Suggestions for OpAmp model

I have suggestions for the OpAmp model to use for HW10 if you have problems.

Try using "UniversalOpamp2" instead of LT1368.
The default parameters for this model is sufficient for our simulation.
The ideal opamp seems to not work well for many people.

If the design is done correctly, there should be no convergence problems, and no need for initial conditions.

Beom

PWM comparator questions

I have not been able to find a comparator that will work in this application. I'm sure there is one, but I'm not having much luck with getting the output for d correct. I have the sawtooth, and Vm , but I'm struggling with the comparator.
Suggestions?

Extra Credit

Hi All,

I am a little confused as to what the difference is between the last part of problem 1d and the extra credit is... they both ask to verify problem 1 using spice...

Can anyone elaborate on what the difference is between the two?

Thanks!

HW10 P16.1

In part (d), I am a little confused about the question statement. Are we supposed to find two power factors, one for input (surface S1), and one for output (surface S2)? Thanks.
- KT

HW 10-1 Creating PWM Spice block.

Hi,

I've read in the text that the PWM block is modeled in Spice as simply vx/VM where vx is the input and VM is the sawtooth peak value. The text also says to limit the output to between 0.1 and 0.9. A couple of questions:
1) I used the LTSpice voltage dependent voltage source, put the output of the compensator into the component and set the scale factor to 0.5. Does this work?

2) Where do the limit values 0.1 and 0.9 come from and how can they be implemented?

Thanks.

-Drew

failure to find a DC operating point

* C:\Program Files\LTC\LTspiceIV\homework 9.asc
Vg 1 0 12 AC 1
L 2 4 4µ
R 4 0 1
C 4 0 100µ
XU3 Vc N003 N005 N006 5 CPM Rf=.1 Va=0.3125 L=4u fs=100k
E2 N005 0 1 4 1
E3 N006 0 4 0 1
H1 N003 0 value={i(L)}
XU1 1 2 2 0 5 CCM-DCM1 L=4uH fs=100k
XU2 N001 N004 Vc opamp Aol=100K GBW=10Meg
R1 4 N001 10k
R2 N001 N002 9.1k
C2 Vc N002 11n
C1 Vc N001 700p
Vref N004 0 5
.lib switch.lib
.ac dec 101 10 100k
.lib opamp.sub
.backanno
.end

--------------------------

With a higher value for R (around 10 ohms) spice can resolve to a DC operating point. At R = 1 I get "Failed to find DC finding operating points for AC analysis".

Any suggestions?

I have tried (most) of the suggestions listed. 

compensation circuit questions

On homework 7 we were given a type 2 compensation network for problem 2 as part of a Buck voltage regulator operating in discontinuous conduction mode. As we are using this same compensation circuit for Homework 9 I have a few questions regarding the solution given on homework 7.

I have a calculated transfer function of vc/v is equal to: (sC2R2 + 1)/[sC1R1*(1+sR2C1C2/(C1+C2))].  If the pole at C1R1 is designed to be equivalent to C2R2 then (sC2R2 +1)/sC1R1 becomes 1 + w/s, an inverted pole.

In the solution, it states that "at mid-frequencies around fc, Gc(s) = R2/R1". Can anyone give me any help on where this comes from and how we determine it?

Additionally, the equation used to determine the crossover frequency where T = unity gain is given as:
1/Vm * R2/R1 * Gd0 * fp/fc = 1
where does the fp/fc come from?


UDPATE: I actually found a good paper on the Intersil site that discusses this compensation network.
www.intersil.com/data/tb/tb417.pdf

I'm still confused about the unit gain equation, and anyone who can give help on this would be appreciated.