Question on AMZ Clipping Saturation control

[bipedal]

Am thinking of adding the following clipping saturation control at the end of a simple transistor booster circuit...
, discussed at http://www.muzique.com/lab/sat.htm.

I'm not entirely clear: at one end of the pot's travel, will this arrangement and the 10k pot value effectively prevent the diodes from impacting the signal path / sound?  Or will it simply tame (but not entirely remove) the clipping?  Would a larger pot value have any noticeable impact?

I like the sound of the clean booster as is, and want to preserve that.  Just exploring options for a single circuit design that can range from the booster's pure output tone, with the ability to dial in a slight bit of diode clipping "grit" via a pot (rather than a switch that connects/disconnects the diodes from the signal path...).

- Jay

[Mark Hammer]

I'm not entirely clear: at one end of the pot's travel, will this arrangement and the 10k pot value effectively prevent the diodes from impacting the signal path / sound?  Or will it simply tame (but not entirely remove) the clipping?  Would a larger pot value have any noticeable impact?

I like the sound of the clean booster as is, and want to preserve that.  Just exploring options for a single circuit design that can range from the booster's pure output tone, with the ability to dial in a slight bit of diode clipping "grit" via a pot (rather than a switch that connects/disconnects the diodes from the signal path...).

Let's do the "gedanken" (thought) experiment.

Imagine the wiper of the pot is moved so that it is near the "bottom" of the pot as shown.  What will you have?  Well, the diodes will go to ground as always, and pretty much exactly as you see them in a Distortion+.  The 1k resistor and the full resistance ofthe pot will be placed in parallel and function like the 10k series resistor in that same classic circuit, except that the effective parallel resistance of 1k and 10k will be 0.91k instead of the 10k seen in the Dist+.  As you move the wiper towards the upper end of that pot (as shown in the drawing) there is a voltage divider formed by whatever is between the wiper and the diodes, and the 1k plus whatever is on the other side of the wiper.

Let's move the wiper to the exact middle (and assume it is a perfect linear pot).  We will see 5k between the wiper and the diodes.  We will see 1k and 5k in parallel between the wiper and the input (0.83k).

So what is happening?  the signal hitting the diodes is attenuated just a little by the voltage divider action of that pot/resistor network, but never that much.  The 1k fixed resistor sees to that because it sets limits to the parallel resistance of pot and fixed resistor.  The larger change is the resistance from the wiper to the diodes.

So what would happen if we changed things a wee bit?  Let's use a 4k7 fixed resistor and 47k/50k pot.  When the pot  wiper is moved fully to the diode side, we will have the diodes tied to ground directly from the wiper, as per above.  The fixed resistor and pot will have an effective parallel resistance of 4.27k (using a 47k pot), less than what we see in a Dist+.  We can call this the "classic" position of the control.  As we move the wiper, though, we add more resistance in series with the diode path to ground, reducing the clipping effect of the diodes.  How much of a voltage-division effect are we adding, though?  Well, as the first instance, the fixed resistor (the lower value) constrains how much resistance there will be between the input and the wiper.  But since 4k7 is a larger value than 1k, for at least a somewhat larger portion of the pot's range, thje resistance n the input side of the wiper will be larger than the resistance on the diode side.  This will have the effect of softening the signal hitting the diodes, and reducing the clipping.

Note as well, that the constraining effect of the fixed resistor doesn't have to be just on the input side of the wiper.  You could easily stick a second fixed resistor on the diode side of the wiper to constrain how much resistance change occurs on that side too.  This would be particularly useful if you found yourself wanting to use this mod but lacked a low-ish value pot.

So, the answer to your question is yes, you can use other pot values, as well as other fixed resistor values.  However, they will alter the degree to which the signal arriving at the diodes is altered, as well as the manner in which resistance change on the diode side of the wiper is distributed.

[StickMan]

Isn't that circuit going to depend to a large degree on the impedance of whatever is on the other side of output?

For instance, if you plug it straight into an amp with a 1Meg input impedance isn't it going to clip a lot harder than if you plugged it into something like a fuzz face with a low input impedance?

[bipedal]

Hello Mark -- as always, thanks very much for the comprehensive response.  I invariably learn something new and useful from your comments on the forum.

I may have articulated my original question in a more complicated manner than intended...

If we look at the simpler version of this control: , the resistance between the output signal path and the diodes-to-ground can vary between 0 (diodes "all in") and 10k.

Now, if I were to use a switch (rather than a pot), I could make or break the connection between the signal path and diodes-to-ground. Inserting 10K resistance between the signal path and the diodes is certainly not the same as breaking the connection, but in this example, I'm wondering if 10K resistance would effectively mimic (at least in an audible way) a break, so to speak.

In other words, if the pot is turned all the way to one end -- with no resistance between signal path and circuit out, and 10K resistance between the signal path and diodes-to-ground -- would any of the signal still "see" the diode loop at all?

As you can tell, I'm seeking to keep things as clean as possible at one end of the saturation pot's movement; i.e. an arrangement that allows the user to minimize signal alteration caused by the diodes, or turn up the saturation pot to add in just a bit of diode clipping if desired.  At this point, it's pure guesswork for me whether 10K is sufficient to accomplish this or if I ought to consider a much higher pot value.  (The answer is probably somewhat dependent upon the signal level prior to the voltage divider, no?)

Cheers,

- Jay

[JDoyle]

Hello Mark -- as always, thanks very much for the comprehensive response.  I invariably learn something new and useful from your comments on the forum.

I may have articulated my original question in a more complicated manner than intended...

If we look at the simpler version of this control: , the resistance between the output signal path and the diodes-to-ground can vary between 0 (diodes "all in") and 10k.

Now, if I were to use a switch (rather than a pot) to make or break the connection between the signal path and diodes-to-ground, that resistance effectively changes between 0 and infinite. Inserting 10K resistance between the signal path and the diodes is certainly not the same as breaking the connection, but in this example, I'm wondering if 10K resistance would effectively mimic (at least in an audible way) infinite resistance, so to speak.

In other words, if the pot is turned all the way to one end -- with no resistance between signal path and circuit out, and 10K resistance between the signal path and diodes-to-ground -- would any of the signal still "see" the diode loop at all?

As you can tell, I'm seeking to keep things as clean as possible at one end of the saturation pot's movement; i.e. an arrangement that allows the user to minimize signal alteration caused by the diodes, or turn up the saturation pot to add in just a bit of diode clipping if desired.  At this point, it's pure guesswork for me whether 10K is sufficient to accomplish this or if I ought to consider a much higher pot value.  (The answer is probably somewhat dependent upon the signal level prior to the voltage divider, no?)

Cheers,

- Jay

Put a JFET buffer before and after the diode circuit and DC isolate the diodes with caps; use 10k Source R for the first buffer, a 100k pot for the diodes, and a 1 Meg gate resistor for the buffer after. This means that each stage is now designed to avoid loading the previous circuit or being loaded by the subsequent circuit. That will pretty much take any variable out of the equation and do exactly what you want.

Regards,

Jay Doyle

[Caferacernoc]

"As you can tell, I'm seeking to keep things as clean as possible at one end of the saturation pot's movement; i.e. an arrangement that allows the user to minimize signal alteration caused by the diodes, or turn up the saturation pot to add in just a bit of diode clipping if desired.  At this point, it's pure guesswork for me whether 10K is sufficient to accomplish this or if I ought to consider a much higher pot value.  (The answer is probably somewhat dependent upon the signal level prior to the voltage divider, no?)"


All other parameters being equal, a larger value pot will make it easier to "shut off" the clipping.

[StickMan]

Bipedal,  I think you need to remember that whatever is "Output" isn't going to free of resistance.  So even if the wiper is all the way at the top, some portion of the signal will still go through the diodes.  As to whether there's enough to overcome the forward bias of the diodes is going to depend on the strength of comes out of "In".

As Caferacernoc said, increasing the value of the pot is going to increase the chance that you can choke off the signal going to the diodes.

[Mark Hammer]

How much of the clipping one shuts off though, will always depend on the absolute average amplitude of the signal, and the forward voltage of the diode.  So, the effect of that resistance in series with the diodes, unless infinite or darn close to it, will be different for a 3v P-P signal using Schottky diodes, than it would be for a 1.2V p-p signal, using 1N4001s.

Bipedal,  I think you need to remember that whatever is "Output" isn't going to free of resistance.  So even if the wiper is all the way at the top, some portion of the signal will still go through the diodes.  As to whether there's enough to overcome the forward bias of the diodes is going to depend on the strength of comes out of "In".

As Caferacernoc said, increasing the value of the pot is going to increase the chance that you can choke off the signal going to the diodes.

Quite right.  If the diode "needs" to see 560mv before it starts conducting, and the pot divides the input signal down such that it would rarely pass 400mv, the odds of the diode conducting that signal at the wiper are reduced dramatically.

[Khas Evets]

If you want to keep it simple, use a switched pot. When it's switched off, break the diode connection.

[John Lyons]

Interesting topic, and uselful.
In the first example the 1k is tapering the 10K pot to a reverse audio taper correct?
Putting the 1K from the bottom (diode side) to the wiper would be simulating an Audio taper.
This is just adjusting the "feel" of the pots travel.

John

[JDoyle]

How much of the clipping one shuts off though, will always depend on the absolute average amplitude of the signal, and the forward voltage of the diode.  So, the effect of that resistance in series with the diodes, unless infinite or darn close to it, will be different for a 3v P-P signal using Schottky diodes, than it would be for a 1.2V p-p signal, using 1N4001s.

Bipedal,  I think you need to remember that whatever is "Output" isn't going to free of resistance.  So even if the wiper is all the way at the top, some portion of the signal will still go through the diodes.  As to whether there's enough to overcome the forward bias of the diodes is going to depend on the strength of comes out of "In".

As Caferacernoc said, increasing the value of the pot is going to increase the chance that you can choke off the signal going to the diodes.

Quite right.  If the diode "needs" to see 560mv before it starts conducting, and the pot divides the input signal down such that it would rarely pass 400mv, the odds of the diode conducting that signal at the wiper are reduced dramatically.

If the pot value is at least 10x the output impedence of the preceding stage, and the stage following the diodes has a input Z of 10X the pot value, then I think the easiest way to think of the 'Saturation' control is as an extremely crude mixer which is completely out of the circuit when the signal on the input to the pot is below the diode's threshold. When the wiper is farthest away from the diodes (with the signal level this low, the pot has no effect and can be considered an open circuit), and when the signal is above the conduction threshold, the wiper will then be able to go from fully clean (the output of the previous stage divided down by the pot or: [(Vin)x((ZpotB)/(Zout)+(ZpotA)); where Vin is the output of preceding stage, ZpotB is the resistance from the wiper to the diodes, Zout is the output impedence of the preceding stage, and ZpotA is the resistance from the wiper to the input of the pot]) to fully distorted when the wiper is directly on top of the diodes. The pot is basically mixing the two signals, clean and distorted, in the ratio set by the divider action of the pot and it's interaction with the various impedences.

Think of it like this, if you have the above conditions met with the input impedences each being 10x greater than the output impedences of the preceding stage, you have the diode circuit isolated enough where you can examine the exact action without worry of signal impedences confusing your thought process. Then you can easily infer the changes that different impedences around the circuit cause. Think of the circuit you want to analyze as a 'Black Box' with certain fixed properties, ( i.e. If Vin<0.6V then Zin=Infinity, if Vin>0.6V then Zin=Zpot, etc.), then add in the source and load impedences and figure out what each does to the Voltage at the two signal nodes we care about, the input to the pot and the output at the wiper of the pot. Once you have this generic version of your ideal circuit, you can then start including the error factors in your thinking such as the effect of the current drive on the diodes and how they conduct (the pot limits the amount of current through the diodes), the mixing action of the pot (way, way, WAY out of my mathematical depth here...), etc.

When thinking about circuits, for nearly any type of circuit EXCEPT one specifically designed to create distortion, it will be perfectly OK 99% of the time to think of and design for 'ideal' devices as outlined in any beginner EE textbook, if you stay within the limits of any real world device as outlined in it's datasheet. It is when you ARE specifically designing for distortion that that process goes out the widow and you start looking around the edges of each device's many limits. But until that limit is reached and onset of distortion occurs, for our purposes and in the audio realm, assume that the device is ideal. Once it starts to distort, you are on your own...