attempting to design a tube-based fuzz circuit for guitar

The fuzz effect is one of the oldest guitar effects, and still one of the best (in the opinion of some, such as me!).  It was an integral part of Jimi Hendrix's sound, where it amped up both the intensity and the delicate sensitivity of his playing style.  And many will point to Keith Richards using it on "Satisfaction"; but that, to me, is an example of one of the many ways fuzz can sound bad.  It's not a sound that works in all contexts; indeed, it tends to require the guitarist to "play to the effect", choosing riffs, chords, and intervals which respond well with the fuzz effect, and avoiding those which don't.

Fuzz is often lumped in with other guitar distortion effects, and is considered to simply be a more-intense form of distortion, on a one-dimensional spectrum that runs from mild ("overdrive", also sometimes called "edge" or "compression") through medium ("distortion"), through extreme ("fuzz").  However, there's something magical and not fully understood about fuzz, which has subtle time-varying characteristics and all sorts of depth and character, which doesn't seem to simply be a more-extreme form of plain distortion.  As a thought-experiment, I don't believe it would be possible to obtain a true fuzz tone by simply chaining two or more standard distortion pedals together.  Also, the original fuzz circuits such as the Fuzz Face (see below) did not have particularly high gain.  The distinctive fuzz "sound" starts to come into play at relatively low gain levels, though it is certainly a continuum which is traversed by increasing gain.  It seems to be something different from mere "clipping", which is the basis of the other distortion effects.

Taking a look at the Fuzz Face schematic, it looks deceptively simple.  Also, there are some "wrong" aspects to the circuit, such as the low input impedance.  This means that the Fuzz Face is highly interactive with the guitar pickups, and normally it has to be the first effect in the signal chain after the guitar; feeding it with a buffered signal tends to ruin the sound.  Some people believe that this low-impedance input is integral to the fuzz sound, but I am tending to believe otherwise.

In this blog, I am interested in analyzing the nature of the fuzz sound and what causes it.  From a practical standpoint, I would like to design a fuzz circuit which uses tubes, instead of transistors.  This is partly just to see if it's possible.  I'm not necessarily expecting to produce a "better" sound than the existing transistor fuzz circuits: I will be happy if I can simply produce a usable fuzz tone at all.  And I think there may be market interest in such a circuit if it can be created.

I have in the past tried to produce fuzz tones with tubes.  I have in fact managed to make some pretty good tones by chaining together multiple pieces of tube audio gear, particularly "hi-fi" gear -- which may offer an important clue (these circuits frequently contain negative-feedback loops in various forms, whereas traditional guitar preamp circuits seldom do).  But when I've tried to actually build a standalone circuit, using two triode tube stages in place of the two transistors in the common fuzz circuits, I have not had good results.  At best, I have created plain distortion devices, really no different than what can be produced by a high-gain tube amp.  The magical fuzz tone has been elusive.  I notice a few other people on the Net have posted schematic diagrams of purported "tube fuzz" circuits, but every one I've seen has been a variation of the same type of circuit I attempted: just two tube gain stages in series.  I doubt that any of these have produced any materially different results than I found myself.

As I now examine the Fuzz Face more carefully, I find that a crucial aspect seems to be the feedback path.  And it seems very important to analyze both the DC and the AC behaviours of this feedback loop.  It looks to me like in the DC realm, the loop reactively biases the transistors to be just barely conducting: which comports with the "gate-like" behaviour of many fuzz tones, where they seem to be clinging to the edge of cutoff.  The gain control gradually cuts out the AC component of the negative feedback, as well as increasing the AC gain of the second stage, while leaving the DC loop operating the same.  Up until now, I haven't tried to replicate this with tubes, and I haven't seen anyone else do so, either.

The fuzz face operates at very low impedance, and the feedback is essentially current feedback into the base-emitter diode.  I'm thinking maybe a tube circuit could replicate the important aspects of this behaviour, while operating at a much higher impedance in voltage-mode, and actively seeking the optimal bias points of two tube stages, which are of course quite different from transistor bias voltages.

I've started putting together a prototype circuit to try out these ideas, and I'll report on any progress or other interesting results, here.


Comments

Post a Comment

Popular posts from this blog

more changes to the schematic

Image
  ...Just keeps looking better and better. A few resistor and cap values changed.  But the biggest change is, now I'm going to use two 120V Zeners in series to drop the regulated 260V B+ rail down to 20V, so that the range of the bias control is more reasonable.  (I could, more plausibly, also use one 20V Zener to do this, but now that I have the 120V units in stock...) I guess I could put a capacitor on this newly-created voltage reference, and then feed it to both legs of the dual-gang bias pot; but I'll probably just duplicate it and keep everything separate, for (conceptual) simplicity.  I'm not initially building out the second section of this circuit anyway, until I can test the first and find out whether it actually behaves any better than the "take 2" circuit.
Read more

blue smoke release

Yesterday, testing the most recent circuit, I had a bit of a setback.  First off, the section I had modified seemed to have practically no gain.  As in, just slightly over 1; maybe 2?  Switching from bypass to enabled, I could get a little more volume if the level was really high, but there was very little difference.  And, importantly, the gain pot seemed to do absolutely nothing.  No change in gain, from min to max.  So this seems like maybe I had a circuit error, relative to my latest schematic.  Kicking in the second section (not yet modified from the older circuit), I had a fair bit of gain, just like it used to behave.  So that section was still operating "normally". But then, as I twiddled the gain and bias knobs, I seemed to hear a crackling sound, like too much DC on a pot (?), and then all of a sudden there was that dreaded, quiet little sound of something blowing up.  I cut the power.  There was a smell of too-hot, blown compo...
Read more

level shifting and DC gain

Yes, I think the Zener diode level shifters may be an important piece to this puzzle.  Without the Zeners, i.e.  with the simple resistor dividers as level shifters, I'm up against a no-win situation as regards gain.  The tubes must be operated near voltage midpoint, i.e., with large cathode resistors.  This implies gain of 1.  In this configuration, all three resistances, i.e., the plate resistor, the effective resistance of the tube plate-cathode path, and the cathode resistor, are approximately equal.  To shift the plate voltage down to the proper grid voltage, requires at least a 1:1 resistor ratio, i.e., gain of 0.5. There seems to be an inevitable no-win voltage-scaling situation, in the absence of Zeners or other non-resistor level shifting.  Operating the tubes so that the target voltage at the grids is below midpoint to increase the DC gain above 1, means the plate output voltage must then be scaled down with a more-aggressive resistor ratio, ...
Read more