The CNFB Method

FractalAudio

Administrator
Fractal Audio Systems
Moderator
I've finally perfected the "Chase Nonlinear Feedback" (CNFB) method for the modeling of nonlinear networks. And it works amazingly well. Has the accuracy of high-order integration methods with less computational burden.

Can simulate diodes, triodes, pentodes, etc. Far less error-prone than other methods (like K-method or DK-method, etc.) as you don't need to enter large matrices or tables.

It works on the principle that nonlinear devices can be thought of as linear devices with nonlinear feedback. You compute the states of a linear network and apply nonlinear feedback to get the output. It's also inherently stable. If the analog version of the network is stable, the CNFB implementation is stable.

The plot below is a simple example. This is a single-sided diode clipper with "memory" (the memory being a capacitor across the diode). The dotted line uses classic nonlinear ODE techniques solving the network using Trapezoidal Rule integration. The dashed line uses the CNFB method. The results are virtually identical but the CNFB method executes in about 60% the time (12 operations per loop vs. 20). As the number of nodes in a network increases the computational advantage increases proportionally.
cnfm.png

Here's a more complex example. This is a plot of a 6L6GC push-pull power amp into a reactive load (blue) compared to the same power amp simulated in SPICE (red). Doing this with conventional methods (nodal K, DK, WDF, etc.) induces major thinky-pain. I did this with the CNFB method in a couple hours.

cnnfb_power_amp_3.png

Could be a revolution in nonlinear network modeling.
 
Verrrrry Interestink!

MEME2022-01-05-09-27-34.jpg
 
Will this new method "only" contribute to lower processing power consumption of amp and drive instances or will it, on top of the computational benefits, also introduce greater accuracy in modeled circuits that are pushed to their analog limits?

Whatever the case may be, I'm just happy to witness and hear yet another great epiphany. This is exactly why Fractal is still at the top of the guitar amp modeling game.
 
Woohoo - Extra performance!!
Thanks Cliff

I've finally perfected the "Chase Nonlinear Feedback" (CNFB) method for the modeling of nonlinear networks. And it works amazingly well. Has the accuracy of high-order integration methods with less computational burden.

Can simulate diodes, triodes, pentodes, etc. Far less error-prone than other methods (like K-method or DK-method, etc.) as you don't need to enter large matrices or tables.

It works on the principle that nonlinear devices can be thought of as linear devices with nonlinear feedback. You compute the states of a linear network and apply nonlinear feedback to get the output. It's also inherently stable. If the analog version of the network is stable, the CNFB implementation is stable.

The plot below is a simple example. This is a single-sided diode clipper with "memory" (the memory being a capacitor across the diode). The dotted line uses classic nonlinear ODE techniques solving the network using Trapezoidal Rule integration. The dashed line uses the CNFB method. The results are virtually identical but the CNFB method executes in about 60% the time (12 operations per loop vs. 20). As the number of nodes in a network increases the computational advantage increases proportionally.
View attachment 93791

Here's a more complex example. This is a plot of a 6L6GC push-pull power amp into a reactive load (blue) compared to the same power amp simulated in SPICE (red). Doing this with conventional methods (nodal K, DK, WDF, etc.) induces major thinky-pain. I did this with the CNFB method in a couple hours.

View attachment 93826

Could be a revolution in nonlinear network modeling.
 
For the non-technical people among us (like me), what’s the impact of this revelation? Will we be able to cram more stuff in a preset? Will there be audible improvements?

I love that you post about what you’re working on. And admire the drive to constantly improve things when most of your customers are already happy...
 
For the non-technical people among us (like me), what’s the impact of this revelation? Will we be able to cram more stuff in a preset? Will there be audible improvements?

I love that you post about what you’re working on. And admire the drive to constantly improve things when most of your customers are already happy...
My guess is lower CPU use for (at least) Amp and Drive blocks... Although that's only a guess.
 
"Chase Nonlinear Feedback" (CNFB). Chase as in "Trail" or as in "Cliff Chase©️" :cool:

Congratulations for the development!! 👍
 
Hey Cliff,
The spice simulation in the second chart seems to indicate a marginally consistent faster leading edge - would that be audible?
Thanks
Pauly

I've finally perfected the "Chase Nonlinear Feedback" (CNFB) method for the modeling of nonlinear networks. And it works amazingly well. Has the accuracy of high-order integration methods with less computational burden.

Can simulate diodes, triodes, pentodes, etc. Far less error-prone than other methods (like K-method or DK-method, etc.) as you don't need to enter large matrices or tables.

It works on the principle that nonlinear devices can be thought of as linear devices with nonlinear feedback. You compute the states of a linear network and apply nonlinear feedback to get the output. It's also inherently stable. If the analog version of the network is stable, the CNFB implementation is stable.

The plot below is a simple example. This is a single-sided diode clipper with "memory" (the memory being a capacitor across the diode). The dotted line uses classic nonlinear ODE techniques solving the network using Trapezoidal Rule integration. The dashed line uses the CNFB method. The results are virtually identical but the CNFB method executes in about 60% the time (12 operations per loop vs. 20). As the number of nodes in a network increases the computational advantage increases proportionally.
View attachment 93791

Here's a more complex example. This is a plot of a 6L6GC push-pull power amp into a reactive load (blue) compared to the same power amp simulated in SPICE (red). Doing this with conventional methods (nodal K, DK, WDF, etc.) induces major thinky-pain. I did this with the CNFB method in a couple hours.

View attachment 93826

Could be a revolution in nonlinear network modeling.
 
Reading this stuff is mindblowing. I think I can literally read your enthusiasm on this discovery haha this is great

Even if i dont understand #$%* it still is mindblowing

Goodluck on the revolution Cliff



Cheers
 
The spice simulation in the second chart seems to indicate a marginally consistent faster leading edge - would that be audible?

I might be wrong, but the real 6L6GC (blue line) looks pretty identical to the simulated red line in the complex example (i.e. no leading edge).
 
Back
Top Bottom