the sag

[Athin]

Hi,
   I was wondering is there a way of simulating sagging, making a sag effect in a box. I've noticed that even if the guitar ISN'T overdriven but there is a strong sag effect it sounds sorta overdriven, and smooth as ice [and cool as that as well]. I think I could use the technique the guys from runoffgroove used for tube preamps and apply it to tube power amps [though I think I'll use mosfets rather than jfets]. I'm not sure it'll be capable of sagging though [actually I think it won't] I'd need some sort of current limiter and I don't think a resistor would be enough [someone experienced - help me out here]. Mosfets are also had to "set up" [thir power/bias/(everything ) section] - any links with info would be appreciated.  I'd like to learn about mosfets a bit, if you know of any interesting/valuable articles - post them/links pls.The only thing I've found is mosfet follies ie. 1001 wierd ways of using mosfets, from geo I think.

[petemoore]

What about a limited current power supply, then a bunch of amplitude dependant circuitry, simply to suck current and cause sag / but not in the signal path...sketchy, but a suggestion to ponder.
 The 'side' circuit, which uses the same PS as the signal path circuit, possibly could be made to suck more current when the input signal amplitude is high [like when you're striking a note on guitar] and allow less current into the signal path's circuit...which would could [?] possibly be made to drop in output [without gating] when the current supply drops.
  :idea:   /   :?:  I don't know....

[brett]

My CD4049-based overdrive uses a 220 ohm resistor in the supply line to drop voltage and limit current.  Works quite well, but there's obviously more you could do.

cheers

[Alpha579]

would this work?
use a Jfet as a constant current source ie drain connected to supple, gate connected to source via resistor, then get supply from source...
BUT heres the twist, you use a fet as a variable resistor instead of the fixed resistor, then you use an envelope detector to control the resistance of the fet, thus changing the current dragged from the supply...
Probably completely wrong, but it was worth a shot...

[Athin]

Huh, I've named the effect wrong,  :oops:  what I AM looking for is the zero-pass-thru effect , which causes the waveform to look a bit more like a tangensoid than a sinusoid. The effect is caused only in class B operation, so baisically I've answered my question, thouhg a sag effect would be nice too. If I get any results - you'll be first to know.

[stm]

I don't get what you mean by "tangensoid" instead of "sinusoid"    Please explain.

Perhaps you mean something like "zero crossing distortion", which is found on class-B amplifiers?  In this case an antiparallel pair of diodes in series with the signal path will produce this effect.

Silicons loose to much voltage and you have no output when the signal weakens.  On the other hand, I have obtained interesting results with two antiparallel Ge's (in series with the circuit) and a 10k resistor to ground.  They add a little "grit" to the sound, yet amplitude is unaffected!

:idea: This circuit introduces more distortion when the signal weakens, as opposed to "normal" diode distortion.  Perhaps if you mix a "normal" stage and a stage like this you can get a signal that always has some noticeable distortion, irrespective of signal level.

By the way, the Boss HM distortion unit uses a pair of antiparallel Ge diodes to eliminate background noise when you are not playing, a kind of simplified noise gate.

Regards.[/code]

[Athin]

stm:
what I mean by tangensoid is the red curve on this pic: http://www.mif.pg.gda.pl/homepages/tom/images/3h_180.gif
it's the 3rd harmonic phase shifted 180 deg. it is supposed to be caused by either underbiasing or sagging.
I suppose "zero crossing distortion" is a good name for that  
the 10k to ground after the pair of antiparallel diodes, right? Any idears of getting this without Ge's? I think I'll try the "simulating the power amp with ss parts" approach first and when I get my hands on Ge's......

[stm]

OK, looked the curve.

The diode arrangement I mention produces this effect. By the way, in my last post Ge stands for a germanium diode like 1N34 or 1N60.  They are also solid-state technology, and I think they are the first thing to try. If you don't like germanium diodes, perhaps a pair of signal shottky diodes like the 1N5711 will be better than the 1N4148 or 1N914 silicons (200 mV instead of 600 mV direct forward drop).

You also mention that the third harmonic is shifted 180º, producing the peaks in the middle of the sinusoid. It is funny but true that if the third harmonic has its phase reversed 180º, it will *sound* the same (the human ear is insensitive to a fixed change in phase), but the tops of the sinusoid would look squashed instead of peaked.

Anyway, it is not the same to have the diodes in series with the signal or in parallel with it, since as I said before, the first option introduces more distortion with low level signals, while the latter produces distortion only when the signal level exceeds a certain threshold.

Good luck.

[Mark Hammer]

There is current reduction and there is current limiting.  When people describe "sag"' what they are referring to is a kind of "sponginess" from a working power supply that tries to meet current demands and just gives up or gasps.  This is different from a "dying battery" which has a hard time meeting most current requests.

Correct me if I'm wrong but current-limiting resistors just reduce how much current comes through in a linear fashion, and to the best of my knowledge they have no sort of "Oh, is THAT how much current you want?  Forget it!" relationship with batteries or wallwarts.

The Electronic Musician Fuzz designed by Charles R. Fischer (scanned and posted at GGG) uses a current regulator to control the clipping element, which may come closer to the idea of "sag" than simple series resistors., particularly insomuch as the sage does not require any particular battery state.

[stm]

The sponginess you mention pertains to a combination of different phenomena that take place both in the power supply and in the output stage of the amp:

1) The diode valve charges a supply filter capacitor. Upon a sudden power demand, this capacitor is able to provide a peak of power thanks to its stored energy.

2) Once this energy starts depleting, the diode valve and power transformer must provide the required power to keep up with the output demand. Since diode+transformer impedance is much higher than a capacitor impedance, a voltage drop occurs, and thus output voltage of the power supply drops.

3) Things don't stop here.  Power amplifiers (P.A.) with pentodes have a compression effect because the second grid is tied to the power supply voltage, and overall gain is a function of this voltage. This mean you have less gain in the P.A. and thus you have a neat compression effect!

4) In addition, since the supply voltage dropped, the clipping point of the P.A. is reduced, adding a bit more of limiting. But tubes clip gently, so this introduces some low order harmonics (mainly 3rd and 5th) and adds up to the compression.

5) The reason why this sounds good is because when you play a chord the amplifier remains at a power level that self adjusts for a while on that sweet spot with mild overdrive. In addition, the initial response given by the supply capacitor allows for signal dynamics. Also, this compression effect raises the apparent level, since after the initial attack the output signal doesn't drop too fast due to the inherent compression.

In summary, the supply sag, the compression and the limiting effects are related to the final output power on the speaker, and to the time, since initially the power supply may be able to provide a higher output.

I am working on these ideas to come up with an effect that emulates supply sag in a comprehensive manner, i.e. including a little of compression and reducing the threshold of limiting at the output. I believe this will lead to a nice warm tube emulation sound.

When I come up with something I will post it for sure.

[Eric H]

There is current reduction and there is current limiting.  When people describe "sag"' what they are referring to is a kind of "sponginess" from a working power supply that tries to meet current demands and just gives up or gasps.  This is different from a "dying battery" which has a hard time meeting most current requests.

Correct me if I'm wrong but current-limiting resistors just reduce how much current comes through in a linear fashion, and to the best of my knowledge they have no sort of "Oh, is THAT how much current you want?  Forget it!" relationship with batteries or wallwarts.

Depends on your definition of sag
When it comes to tube guitar amps the sag many refer to is generally caused by the resistance
within the rectifier tube , and this can be reasonably simulated (in a tube amp) with  (drum-roll) series resistance before, or after the SS rectification.
It is __not__ caused by under-rated power-transformers (contrary to some popular wisdom) --under-rated PT's simply die

-Eric

[Mark Hammer]

Thanks.  Nice explanation.  Very informative.