LM5046: Power management forum

Anthony Turpin

Part Number: LM5046
Other Parts Discussed in Thread: UCC28951, LM5035, LM5045, LM5036, LM25037

Hello everyone,

I am a beginner with the PSFB topology, but I have to design a boost converter with it. My input voltage range is from 40V to 60V and the output voltage is 400V, for a 1kW output power.

I can find only very few examples of PSFB for a boost application, but I am almost sure it is the right topology. However, what about the LM5046 controller ? Will it be suitable for my application ?

Let's assume I have right until here, I would have a design issue. One of the condition to perform ZVS is given by the following equation :

With a C_r of 3 nF, this leads me to a needed leakage inductance about 50 nH if I want ZVS at 50% load ! It will practically impossible to have so few leakage inductance.

If I want to a ZVS at 10% load, I would need 1.8uH of leakage inductance. It still seems to be weak compare to what I can find in the UCC28951 datasheet (page 46)

So I think there is something I am wrong. Do you have an idea ?

Thank you very much for your help

over 5 years ago

0 Colin Gillmor over 5 years ago

TI__Mastermind 25570 points

The PSFB and HSFB will both boost the input voltage up to a higher output voltage - it just depends on the transformer turns ratio.

The equation is an inequality - it calculates the minimum leakage inductance you need to achieve ZVS but there is no real problem in having more than the minimum - even significantly more then the minimum is ok. As the leakagt inductance increases, the time needed to reverse the current in the primary as the H bridge changes the polarity of the voltage across the transformer increases. During this time no energy is transferred from primary to secondary so the term 'duty cycle loss' is sometimes used to describe its effect. The duty cycle loss limits the maximum duty cycle you can get at the bridge but this is not normally a problem.

The leakage inductance needed to achieve ZVS is not a function of the controller used - it's determined by the capacitance at the switched node and by the voltage across the bridge. so the calculation for the LM5046 would be the same as for the UCC28951.

Finally, at low input voltages like yours the benefits of ZVS can be quite small and the losses due to the circulating current needed to achieve ZVS can exceed the savings due to ZVS. For this reason the Hard Switched Full Bridge (HSFB) is an alternative worth considering. LM25037, LM5045, LM5036, and LM5035 are all parts worth looking at - although you will have to add external drivers to some of them.

Have a look at the power stage designer tool, available at http://www.ti.com/tool/POWERSTAGE-DESIGNER

It will allow you to compare the HSFB and PSFB currents (mainly transformer I_pri and the MOSFET currents) and give you some guidance on component selection.

Regards

Colin

0 Anthony Turpin over 5 years ago in reply to Colin Gillmor

Prodigy 20 points

Hi,

Thank you so much for your answer, you gave me very important details that prevented me to progress in trust

Colin Gillmor said:
Finally, at low input voltages like yours the benefits of ZVS can be quite small and the losses due to the circulating current needed to achieve ZVS can exceed the savings due to ZVS. For this reason the Hard Switched Full Bridge (HSFB) is an alternative worth considering. LM25037, LM5045, LM5036, and LM5035 are all parts worth looking at - although you will have to add external drivers to some of them.

You're right, but I have chosen this topology also because we have heavy EMI constraints

0 Colin Gillmor over 5 years ago in reply to Anthony Turpin

TI__Mastermind 25570 points

Hi Anthony

EMI is always an important consideration !.

Do get in contact again if you have any further questions. For now - I'm going to close this thread.

Regards

Colin