Buck Converter PSpice Help

[willem445]

New member here, having some trouble designing a buck converter for a class project.

Here is what I have in PSpice



And this is the error im getting.





I think it has something to do with all my currents and voltages in the circuit in PSpice (first picture) showing currents of like 1.4e-24A and voltages of -1.03e-21V, when I use a voltage probe or current probe, it shows the correct voltages and currents. This only happens when I use the Vsin supply.

Any ideas on how to fix this?

Also the design requirements need me to step down 120V 60Hz to 60V 5W. I was planning on stepping it down to 90V with the transformer and then step it down to 60V with the buck converter. Is this a good way to go about it?

Thanks,

 

[FvM]

The circuit fails to converge within the first picoseconds, in so far the voltage and current numbers are pretty meaningless. Not all circuit parameters are visisble, I don't see an obvious problem at first look.

Convergence failure is a well known SPICE problem since ever.

May be a problem with "floating" bridge rectifier nodes, could be sufficient to connect a high ohmic resistor from transformer secondary to ground. Or increase some relative and absolute tolerances in SPICE parameters.

 

[BradtheRad]

Your C2 capacitor is a large value. It takes many cycles to charge.

Try a Farad value between 1 and 100 times your inductor value.

 

[willem445]

Thanks for the replies, we went over the design part in class so all of my values I picked should be correct to get 60V and 5W output. Still having the convergence issue.





Is it something with the MosFet? The notes say to use QbreakN for the switch. Any ideas?

 

[BradtheRad]

Try adding a high-ohm resistor across each diode.

Also add a series bias resistor at Q1.

Also add a low-ohm resistor inline with each capacitor.

 

[KlausST]

Hi,

Maybe try to select a real NPN transistor type.
A current limiting resistor at it's be is useful.

Klaus

 

[FvM]

Presently, V2 shorted by the transistor base-emitter junction is your problem. But after solving it, you'll find other issues like absurd 100 MHz pwm frequency.

 

[willem445]

Presently, V2 shorted by the transistor base-emitter junction is your problem. But after solving it, you'll find other issues like absurd 100 MHz pwm frequency.

Oh duh! So now my simulation completes without any errors but im still not getting an output voltage.








Im getting the correct 90V inputting into the FET but the output of the FET is 100V @ the switching frequency.

As for the FET settings I changed PER to 1/10k and PW to .75u (Duty Cycle / Frequency). Is this correct?

BTW, thanks for the help everyone, I was not expecting any replies at all. Glad I discovered this forum!

 

[willem445]

Hi,

Maybe try to select a real NPN transistor type.
A current limiting resistor at it's be is useful.

Klaus

What do you mean by "limiting resistor at it's be" What is "be"

 

[KlausST]

Hi,

"be" is a typo. I meant "base"

A resistor that limits the current into the base of the bjt.

Klaus

 

[FvM]

To step 100 V down to 60 V, you'll want 60 % duty cycle (pw = 60 µs), at least if the converter is working in continuous conduction mode (CCM).

1N4002 isn't well suited for 10 kHz switching frequency, you'll either use a generic rectifier diode, or a real fast type.

Every detail of circuit operation can be checked by looking at the simulation waveforms.

 

[willem445]

To step 100 V down to 60 V, you'll want 60 % duty cycle (pw = 60 µs), at least if the converter is working in continuous conduction mode (CCM).

1N4002 isn't well suited for 10 kHz switching frequency, you'll either use a generic rectifier diode, or a real fast type.

Every detail of circuit operation can be checked by looking at the simulation waveforms.

Phew, finally got it to work. Was accidently typing .66u in for my duty cycle instead of 66u and changed to a generic diode.

Here is what it looks like now.



Is it normal to take roughly 150ms to reach 60V steady state?

 

[FvM]

It's generally no good idea to operate a switching converter with fixed precalculated duty cycle. Instead you better use a pwm controller with voltage feedback. It can achieve faster settling of intended voltage setpoint.

 

[willem445]

It's generally no good idea to operate a switching converter with fixed precalculated duty cycle. Instead you better use a pwm controller with voltage feedback. It can achieve faster settling of intended voltage setpoint.

Yeah we just barely covered the PWM to adjust the duty cycle. Essentially we know how it would be done but we haven't covered anything on how to design and implement it yet. Thanks a ton for your help though. I should be able to take it from here.