LM5123-Q1: Boost doesn't work as supposed to

Hi Harry,

Thanks for your answer.

1. I followed my E-mailing with you and saw that there was a recommendation to use 0.861uF but that was while considering that the output capacitance is 220uF instead of beyond 1000uF. I finally managed to insert my board 1360uF for boost output capacitance and even the excel shows that keeping the 1360uF allows us 40% ripple with inductance of  2.95uH:

 2. In addition, consider that L=1uH gives a very bad efficiency while the boost in my system is actually intended for boosting batteries voltage, meaning when batteries are in discharge their voltage is between 8v-12v and the boost shall convert it to stable 16v with good efficiency so the batteries won't be wasted. 

According to the excel 1uH gives 90-92% which is very bad compared to 97-98% gotten with 5.3uH. Even 3uH is much more efficient than 1uH.  

If so then does 1uH still recommended just for increasing the ripple current though we can work with a higher output capacitance such as 1360uF ? How much does the Inductor average current is increased in such a case?

3. BTW, you wrote that you usually recommend +- 30% which means an amplitude of 60% (this is what you enter in the QuickStart calculator). 

What exactly shall be entered to QuickStart? 30% or 60? Is that the ripple recommended to my conditions ? 

Vin: 8v - 12.5v, Iout=10A, Vout=16v. 

4. Can you please review my design I showed you and gibe your FB  ?

Thanks & BR,

Hi Ohad,

I will take this over from Harry. Just to be on the same page, can you please share the schematic you use as a PDF, as well as the filled out quickstart calculator? So i can do a efficient review. 

Ohad Fundoianu said:

I finally managed to insert my board 1360uF for boost output capacitance and even the excel shows that keeping the 1360uF allows us 40% ripple with inductance of  2.95uH:

I am not sure how this is meant, the inductor current ripple is not related to the used output capacitance. We recommend to have a inductor current ripple between 40%-60% for achieving a good regulation. A smaller inductance will cause a higher peak current through CS resistor as well as the FETs. This will cause more conduction losses and decrease the efficiency. However i see that you are switching at 200kHz only, right? 

Because in this case 3uH would alsready be low enough to allow a sufficient current ripple.

Best regards

Moritz

Hi Moritz, 

Thanks a lot for the answer. 

I sent Uriya who represent TI in my area the pdf and he shall send it to you by mail. I can not send it on E2E.  

there is one thing I didn't understand from Harry. Suppose I choose inductor current ripple of 45%, what number shall I put in the "Quick start" GUI? Shall I put 45%? Or 90% since +-45% means amplitude of 90%? 

In addition, in that case you described where Inductor is 3uH, is it fine to choose the sense resistor to 1mOHM? It was mentioned that too little sense resistor can take the boost out of regulation, does 1mOHM is fine?  

BTW, we do now some more considerations, will the Boost be able to support Vout=16v with Iout=19A ? meaning 304W

Moritz, will it be possible to arrange a meeting today with you in order to clarify some unknown things? 

Ohad

Hi Ohad,

Thanks, i will try to get the PDF from him.

Ohad Fundoianu said:

Shall I put 45%? Or 90% since +-45% means amplitude of 90%? 

You should put 45% in the calculator then. 

Ohad Fundoianu said:

It was mentioned that too little sense resistor can take the boost out of regulation, does 1mOHM is fine?  

You should make sure that the expected peak current will not lead to a higher current sense voltage than 60mV. (plus add a bit of margin)

When you select an inductor value in the quickstart calculator, the peak inductor current will be shown 2 lines below. 

Based on this and the current limit margin you set in line 28, a value for Rcs will be recommended. 

For a proper regulation, the controller needs to see a voltage ramp at the CS resistor. Lets say you peak current would be 5A only and the sense resistor would be 1mOhm, there would be nearly no voltage seen and you would not achieve a stable operation.

Ohad Fundoianu said:

BTW, we do now some more considerations, will the Boost be able to support Vout=16v with Iout=19A ? meaning 304W

Theoretically yes, but for this power level you could also think about using 2 phases, because the currents will be quite high.

Ohad Fundoianu said:

will it be possible to arrange a meeting today with you in order to clarify some unknown things? 

Please contact your responsible sales person for setting up a call in case it is still needed.

Best regards

Moritz

Hi Moritz, 

I have looked on the functional block diagram to better understand how does exactly the Vcs value affects.

I see two comparators, the down and the up. It seems the down check that Vcs is not above 60mv. 

What does the down comparator check ? Does it check that Vcs is not below 10mv ? Does 10mv is the Vcs threshold that any value lower than that will cause the boost to be not stable? If not then what is the minimal threshold ? If yes then what is the meaning of that? Does it mean the Boost output voltage will taper down or up? In how many? 

What is the minimal output power required for not getting out of stability?  

Ohad

Hi Ohad, 

As this is discussed in a call today, i will put this thread on hold.

Best regards

Moritz