self oscillating flyback smps

[blackmamba]

hey all!

Am working on a self oscillating flyback smps. i get the schematics from smps handbook.

when i start simulating the circuit using LTspice it doesn't work
i couldnt generate the proper signal to fire the switching transistor.

could any one look at my schematic and help me..
i think its the values of the elements or somthing about the flyback transformer !!

and thanks in advance..

[fernando_g]

I've worked with this topology in the back, and although it appears deceivingly simple, it is not.

For starters, the 2N3055 is way too slow. Replace that with a suitable MOSFET instead.

[blackmamba]

But i couldnt establish a signal at the base ..

I thought it may be the values of the components which control the boundary between CCM and DCM or the transformer modeling..
because i tried a mosfet and i had the same proplem, there's weird signal at the gate!!

[fernando_g]

What did the simulation log show?
Any Convergence problems?
Try reducing your step size

[Elvee]

I notice C2 has no suffix. This means the simulator will assume it is a 200 farads capacitor.
Could this be the problem?

[blackmamba]

Thanks for the replies..

due to the snubber capacitor C2 may be because i changed the values thousand times so may be i miss this one.. sorry

after i changed the switching transistor to IRF mosfet i got a switching signal but it still has a random behavior.

i will attach the simulation file to make it easier if anyone could help..

[Elvee]

Obviously, with your values, you arrrive at insane voltages and currents: when properly simulated, i.e. a time long enough for the initial conditions to settle, and a time step short enough, the voltage on the drain reaches over 330V.

Off course, it is practically unloaded, and there is no regulation.

Have you properly defined target values for the frequency, conduction time, peak current, output power, etc?
Where will you introduce the regulating/feedback element?
What regulation strategy will you use?
These are the prerequisites, and once they have been laid down, you have to design the circuit and calculate the values accordingly.
A more informal approach will not work.
I have substituted more realistic values, and now it works in a more sensible manner, but all the above-mentionned issues remain open.
I'll try to disguise the .asc file into a .txt. To use it, restore the correct extension.

[blackmamba]

hey..
the circuit is low power and has variable frequency so i prefered to let these elements open..

and the regulation in the schematic am using has a feedback signal throw a shunt reglator (TL431) and opto coupler but i thought it is not necessary to complicate the circuit as
long i cant get a stable output voltage .

[megajocke]

Try this The skip inital operating point solution should be turned off or else circuits like these might not start.

The component values you chose gives a way too high frequency for this kind of circuit. My changed circuit optimally should have the inductances increased even further to give lower operating frequency.

The darlington/CFP is needed or the current trip point will change too much between oscillating mode and stopped mode making burst operation start at pretty high load levels.

The 470 ohm resistor makes the turn on/off pretty slow, this is why frequency should be kept low. The resistor could be reduced but watch out for high turn-off-transistor current and high zener current at high input voltage.

[blackmamba]

thaanks .. its very much more than i expected