evalf(solve(...)) and fsolve return solutions in different orders.  One gives the first real solution and the other gives the second real solution.

Thanks,

@Joe Riel 

It looks like the call to Solve has some side effect that causes BodePlot to fail

restart;
    
with(Syrup);
Syrup:-Version;
with(DynamicSystems):

check_bode := proc()
    sys := TransferFunction( 1/(s-10) ):
    BodePlot(sys);
end proc:

################################
## This works
check_bode();

ckt := [V, Rsrc(50), C1(15e-9), L1(15e-6), C2(22e-9), L2(15e-6), C3(22e-9), L3(15e-6), C4(15e-9), 1, Rload(50)];
################################
## This works
check_bode();

(sol, other) := Solve(ckt, 'ac', 'returnall');
################################
## This fails
check_bode();

################################
## This works
restart;
with(DynamicSystems):
sys := TransferFunction( 1600000000000000000000000000000000000000000000000/(29403*s^7+78408000000*s^6+
670032000000000000*s^5+1298880000000000000000000*s^4+
4400400000000000000000000000000*s^3+5328000000000000000000000000000000000*s^2+
7360000000000000000000000000000000000000000*s+
3200000000000000000000000000000000000000000000000) ):
BodePlot(sys);

@Joe Riel 

Yes, I omitted some print statements.  In the cut-paste process, I omitted:

(sol, other) := Solve(ckt, 'ac', 'returnall');

I am using Maple 2018. My Syrup version is 0.1.16.

@Carl Love 

@tomleslie 

Reducing the function f to 1 reduced the problem to the volume of a sphere.

I work in Emacs, so I do a lot of cut/paste.  But, I do have a typo problem.

Thanks,

Tom Dean

@tomleslie 

Thanks for correcting my (silly) error.

Looking at my schematic, I numbered both the top ends of R5A and R5B then gave the output of U1A and U1B different numbers!

Sorry for the noise.

Thanks

@Joe Riel 

I tried this with ngspice-28.  The result was similar to using the IdealOpAmp in the above ckt.

I re-ran the analog.com filter design tool and got the same design.pdf.  I sent an email to analog.com.  No reply, yet.

@Joe Riel 

Here is the design file from analog.com Design.pdf

Look at the comments in my question.  This is a 4th order Butterworth.  Analog.com stated 10db gain and 3db down at 2kHz.  Looking at the component values, I questioned the performance.

Syrup gives very different results.

I get similar results to yours.

The netlist is from analog.com.  When I first saw the component values, I suspected the results, so I tried Syrup.  Syrup shows a much wider bandpass and 30db less gain.

@tomleslie  Thanks.  I just cut-paste and run,  Should read first.

@dharr 

limit(subs(theta2=theta1,soln[3]),theta1=0,left); ## or right

This is identical to one of the pdsolve examples, except the example uses

(D[2](u))(x, 0) = 0

as one of the initial conditions.

@Rouben Rostamian  

Linux, Maple 2016, 4.2GHz I7.

All I see is the cone with the minimal height cylinder.  Any ideas?

@Rouben Rostamian  

I use X with twm.  I do not use a desktop.

When an application's window is minimized, it is displayed as an icon.  This is not a maple toolbar thing.

The icon used for maple is maple2016/bin/Maple2016.png.  There is no maple2016/data/start directory.  There is a maple2016/data/start

searching the maple2016 directory tree for *.png finds 241 files.  only one file has maple in its file name, of any case, and that is maple2016/bin/Maple2016.png.

Most, if not all, the icons in the maple2016 tree appear to be for toolbars or the start page.  The largest is 75x75 pixels.

@rlopez 

The Tools/Options dialog has a setting for large icons on the toolbar.

If I start maple, choose Help, then minimize (iconize) the window, the displayed icon is very small.  Same for maple.  Start maple, then minimize (iconize) the window, the displayed icon is very small. 

I think the help icon is built into maple help??

The maple icon is maple2016/bin/Maple2016.png.    If I could find a larger version of that icon, it would make my use easier.