LTspice problem

Thread Starter

flaviotmz5

Joined Aug 23, 2021
14
My simulation with a dc forward converter has a problem when I add any parallel capacitance to the inductor L3 even when I add a real diode after the L3, the problem with the parallel capacitance is that surge an extremely high peak current in the MOSFET, I don't think that's suppose to happen, it's an LTspice problem..., any ideas how to solve it?
 

Attachments

Papabravo

Joined Feb 24, 2006
21,159
... surge an extremely high peak current in the MOSFET, I don't think that's suppose to happen, it's an LTspice problem..., any ideas how to solve it?
That's an extraordinary claim with little or no evidence presented. Show us an analysis of why you think it is not supposed to happen.
I see a peak current of just under 12A. What current were you expecting in the MOSFETS, with a 50 VDC input supply?
Everything looks perfectly normal to me; what design requirements did you start with?
 
Last edited:

Thread Starter

flaviotmz5

Joined Aug 23, 2021
14
That's an extraordinary claim with little or no evidence presented. Show us an analysis of why you think it is not supposed to happen.
I see a peak current of just under 12A. What current were you expecting in the MOSFETS, with a 50 VDC input supply?
Everything looks perfectly normal to me; what design requirements did you start with?
12 A it's okay by design, but when I add a parallel capacitance in L3 surges a peak 54A.
 

Attachments

Papabravo

Joined Feb 24, 2006
21,159
12 A it's okay by design, but when I add a parallel capacitance in L3 surges a peak 54A.
Which capacitor are you talking about?
Never Mind. What you did was add a parallel capacitance to L3 without making that fact visible and obvious on the schematic. What is the resonant frequency of 1300 uH and 1nF? Why did you choose to do that? Why did you give us the wrong schematic in your original post? What you have is the undamped transient response of a resonant circuit. It lasts about 130 nanoseconds which may or may not be significant.
 
Last edited:

Thread Starter

flaviotmz5

Joined Aug 23, 2021
14
Yeah. I figured that out. If you're going to do things like that you should make it obvious on the schematic. See revised post #6
Sorry for the inconvenience, I add 1n just for the simulation to go quicker because I actually want is to add the diodes model in D3 and D4.

I'm attaching the simulation with the model of D3 and D4 and the problem persists. But I think because it's just 130 ns the semiconductors will be fine right?

Sorry for the trouble, really I just look at the simulation and don't think too much, sorry
 

Attachments

Papabravo

Joined Feb 24, 2006
21,159
Sorry for the inconvenience, I add 1n just for the simulation to go quicker because I actually want is to add the diodes model in D3 and D4.

I'm attaching the simulation with the model of D3 and D4 and the problem persists. But I think because it's just 130 ns the semiconductors will be fine right?

Sorry for the trouble, really I just look at the simulation and don't think too much, sorry
There is more going on than you seem to be aware of. I get the feeling that you are adding components and parasitic elements at random without any meaningful analysis or understanding of physically what is going on in this circuit. The choice of component values is based on a number of considerations, none of which you seem to have take account of. Did you pull the value of the parallel capacitance on L3 out of thin air, or does it represent an actual transformer?
 

Thread Starter

flaviotmz5

Joined Aug 23, 2021
14
There is more going on than you seem to be aware of. I get the feeling that you are adding components and parasitic elements at random without any meaningful analysis or understanding of physically what is going on in this circuit. The choice of component values is based on a number of considerations, none of which you seem to have taken account of. Did you pull the value of the parallel capacitance on L3 out of thin air, or does it represent an actual transformer?
I've already built up the transformer, and the values are real. The capacitance now I need to revise. But it's not random do not worry about it.
 

Papabravo

Joined Feb 24, 2006
21,159
I can see the things you did (options, C3, D3, & D4) to eliminate the current spike at the turn-on edge.
What was causing the current spike?

ETA: You also changed the properties of the pulse generators V2 & V3 to add some delay, rise time, and fall time. LTspice gets cranky with discontinuous waveforms. I made those changes as well but the current spikes on the drains of M1 & M2 are still and they are worse. Also they are short duration and maybe they don't matter. Everything else looks "normal"
 
Last edited:

Papabravo

Joined Feb 24, 2006
21,159
Got it -- thanks
No real transformer can have a coupling coefficient of 1.
V2 & V3 need series resistance.
 
Last edited:
Top