Electronics > Projects, Designs, and Technical Stuff

Boost Converter in CCM?

(1/1)

Dmeads:
Whats up! trying to learn Switching converters with LTspice and I was wondering if this one is operating in continuous conduction mode?

The picture is of the current in the inductor (blue) and the current in the diode (green) IMPORTANT NOTE: I offset the diode (green) waveform by -5 Amps so you can see what is going on otherwise they over lap.

I am assuming that this converter is in CCM because the current in the inductor doesn't go down to zero before the diode starts conducting because of the switch turning on.

Is this correct?

I calculated my duty cycle and stuff from this link: http://www.ti.com/lit/an/slva372c/slva372c.pdf which was helpful but i needed to adjust a little bit. I increased my inductor size by 4 times to reduce initial current spikes in the inductor.

switching frequency is 400kHz and I gave a random ESR to the inductor of 0.2 Ohms cuz in reality it will have thick wire.

picture and file are attatched.

Thanks!

TimNJ:
You can easily distinguish CCM from DCM by checking if the inductor current ever reaches 0. If it reaches 0 during the switching cycle, it is discontinuous (DCM). In your case, IL1 is continuous, so I'd say its CCM.

The positive slop of the current wave should be equal to Vin/L, and the negative slope should be equal to -Vout/L. Your slopes are fairly close to that.

On a practical note, notice that your inductor ripple current is extremely low at less than 2%. This might sound like a good thing, but in reality it means your inductor will have to be physically very large to not saturate at 15A and 240uH. Most designs will shoot for 10 to 30% ripple current as a trade off. That is, drop the inductance, increase core gap (if there is one), and so on.

jmw:
Yes that's CCM. Zoom in on I(L1) and you'll see the expected sawtooth waveform. The boundary condition for DCM operation is: \$K < K_{\mathrm{crit}} = D(1-D)^2\$ where \$K = \frac{2 L f_s}{R_L}\$.

Ferrite core inductors saturate and 14.5 A is going to require something substantial. Winding loss alone at 200 mohm is going to be almost 40 W. Also, I assume you'll eventually want to operate closed-loop and 84% duty cycle is close to the limit of what a lot of controllers will support.

T3sl4co1l:
Zero rise and fall time is rather unrealistic, especially with a zero(?) ohm source.  20ns and 5 ohms would be pretty typical for a gate driver.

I think that will address the huge fast spikes.

Tim

FalkMusick:
To reduce the simulation time, you m might consider using the spice initial condition command. For example .ic V(Vout)=160 I(L1)=12

Navigation

[0] Message Index

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod