CIC interpolators are WAY out of my realm of expertise. ?But I am puzzled by a few things.
Why is the Gain of the first VCCS (G1) so very low? ?You have volts going in, and nanoamps & nanovolts coming out. ?From there forward the transconductance gains are all 1 with 1 ohm loads, so the rest of the chain is down in the nanovolts range ... but with volts of signal elsewhere in the same circuit. ?Mixing nanovolts and volts in the same simulation, and expecting proportional accuracies (down in the femtovolt range!), does not always work well in analog circuit simulation. ?Maybe that's part of the problem. ?You might need to tighten up the tolerances (abstol, reltol, vntol or volttol, etc.). ?Or scale the amplitudes better.
Current source I1 looks odd. ?Is it there for a reason? ?Does the S/H not have a voltage output?
V4 and V5 have "zero" fall times. ?As you should know, you can't get zero rise or fall times because they are impossible. ?LTspice substitutes much slower values when you specify "zero" (because "zero" is shorthand for "use a default value here"). ?If you want very fast fall time, use a small, non-zero value.
The time-delay elements in the lower string are curious, but I hope you really meant for them to be used that way, as inverting delay elements (taking advantage of the way SPICE implements lossless transmission lines). ?Do they present the right kind of load to the current sources driving them? ?(I haven't used them this way so I don't know.)
Have you considered using B-elements as the delay lines, instead of T-elements? ?I have a suspicion that the results might differ at the microscopic level.
Looking at node 0.4, it appears LTspice isn't using fine enough internal timesteps. ?You may need to specify a Maximum Timestep.
How did you get the "COUNTER" behavioral element to use a gate-type symbol? ?I suppose it doesn't matter ... does it?
The error log has several "singular matrix" flags, pointing to nodes 'q' and 'o'. ?Your guess is as good as mine what it means, or whether it is important.
Regards,
Andy
