -40° and 20A. I am still adjusting these parameters, as well as the parameters of 12V output. I am curious why there is such a change at low temperature, because I can operate normally at normal temperature no matter how bad the parameters are used. Does TI have any good ideas? I did not find relevant information on the Internet.The following is my PCB layout and the quickstart of the original parameters. Sorry, we cannot contact the FAE of TI.Finally we need to get 20A for each output voltage,Thank you for your help.2402.LM5149-LM25149 Quickstart Calculator_rev3.xlsm

Thanks

best regards

all 2oz copper

Cout reminded me that I have now used 5 ceramic capacitors of 16V 1210 22uf and 2 polymer tantalum capacitors of 25V 100uf, so the inductance value of 1.1uH needs to be increased when the output of 5V and 3.3V, and the inductance value of 12V needs to be increased by nearly 4uH to stabilize. And I have a question. How far away from the output loop are the capacitors to count as Cout, because there will be a lot of capacitors out on the board, will they also count as Cout? Or does it have to be close enough to count.

Thanks

I have good news, after many days of trying, I found that I had neglected the derating of the input and output ceramic X5R capacitors. After I calculated the Bode plot with the correct parameters and adjusted the poles and zero by changing the inductor and output capacitor to make the Bode plot stable, such parameters were feasible. And do you think my PCB is OK? Thank you again for your reply.

THANKS