Design of PWM for Cuk Converter using LM5025

[dtothec]

I hope this is in the right place and apologize if it isn't.
Not too long ago I posted a thread about a Cuk converter that I was tasked with designing. Currently I have hit a snag yet again and hope that its as simple a fix as the last problem I had.

I am using an LM5025 as a PWM to help control the Cuk converter. I found a post online that showed a circuit schematic using only the LM5025. I built that circuit in PSpice and simulated it. Then I copied my Cuk converter over into the design and simulated it again. Both circuits can run just fine on their own, but when I connect the two I get problems. The voltage source connected to the gate of my MOSFET is a Vpulse with peak-to-peak voltage of 5V. I have also made a voltage divider at OUT_A of the LM5025 to get the voltage there to around 3.7V. I set the transient analysis as: Run time = 1ms, save data after = 0.1ms, max step size = 0.1ms.

When I connect the two circuits together I make a connection from the gate of the MOSFET (I remove the Vpulse) to the voltage divider. However the simulation never runs and always gives an error that says "ERROR(ORPSIM-15138): Convergence problem in transient analysis at Time= 208.7E-09". I'm unsure of what that means. I also believe that there might not be enough/too much current running through the MOSFET when the connection is made. If anyone could give any advise or point me in the right direction I would be very grateful.

Thank you for your time,
dtothec

I have attached pictures of both the Cuk converter and the LM5025 circuit, and the calculations I used to design the converter.

 

[schmitt trigger]

" have also made a voltage divider at OUT_A of the LM5025 to get the voltage there to around 3.7V."

Why would you want to do such a thing? And with 1 meg resistors?

 

[dtothec]

When I run my Cuk converter if the Vpulse is set higher that 5V then it doesn't simulate. The voltage divider is so that OUT_A's voltage is halved, so that it will drive my converter. The only reason I used 1 meg resistors was that I just copied what was already there. As far as I know they can be changed.

 

[schmitt trigger]

Why did you select that controller?
The LM5025 is an active clamp controller, I don't see the clamp in your powertrain.

Even if there were a clamp, I'm always suspicious of claims of successful simulations for PWM controllers running open loop.

The powertain isection, one can simulate more or less by itself, by setting voltages and pulses. But a controller requires feedback, from several sources to simulate properly.

For instance....the comp terminal you got a fixed 5v supply. However, the datasheet clearly indicates that: "The external opto-coupler sinks current from COMP to control the PWM duty cycle."

 

[mvaseem]

Can you attach the netlist and cir files. May be I can try to play around the circuit and various options to see if it can be resolved.

 

[atripathi]

Can you share the circuits (as suggested by Mvaseem) or complete circuit (where two of these modules are connected)? This would enable people to have a closure look.
Meanwhile - try following setting and see if this helps

- ITL4= 40
- ABSTOL=1n
- RELTOL= 0.005

MAX TIME STEP = 50n --- If this works, in next run try increasing this to higher value but this should be at least one order lower that switching time period.

 

[schmitt trigger]

The original poster has apparently disappeared without commenting the suggestions