[SOLVED] Push-pull chip closed loop problem?

[Tõnn]

Hy!
I have circuit what I experiment in Pspice for my graduation work. This is 12V-24V DC converter with push-pull chip. I got with 24ohm load resistor exactly 24V, that means that I have 24W converter and that is what I want. But if I lower the load resistor the voltage going to rise. I can not getting work closed loop system what is controlled in error amplifier. I had turned off current sense because I do not know if I need that, and I do not know how to get it work. But the main problem is that I do not have closed loop system. I have connected the closed loop but in this way it does not work. I change a output voltage with voltage divider to the 5V and compared it with reference voltage which is also 5.1V and with that the output voltage must not be change. Pulse width must be change in the different voltage but it not. I do not know what I have to do with comp output, I find those capacitor and resistor values in the another schematic (I found that circuit in UC1846 Texas Instrument datasheet page with Pspice files). Xfromer is in 1:2 relationship. In the PWM chip datasheet there is not clear understanding what values I have to put. In my option, my circuit should work.
Any thoughts?
Sorry my English language, it is not good.

 

[andre_luis]

hi Tõnn


I tried to simulate a SMPS too many years ago, also at University - modeling it at SPICE.
And I observed that simulation result did not match expected working.

The reason was that PSPICE configuration parameters was not defined properly. ( I discovered It later, after abandon this project ).
Basically, due SMPS simulation requires a wide magnitude float point numbers at numerical analisys, the reduction to default size, increases speed simulation, but affect iteration convergence.

+++

 

[atripathi]

First try simulating the complete circuit in open loop and see if you are able to control the output pulse duty cycle by varying the voltage at appropriate pin (may comp pin). Once you achieve this try feedback loop.

 

[Tõnn]

OK, I managed to get working current sense and closed loop, because if I have lower resistance and the voltage is not rising. But I got only 5V in output with closed loop and the duty cycle is also quite minimal. How I have to use a comp pin to increase duty cycle? What components I have to put here?

 

[mtwieg]

I've always hated canned spice models of smps controllers. I've always built them from scratch, so that I know exactly how they work. It's not that difficult, and I've done it with many controllers similar to this one. Plus, while making the model you will get a much better understanding of how it works.

 

[Tõnn]

Well I simulated AE desing system circuit and getting results. There is something strange, after some time the voltage become stable and the current sense voltage also decrease almost 0. I just cant understand why, it seems like some time is going off and then the voltage is stabilised. OK I understand that the error amplifier compare error voltage with reference voltage but if I try that in my circuit with different transformer I got 5V for start but it is rising.

Edit: It depends on the shunt resistor, I chosen simply to small resistor and also the closed loop is much longer, which is quite logical.

 

[atripathi]

In most of PWMs comp pin is direct input to comparator, comparing oscillator ramp with error generated to PWM pulse. The ramp is mostly in range of 0-3 volts. providing a DC voltage in this range on comp pin would directly control PWM pulse. However this would depend upon how the model have been build.

 

[Tõnn]

Are anyone have a clue how I calculate Rsense? I have found how I calculate voltage in Rsense but in this equation the Rsense is inside the equation. The Rsense is the current sense input what determines the pulse witdth.