LTSpice Boost Converter Simulation and Improvement

[arthurartmaster]

Hi All I have setup a simulation in LTSpice with components selected that I wish to create a real PCB for but wanted to simulate it first.
I am having a it of an issue with the design I used LTpowerCAD initially to workout what components to use, It takes in 12V and outputs 40V @ 7.5A. However when simulating it does not produce the output waveform expected I feel like I am missing something trivial any help would be appreciated. Simulation is attached with this post. Any suggestions for improvements or reduction of any noise would also be appreciated.

Kind Regards

Art

 

[ronsimpson]

I have a file that does work so I compared your work.
First I move "mode pin" to ground with no effect. (did not move back)
Then I realize you have "soft start" set to 4.5mA delay on start. So move from 0.1uf to 0.01uf and it starts in 0.45mS
Now it is working (almost)

If you build one of these in the real world. Your MOSFETs are only 40volt. And the boost diode is only 40volts. SPICE will allow you to put 41 volts on a 40V part. Don't do that on this Earth.

There seems to be a large amount of ripple on the output. Too much current on small capacitors.

 

[arthurartmaster]

I have a file that does work so I compared your work.
First I move "mode pin" to ground with no effect. (did not move back)
Then I realize you have "soft start" set to 4.5mA delay on start. So move from 0.1uf to 0.01uf and it starts in 0.45mS
Now it is working (almost)

If you build one of these in the real world. Your MOSFETs are only 40volt. And the boost diode is only 40volts. SPICE will allow you to put 41 volts on a 40V part. Don't do that on this Earth.

There seems to be a large amount of ripple on the output. Too much current on small capacitors.

Great thanks so much thought it was something trivial I just wanted to get a working model first then go in and change components to meet ratings for definite. For the output ripple can I improve it by using ferrite beads etc? Also should I be increasing the output capacitance then?

 

[arthurartmaster]

I have a file that does work so I compared your work.
First I move "mode pin" to ground with no effect. (did not move back)
Then I realize you have "soft start" set to 4.5mA delay on start. So move from 0.1uf to 0.01uf and it starts in 0.45mS
Now it is working (almost)

If you build one of these in the real world. Your MOSFETs are only 40volt. And the boost diode is only 40volts. SPICE will allow you to put 41 volts on a 40V part. Don't do that on this Earth.

There seems to be a large amount of ripple on the output. Too much current on small capacitors.

Do you have any suggestions of how I can improve the model also does the PGood pin go high once the desired output voltage is reached?

Kind Regards

Art

 

[ronsimpson]

does the PGood pin go high once the desired output voltage is reached?

PGood is open collector (open drain) So you need a pull up resistor. but yes!

You can add a CLC filter on the output. "Pie" ∏ There found the symbol.
I am stuck at 997 likes (for being helpful) Trying to get to 1000. The last 30 people did respond.
I think you need more capacitors on the input and output. Look at the peak currents. It is very high. That much current is causing volts of ripple. The output cap needs to hold a 7.5A supply up for one cycle. You need more storage. Pretty simple math but you can just look at the ripple and know that doubling the caps will cut the ripple to 1/2 (10x = 1/10)

 

[ronsimpson]

It takes in 12V and outputs 40V @ 7.5A.

I am thinking about math. 40V X 7.5A = 300 watts. So you will need to pull 25A from the 12V supply. Actually a little more current because the PWM and the MOSFETs are not 100% efficient.

 

[arthurartmaster]

I am thinking about math. 40V X 7.5A = 300 watts. So you will need to pull 25A from the 12V supply. Actually a little more current because the PWM and the MOSFETs are not 100% efficient.

Yeah from the spec I was given for this project they wanted power of 300W delivered