Patent US5032796A Harmonics vary as a function of the amplitude and duration

[Vivek]

 Patent US5032796A is regarding changing the duty cycle of asymmetrically chopped waves, based on amplitude and duration of the input signal. Supposedly this comes close to "Sag" and hence makes a SS device sound more "tubeish".


Today I finally decided to study it in SPICE.

References :
https://patents.google.com/patent/US5032796A/en
https://i.ibb.co/XSXZhMH/vh140c-offical-schematic.jpg
http://www.muzique.com/schem/projects.htm (scroll down till "Tubetype Distortion from US Patent 5,032,796")



This is the schematic I tried. I removed the Zeners since I wanted to see if the idea could work with 9V supply and then easily incorporated into a pedal.



Here are the input sine wave at 250mv Peak, the audio resultant output and the voltage buildup on the peak cap C1



Here is how things look at start of the sine wave input at 500ms



Here is at somewhat half way of the total Cap C1 DC swing at 540ms ie 40 ms into the input sine wave


and here is when the cap C1 has attained its maximum possible DC swing at 670ms ie 170ms into the input signal.

Time constant of 22uF and 10K ohms is 0.22s however the cap is continuously discharging as well. It appears that 170ms = 3 time constants.



LTSPICE file available in case anyone is interested.

Things to do :
Change time constants, have time constant pot ??
Make DC swing greater, try to increase the difference in duty cycles from start to finish, have a "sag control pot" somewhere
Try collecting peak data from one opamp and use that to change bias on a subsequent opamp

[FiveseveN]

Don't we all want "tubeish" ?

No. And unlike in 1989, if you want accurate tube emulation nowadays it makes more sense to use DSP.
But if you want tubeish why not build a tube preamp?

There's been some discussion on the Flexwave circuit: https://www.diystompboxes.com/smfforum/index.php?topic=83014
This should also probably be on your radar: https://www.diystompboxes.com/smfforum/index.php?topic=122432.0

[Vivek]

Please teach me the way to add full images into the body of the post, rather than only thumbnails.

Thanks

[Rob Strand]

For many year, Patent US5032796A has intrigued me !

I noticed some of the later crate amps no longer load the sag parts.

[Vivek]

I noticed some of the later crate amps no longer load the sag parts.

I heard a pedal made from the AMPEG pre-amp. It demoed Metal music that was very gainy. I just could not make out the effect of the sag feature since the gain was just too high.

Maybe this sag thing is better for blues, rock, hard rock applications ?

[Vivek]

Crate QX-120

https://elektrotanya.com/PREVIEWS/63463243/23432455/crate/crate_gx-120_gx-212_sch.pdf_1.png

are there other amps that use Rubber Diodes with an integration Capacitor ?

[Rob Strand]

Maybe this sag thing is better for blues, rock, hard rock applications ?

Not sure myself.   A good test would be to disconnect the sag diode and connect it to a variable power supply.   Then try light overdrive and heavy clipping and see what range of voltages on the diode actual make an audible difference in each case.

are there other amps that use Rubber Diodes with an integration Capacitor ?

That one just looks like that have replaced the zeners with transistors.  22k vs 47k is like 3V3 vs 6v8.

The difference is 3V3 zener are soft clippers and 6v8 is hard.   The two transistors would be relatively hard. 
So I guess the changes there is a different clipper character.   The overall idea loos similar to the common crate ckt though.

[Vivek]

The difference is 3V3 zener are soft clippers and 6v8 is hard.   

Please help me to understand this.

Are the shape of the IV knees dependent on the Zener voltage ?

[Rob Strand]

Are the shape of the IV knees dependent on the Zener voltage ?

Yes,  low voltages tend to be softer than high voltages.   The flattest voltage lowest impedance is around 6.2V to 6.8V.   I can't remember if the knee region of say a 9.1V is sharper than the 6.2V.

[teemuk]

The SLM patent circuit isn't a "sag" emulation, rather it mimicks DC (bias) offset shifting (and its effects) of generic tube preanps that is result of interaction of AC coupling an asymmetrically clipped signal.

Peavey's "TransTube" circuit and Carvin's SX and QuadTube preamps are effectively based on replicating the very same effect but these ircuit does it naturally and do not require a dynamically varying DC source to control offset bias.
Peavey employs capacitively coupled single-ended transistor gain stages and introduces tube-like "grid-clipping" effect with a shunt diode (effectively enhancing asymmetry). Carvin SX series capacitively couple signals that are asymmetrically clipped by plain opamp stages, and QuadTube is a real tube circuit that just goes over the top with interstage coupling and associated bias shifts by cascading as much as eleven gain stages! (It's not about gain, it's about getting exaggerated amount of shifting).

Speaking of "Sag"... It would change overall clipping thresholds whereas the aforementioned schemes have fixed clipping thresholds but shift the bias point instead, therefore varying overall asymmetry vs. symmetry of clipping. Interstage coupling makes the effect dynamic.

[Vivek]

varying overall asymmetry vs. symmetry of clipping.

Thank you for your response, Sir Teemu !!!

Isn't that a bit like

Putting a man in a room, raising the floor so that his head hits the ceiling
versus
Putting a man in a room, lowering the ceiling so that his head hits the ceiling ?