What is the difference in sound between having more or fewer clipping diodes?

[ItsGiusto]

People talk a lot about modifying pedals by changing diodes, but does this strongly impact the tone, or just the volume/gain?

I was reading this really great post detailing the difference between different version of the Xotic BB Preamp:
https://aionfx.com/news/tracing-journal-xotic-bb-preamp-five-variants/

According to this, the only difference between the BB Preamp v1.5 and the BB Preamp Andy Timmons model (which is also the same as the BB Preamp v1) is that the Andy Timmons model has one clipping diode going each way in the opamp feedback loop, whereas the v1.5 has two diodes in series going each way.

BB AT and v1:




BB v1.5




Obviously, having two diodes in series going each way will double the forward voltage necessary to start clipping. My question is, does this have any impact on sound, other than the fact that you have to turn your gain knob up more in order to get the same amount of gain, and you have to turn the volume down a little bit? Does the tone change at all, or is it basically just a volume difference that can easily be fixed by turning a couple knobs?


Similarly, I think I remember that the main difference between the rat and the turbo rat is that the turbo rat uses LEDs to ground, which have higher forward voltage. Once again, would that change the tone, or just the gain and volume?

[lykwydchykyn]

If you had a theoretically perfect gain stage that was flat and linear across its whole gain sweep, and added theoretically perfect diodes that cut immediately on and off at Vf, then yes, it would only be a difference in output volume and headroom.

But we know how the real world is.

Many of the gain stages you see in guitar pedals are NOT flat and linear across the gain sweep.  They may have a different frequency plot at high gain than at low.  They'll also have some distortion of their own.  So if your diodes clip at low gain vs. high may impact the tone you're getting at a given amount of clipping.

What's more, diodes are not theoretically perfect.  The switch to a conducting state is a curve that varies between different kinds of diode.  They may exhibit some capacitance as well.  So this can impact the dynamics or frequency spectrum of the sound.

All of these are subtle differences, of course, but they add up sometimes to a noticeably different sound.

The best way to get a handle on it is to breadboard a simple gain circuit and try some different clipping diodes.  Sometimes it makes little difference, sometimes a lot.

[drdn0]

To my ears, it changes the distortion characteristics as much as it changes output, which could be a whole heap of factors being at play - how much of the existing waveform is being clipped, whether diode capacitance or knee is playing a role, whether you're inherently turning the volume down so it's interacting with gain stages in your amp or the rest of your signal chain slightly differently, etc

[FiveseveN]

turn your gain knob up more in order to get the same amount of gain

It's always useful to observe the difference between gain and distortion/saturation, but in this case it's essential.

The best way to get a handle on it is to breadboard

Absolutely!

[antonis]

And beware of that cap (47pF) across clipping diodes..
(its primary role is to form a LPF with R5 + Gain pot wiper/lug 1 but it also softens clipping corners..)

[ElectricDruid]

Like other people have said, there are reasons why you might get *some* tonal difference for more diodes vs less diodes, but in general they're massively over-stated.

The internet carries on like type/number of diodes is the critical factor in loads of drive pedals, and that's really not true. It makes a small difference. Altering the frequency response before or after the gain stage makes a lot more difference.

[amptramp]

Remember that with this type of feedback clipper, you still have a gain of unity for the incoming signal when the diodes are driven hard into clipping.  Once the diodes clip, the circuit gain is not as much under the control of the drive pot, it is more under the control of the diode resistance characteristics.  But hard clipping still leaves the gain for the incoming signal at +1.

[ItsGiusto]

Like other people have said, there are reasons why you might get *some* tonal difference for more diodes vs less diodes, but in general they're massively over-stated.

The internet carries on like type/number of diodes is the critical factor in loads of drive pedals, and that's really not true. It makes a small difference. Altering the frequency response before or after the gain stage makes a lot more difference.

That's been my experience, too. I built the BYOC rat clone that has 6 different clipping options a long time ago, but to be honest, many sounded really similar, because the rest of the circuit was the same. The one that maybe had the most impact was the mosfet clipping option. And of course the turbo rat LED option altered volume, but less so the tone.

Also, I understand that there would be some diode capacitance difference, but there's already a capacitor in the feedback loop in the case of the bb preamp. Is the diode capacitance difference really on the same order of magnitude as the capacitor that's already there, enough to substantially change the effective capacitance of that feedback loop?

[Mark Hammer]

Let's say one had unlimited gain, very hot pickups, and tremendous picking strength.  In this scenario, at max gain/drive, a 1+1 silicon or GE, a 2+2 silicon or GE, and a 1+1 LED complement would all end up providing something approximating a square-wave for each half cycle.  Any audible differences wouldn't really emerge unless one turned the gain down, and adjusted your picking strength.

I'm sure that, in some universe, diode type/composition would make a difference for higher frequencies than a guitar pickup is able to provide.  But, given the rather limited bandwidth of guitar pickups, the biggest difference one will hear, due to switching between diode types, lies in how easily harmonic content is generated by how one picks.  As well, since guitar strings provide an initial peak transient and then decay relatively soon thereafter, what you also hear is how long that added harmonic content lasts/lingers  )f course, since that is a threshold phenomenon, the "sound" of one diode type/number can be emulated by adjusting the amplitude of the signal they are affecting.

[m4268588]

[Steben]

yes.

The ratio of input signal and clipping treshold is essential in soft clippers.
I know people freak out when I speak about "treshold + clean signal on top" at max gain but it is IMHO still the most simple explanation soundwise.

With a giving input signal every extra treshold will make for a more distorted max gain sound. yes: at max gain. Because a higher treshold will need gain.
No clipping circuit overrules The diode characteristics as much as the feedback soft clipper.
Above the treshold the diode simply "conducts" compared to the feedback resistor. This brings gain at about unity. Whatever diodes you stack, it will be about unity above the treshold. Look at my drawing below. The ratio between unity region above the treshold (arrows) and the distortion region gets smaller and smaller with bigger treshold. It will sound more snappy, amp like compared to a creamy-or-fluffy TS style. Of course this is hindered by the output swing of the opamp which is related to the supply voltage.
A guitar signal into such a stage with a huge array of LEDS and infinite supply voltage will sound close to a hard clipper. A couple of germanium diodes will sound muffled, less tight.

[antonis]

Look at my drawing below.

I see same slope but different amplitudes..
So whats differs is volume, only..

[Mark Hammer]

Look at my drawing below.

I see same slope but different amplitudes..
So whats differs is volume, only..

Yes and no.
If guitar signal was as steady as a signal generator, you would be absolutely correct.  The thing is that guitar signal produced by strings varies a lot after you pick/pluck/strum, in both its amplitude and its harmonic content.  So while it is possible to control for what comes out of the various diode arrangements, and make them fairly equivalent, by adjusting gain along with diode number/type, that tends not to be what we do.  The result is that, when changes are made to diode complement, the segment of the changing guitar signal that falls above or below the clipping threshold also changes, such that the effect of the diodes will seem and sound different because it is being applied to this or that part of the signal.

And because strings tend to decay towards their fundamental (with a bit of lower-order harmonics in there as well), lower thresholds, or higher thresholds with more gain applied, mean that additional harmonic content (which is what clipping does) 9is being extracted from a vibrating string whose own initial harmonic content has contracted a lot.

All of this is to say that, yes, higher diode forward voltage will yield a louder output.  But it can "sound" different because of the dynamic nature of strings and guitar signal.  That is often misinterpreted as this or that diode clipping "differently".  They generally don't, unless one is feeding them much higher frequencies than a guitar puts out, but they simply do what they do, and it's the interaction of guitar signal and diode that seems to yield different sound quality.

[ItsGiusto]

The result is that, when changes are made to diode complement, the segment of the changing guitar signal that falls above or below the clipping threshold also changes, such that the effect of the diodes will seem and sound different because it is being applied to this or that part of the signal.

I'm still not sure I understand. If the diode has higher forward voltage, doesn't that mean that less of the guitar signal will generally end up clipping, which basically amounts less gain (distortion) but higher volume? Like previously it was clipping at .7 volts, and now it's clipping at 1.4 volts, so the clipping is basically the same as the clipping would be if the guitar signal were a little more attenuated?

[Mark Hammer]

Yes.  Higher clipping threshold means higher "ceiling" for the signal, and potentially higher output.
The thing to keep in mind is that the forward voltage of the diode/s is fixed.  The gain applied to the signal can be varied, and the guitar signal itself varies.  Given that, what you get depends on gain and input signal.

[ElectricDruid]

+1 agree. I think you understand it perfectly. What you described is what Mark was saying.

[ItsGiusto]

+1 agree. I think you understand it perfectly. What you described is what Mark was saying.

That's good. But if that's the case, then I think my original hypothesis is correct: that the number of diodes you have in the clipping circuit doesn't really matter (unless you run it at max or minimum gain), and that the tone you get between such variations is nothing that you can't get out of a slight twist of the volume and gain knobs. Which ultimately means that it MOSTLY doesn't matter whether you have one diode each way or two diodes each way, and Xotic trying to market the BB Preamp Andy Timmons model as some substantially different, refined take on the BB preamp, with slightly more mids, and a substantially different sound is really just marketing BS.

[ElectricDruid]

Just marketing BS?!? Really?! Well, I had no idea!!

[Mark Hammer]

+1 agree. I think you understand it perfectly. What you described is what Mark was saying.

That's good. But if that's the case, then I think my original hypothesis is correct: that the number of diodes you have in the clipping circuit doesn't really matter (unless you run it at max or minimum gain), and that the tone you get between such variations is nothing that you can't get out of a slight twist of the volume and gain knobs. Which ultimately means that it MOSTLY doesn't matter whether you have one diode each way or two diodes each way, and Xotic trying to market the BB Preamp Andy Timmons model as some substantially different, refined take on the BB preamp, with slightly more mids, and a substantially different sound is really just marketing BS.

Again, yes and no.
Sometimes you want to push an amp with a hot signal, but not necessarily a very dirty one.  Just try doing that with a Distortion+...good luck.  It's possible to do that by having low gain ahead of clipping diodes and then following it up with a gain-recovery stage to raise the level.  But it's just easier to use more diodes to raise the threshold, and permit a hotter output that way.  Of course, you'll still be limited by the maximum voltage swing possible with a 9V supply.

[Steben]

Look at my drawing below.

I see same slope but different amplitudes..
So whats differs is volume, only..

Yes same slope but different scale... Look at the 2 images in the middle.
To make things clear: I'm talking ONLY about classic feedback soft clipping...

the same gain with 2 x diode treshold will have less distortion but more output
double gain with 2 x diode treshold will NOT be the same
Is this audible? Depends on taste and amount of different ratio. But a TS with germanium diodes or a TS with 2 x double LEDS and lets say 15V supply DOES sound different. Thing is many don't do this but stick to small differences in treshold. You can't expect big differences with small changes.... It feels like saying input caps do not make difference because no one hears the difference between "tested" 10nf and 15nf .... Which even can be 12n and 13n with the right batch....
The appreciated difference is tighter clipping and less muffled TS stuff. No, I do not like the standard TS sound. There is a reason it is a great booster rather than a perfect sound on its own. Adding gain and more treshold and thus raising the point where the unity slope kicks in adds less soft clipping image.
It is slightly different but even stacking TS can get you there if the output of the first TS is set to guitar output level. But NOT if one sets the first output high... what one would do in a classic booster setup.



It is what one gets a bit too if one adds clippers in the first stage of a BluesBreaker...but I find that trick much better! Mark knows what I mean.

Let this one sink in a bit: