LM5116: LM5116_ Buck converter

6170.Schematic.pdf

I forgot send the schematic of the board. 

Hello Eric,

Many thanks for supporting me.

I checked my Kelvin connection, CSG and CS traces are directly connected to R6 and R7 respectively. The other side of R6 is directly connected to R11(Rsence) but the other side of R7 is connected to R11 by via( attached photo_01 ).

For putting two filter capacitors close to the CS-CSG pins (~100pF). Do you mean, I put two bypass capacitors(100pf) for CS and CSG pins? (like photo_02)

About the RC filter using R6, and R7. Do you mean I create a low pass filter by the R6, R7, and capacitors? ( like Photo_03)

Finally, are populated CS and CSG pins with some capacitors like photo_04.

Thanks again, 

Best regards, 

Leila

Hello Leila 

1. Please try to populate the ~ 100pF - 1nF filter capacitor between the CS and the CSG pins. In order to form a RC filter, you have to  populate ~100ohm - 1kohm at R6 and R7.  

2. This filter configuration is useful when there is a lot of common mode noise, but ideally the ground connection of the C1 and C2 capacitors should be directly connected to the AGND.  

-EL

Hello EricLee,

Thank you!  

I changed R6 and R7 to 220 ohms and also, I put the one capacitor 220pf between the CS and CSG pins (like photo_04).

When I measured the differential voltage of two pins ( VCS-VCSG), It was decreased like the simulation. 

Photo_05:  without RC filter; R6 and R7 are 0 ohm.

photo_06: with RC filter ; R6 and R7 are 220 ohm and C is 220pf.

I connected a capacitor like the photo_07.  It is located between the CS pin of the R7 and the CSG pin of the R6. Is it important to connect directly to these pins in LM5116? 

Best regards,

Leila

Hello Leila

I think you forgot attaching your bench test results. I cannot find photo_04, photo_05, photo_06

-EL

Hello EricLee,

Today I put a load that is 6.8 ohm and 100 watts ( because I designed it for around 4 A).

Unfortunately, I encountered new problems:
First: the output voltage dropped to around 5 volts.
Second: The inductor creates sound continuously. Then I measured the Vcc voltage with/ without load ( attached).
Third: the PCB getting hot.

I used to LM5116 quick start for designing a buck but I set the ripple current on 80 percent and then I selected 23uH. Today I found the operation value it be around 25 to 30 percent ( datasheet section 8.2.2.3 ). When I applied 25% in the Excel file I realized that many changes(attached). Should I apply all of them?

Best regards,
Leila

Vcc with Load.pdfvcc output without Load.pdfcompare of ripple current.pdf

Hello Leila

Please check these items. The VCC should be in regulation with load. 

• Check VIN pin voltage if VIN drops too much VCC cannot  be regulated
• What's the VCC voltage ? the MOSFET plateau voltage should be greater than ~VCC+0.5V. 
• Check if the switching frequency is close to the 250kHz

-EL 

Hello EricLee,

Many thanks for your quick reply.

Vin is constant like without the load and it is 48 volts.

My meaning of the VCC belongs to the VCC pin of the LM5116 that is connected to the diode.

Based on the data sheet of the my MOSFET (Si7852DP), I think the plateau voltage is around 4.5 volts( attached). In my case with load VCC is around 7 volts.

Thanks.

Leila

Hello Leila

The audible noise can be caused by either loop instability or hiccup mode protection. Can you measure inductor current and check which one causes the audible noise ? 

-EL

Hello EricLee,

Thank you!

I measured the switching frequency of my buck converter is 250 KHz( attached). Also, I measured the Inductor Ripple current with the load( 6.8 ohm and 100 watts). It is 3.9A ( based on the below formula:

I_L= ((Vin_max -Vout)*Duty cycle)/(f_sw*L)

Also, I attached the result of the all test points with Load.

I think my mistake belongs to the maximum ripple current, which I considered 80 percent. What is your opinion about the change of the inductor and the rest of the components based on the maximum ripple current of 25%?(attached)
In the quick start of the LM5116, I set Vccx on 12 volts ( attached ). Do you think I need it to be connected to the 12 Auxiliary voltage?

Thanks, Leila

IL result with load.pdfOutput result of test points with load.pdf

Hello EricLee,

Thank you so much for helping me.

In the C_out block, I used four Tantalum Capacitors - Polymer with 35V 47uF,20%, and ESR=30mOhms( link below).

Manufacturer number: T521X476M035ANE030:

https://www.mouser.fi/ProductDetail/KEMET/T521X476M035ANE030?qs=0kYgdqhPduEGIphu%252BgIzpQ%3D%3D&_gl=1*1kie54m*_ga*dW5kZWZpbmVk*_ga_15W4STQT4T*dW5kZWZpbmVk*_ga_1KQLCYKRX3*dW5kZWZpbmVk

Additionally, the last capacitor is Multilayer Ceramic Capacitors MLCC with 35VDC 22uF 20% X5R 1.25mm.

https://www.mouser.fi/ProductDetail/TDK/C2012X5R1V226M125AC?qs=6JAMGB%252BEdkyUk0d2Mu2IrQ%3D%3D

Manufacturer number: C2012X5R1V226M125AC

Do you mean I should put 70% of the 210uf as Cout block?

Thanks again.
Leila

Hello Leila

No, you don't need to use the ~ 70% derating factor if tantalum capacitor. 

-EL

Hello EricLee,

Many thanks for your reply.

If I change some components with 25% ripple current and VCCX  to zero based on LM5116 quick start ( selected components), can I reach to a suitable result?

Thanks again,
Leila

Ripple current _v03.pdf

Hello Leila

Let's see if any improvement. I guess you will be able to see a continuous switching. 

-EL

Hello, 

Thank you so much for helping me. 

Leila

Hello Leila

You don't need to reply if no update. Please let me know if any improvement after replacing the inductor.  

-EL