WEBENCH® Tools/LM5160A: LM5160ADNTT

Itzik,

That link does not take me to the design. 

Can you share your schematic?

-Orlando

hi,

attached.

BR,

Itzik7776.LM5160DNTR_WBDesign147.pdf

Itzik,

It your converter switching stable?

I have seen before where replacing the ceramic COUT with an polymer will cause output voltage ringing, and that can cause instabilities in the converter.

Can you try replacing COUT with a ceramic and also provide scope shots of SW node?

-Orlando

Hi Orlando,

Yes, the converter switching is stable.

• On cathode side of the diode I was saw levels voltage as i expect (according to the datasheet)
• On the anode side of the diode I was saw a stable level voltage of 5.16V (thanks for the diode and the capcitor, but with very very small load)
• After I was put a load of 216mA the level voltage fall to 4.6V

I was do some checking:

• First, I was replaced Ron to a smaller resistor, that bring my frequency go down to 115kHz
• Then on the primary side, close to "SW" pin of the conveter I saw a D.C of 11%, that makes the secondary side (on the anode side of the diode) get a stable level voltage of 5.36V and After I was put the same load as before (216mA) the level voltage fall to 4.995V (diode work fine, perfect!)

next steps and assumptions:

• Now I want to replace Ron resistance as larger as I can for approve converter's efficiency
• I was check  again the current supply issues and saw that the transformer's Irms (0.95A) limit my current consumption:
  • Secondary side maximum current is 0.875A, that meen that the converter on the primary side need to supply 1.3125A, because ratio of 1.5 (> transformer Irms = 0.95A), not good!
  • For proper supply I want to change the ratio to 1, change the level voltage on the primary side around to 5.4V (because the level voltage of the diode), then the primary and secondary current will be aquivalent (fix for the converter and for the transformer)

please your opinion, what do you think?

BR,
Itzik

Itzik,

Decreasing the switching frequency is OK, however I don't know why that improves your output voltage regulation across the load.

Lower switching frequency will lead to increase ripple current, so as you lower the frequency you will have to increase the inductance to reduce the peak current.

If the ripple current is too high, the resulting peak current can trigger current limit, or worse saturate the inductor.

Please see Equation 14 and Equation 15 on pg 19 of the LM5160 datasheet.

What do you mean "close to "SW" pin of the converter I saw a D.C of 11%"? Are you using an oscilloscope to look at SW? SW node switches between VIN and GND. There should be no DC offset, maybe you are seeing the "average" voltage.

Let me know,

-Orlando

Hi Orlando,

On the begining, the freq was 260kHz (although the webench tell 230kHz) and its seems that when I was put a load of 260mA the level voltage reach only 4.6V, it is not make sense with LM5160 abilities, so I search for the most proper freq and get 200kHz.

According to equation 14 and 15 the maximum current reaple is 0.338 mA and maximum IL(peak) is 1.3A (transformer I(sat) = 1.55A).

About the SW node, I am using oscilloscope and just wanted to saw the D.C. (not the level voltage or the average)

The most important issue that when i will reach 0.875A on the secondary side, probably the ratio of the tranformation and the transformer abbilities will limited my circuit and want to know if I will switch the transformation ratio to 1:1 (and make the fiting), is it a good practice idea?

thanks,
Itzik

Hi Orlando,

thanks for the answer, I was used this application notes. thanks...
next I will make the changes and will update you about the solutions...

thanks,

Itzik