Field-Effect Transistors

There are 3 main kinds of transistors - Bipolar, Junction Field-Effect (JFETs), and Metal-Oxide semiconductor FETs (MOSFETs). Both come in N and P varieties with opposite polarities. This page is mostly about JFETs and compares them to bipolar transistors to illustrate why JFETs work better for simple preamp stages. Later on down this page I get into a few MOSFET-based things I've been working on.

There's a lot of talk about why tubes sound better than transistors, often attributed to how tubes clip gently and transistors clip with hard edges. Roughly accurate but there's more to it than that - transistors don't necessarily clip hard and tubes can clip with hard edges when pushed. Rather, the main difference is a tube (and a JFET) is a voltage-driven device that's fairly linear when biased properly, whereas a transistor is a current-driven device that tends to be more nonlinear and usually requires negative feedback to correct, that's where the hard clipping edges come from, especially in the extreme case of opamps.

Here's a simple bipolar transistor preamp stage with varying amounts of drive...




Before clipping it looks ok but the distorted waveform is.. well yuk. The top of the wave rounds and the bottom squares, and when distorted the base rectifies the input signal causing the duty cycle to shift, basically losing power. Some duty cycle shift is OK but with this circuit it only gets worse the harder it's pushed and there's no way around it because the base is biased positive to the emitter, any distortion that causes the emitter to stop following the input signal turns the base into a diode. The resistor values are fairly high in this example but reducing all the resistors by a factor of 10 (and increasing the caps by 10) makes little difference other than increasing the current drain.

Here's an almost identical circuit using a JFET instead of a transistor...




Much better! There's still a hint of rounding on the top of the wave but it's partially because it's not biased perfectly centered. Compared to the bipolar transistor, the FET stays cleaner until it reaches clip then it flattens more or less the same on both sides, and there is no duty cycle shift. Eventually the gate will conduct and shift the duty cycle but it takes over a volt of input as the gate is negative to the source, the signal has to exceed the bias point plus the juction drop before rectification can occur. Tubes do a similar thing (to a greater extent as the signal voltage vs rectification voltage is greater) but the solution is the same - add some resistance between the output of a stage and the gate of the next stage (or in series with the gate) so if the gate conducts at least it won't load the previous stage and cause further issues. But keep in mind that resistors are noise makers by the laws of the universe, use only as many ohms as needed.

The 2N4393 shown in the simulation is available but the common J113 has practically identical specs (Vgs threshold from -0.5V to -3.0V) and works about the same in this circuit. To minimize unit-unit variation I made the source resistor the same as the drain resistor, typically the source biases about 1.3V above the gate with a typical variation of +/- 0.4V (but not guaranteed!), because the source and drain resistors are the same value this limits the drain bias shift to the same amount. There will be some unit-unit variation but it's less than in typical JFET example circuits that omit R3 and use a smaller resistor for R4, those circuits usually require a trimmer for R4. The gain is determined roughly by R6 divided by the parallel combination of R4 and R5 (minus a bit) so if R4 is a trimmer it's going to also change the gain (and also the output level). Also note that dialing up more gain by reducing R5 increases the clean distortion and causes more pre-clip rounding on the top side of the wave. Biasing the gate positive with a larger source resistor limits the output swing but it's worth it to reduce variability, and if trimming is needed it can be done using R3 or R4 with less gain variation. It also simplifies the design process - whatever the supply voltage is, divide it by 3 and bias the gate to that voltage minus the FET's typical operating gate voltage, which for the J113 is about 1.3V. So 1.6V on the gate results in 2.9V on the source, 2.9V across the drain resistor and 3.1V across the FET, in the ballpark. It's a bit more complicated than that, the source follows the signal (unless R5 is 0) so the bottom clip point gets pushed up by the signal on the source - so having a bit more voltage across the FET is a good thing - but there's also loading effects which tends to need less voltage across the FET to balance out the clipping.

Now to analyze the distortion characteristics using FFT plots. Before distortion there isn't much difference...



For both the transistor and the FET the 2nd harmonic is about 32db down, or roughly 2% distortion, the other harmonics fall off about the same. Running clean both these circuits will sound about the same - transistors can be "warm" sounding, at least much better than chips.

Running the circuits into clipping better shows the difference...

 


Besides the FET waveform looking better, the higher harmonics drop off more rapidly. The transistor's spectrum looks harsh and this is fairly mild clipping.

Here are the LTspice (version 4) files for the above simulations...
transistor.asc.txt - the basic bipolar circuit
transistor1.asc.txt - modified for FFT analysis
tfet.asc.txt - the basic JFET circuit
tfet1.asc.txt - modified for FFT
(remove the .txt extension or copy to an .asc file)

LTspice is available from Linear Technology's software page, I use version 4 (under wine) but a newer version is available. Simulation isn't perfect but for what it does it's almost like magic. Just keep in mind that it doesn't model parasitic effects from the circuit layout and the components are assumed to be perfect unless otherwise specified. Usually if a circuit works in LTspice it'll work in the real world although sometimes it might need a few tweaks to compensate for effects that are not modeled. For audio usually the simulation is spot-on.

A 3-way Tone Network using a JFET

Disclaimer - I haven't actually built this circuit... it "should" work pretty much like the simulation but the component values in the tone network will likely need "by ear" adjustment depending on the application.

This circuit implements a 3-band cut/boost-type tone control network using a JFET for the gain stage. With the controls centered the gain is roughly unity (minus about a db). The bass control (on top) has a range of about +/-10db at 80hz (more at lower frequencies), the mid control has a range of about +/-7db at 1khz, and the treble control has a range of about +/-10db at 7khz (more at higher frequencies). This type of tone network probably wouldn't make for good guitar amp tones by themselves (too flat, guitar amps typically scoop the mids), the application I had in mind is putting it after an overdrive circuit that's already equalized to sound about right to allow better control over the overdrive tone.

The circuit has a fairly low input impedance (about 15K) that varies with frequency so it should be driven from a low impedance source. For high impedance sources use a buffer such as this circuit...

              .----------*------< 9V
| |
2.2M |
| |---'
in >--0.01u---*----->| J113
| |---.
| |
| *---1u---> to EQ
1.8M |
| 10K
_|_ _|_

Because the circuit cuts and boosts, keep the input level fairly low (around 100mV RMS) to avoid clipping at extreme settings. The circuit has a fairly low output impedance (about 5K) but if driving an input with an impedance of less than 100K increase the size of C7 to avoid bass loss.

Here's the circuit response with the tones flat and driven slightly into clipping...


Maximum undistorted output is roughly 1.7V RMS, FET variation might reduce this somewhat if not biased perfectly centered. The bottom bias resistor is slightly bigger than the previous examples because the higher current means less gate-source voltage difference and also because the source resistor is fully bypassed so it won't be going up and down with the signal. Note how the gate signal is distorted as the feedback corrects the distortion of this configuration. Controls U1 to U3 are 300K linear-taper controls, 250K linear controls should also work OK. In the simulation the controls ended up being backwards, wiper=0 for all the way up and 1 for all the way down, the correct orientation is for the wiper to travel to the left (towards the input) as the control is increased.

C4 sets the bass control frequency, C6 sets the mid control lower frequency, C5 sets the mid control high frequency, and C2 and C3 control the treble control frequency (could just be a single capacitor in series with R10 but this arrangement permits setting the cut and boost frequencies separately). R5 and R6 set the bass control range, R8 and R7 set the mid control range, R10 sets the treble control range. R9 is fairly large compared to the other resistors to minimize interaction.

Here are AC response plots for various control positions. Flat and with the bass control 90% up (sim wiper setting at 0.1)...

Bass at 10%, mid at 90%...

Mid at 10%, treble at 90%...

Treble at 10%, the last plot is with the bass at 90%, the mid at 10% and the treble at 80%...


Here's the tonecb_ac.asc.txt simulation file and the potentiometer model it needs to run.

An Old-Style Guitar Preamp using JFETs

This is similar to the preamp used in ShoBud amps with more of a rock-n-roll twist - I used a 50K mid control (with an extra capacitor) instead of the usual 10K, and added a master volume between the 2nd stage FET and the output buffer. The original ShoBud preamp had no master and the base of the output buffer transistor (a regular bipolar NPN) connected directly to the drain of the 2nd stage, an arrangement that wasn't designed to be clipped. Although provisions are made to let it get a bit dirty, it's not an overdrive preamp and is more like the clean channel of a typical tube amp. The output of the circuit can be run into any typical power amplifier that needs a volt or less for full output. Or run through a reverb circuit or an effects loop first. This is another circuit I've only "built" in simulation in this exact form but I've used variations of this many times over the last decades and it sounds very much like a tube amp.

Here's the circuit running clean and clipped...



The circuit runs at 18V, the simulation schematic shows 50V because that's what's typically available from the power amp section, reduce R20 if running from a lower voltage to ensure that the zener diode conducts and properly regulates the supply. The simulation shows 2N4393 but almost any N-channel JFET with a reasonably low Vgs threshold should work - J113, MPF102, 2N5457 etc. If clipping is off-centered tweak the values of R5 and R12 (with a J113 model I needed 75K for symmetrical clipping). R6 and R13 set the gains of the first and second stages, reduce for more gain. I configured the first stage has less gain to avoid overloading the input stage.

The tones resemble a normal guitar amp tone stack but I used somewhat different capacitor values, made the mid control bigger to make it more useful for boosting the gain. The extra capacitor C13 keeps it from boosting the highs along with the mids. C6 and R9 add a high treble boost at lower volume settings, R9 controls how much brightness is added. R9 is often a switch for full bright. The simulation shows some rolloff above 8khz, not sure what's up with that since there are no added high-cut caps (probably the "miller" capacitance of the 2nd FET). Guitar speakers cut off around 7khz so it doesn't matter much but this might be one of those areas where simulation deviates from reality.

Here's the AC response with the controls centered (20% log controls)...


The following plots show the response with the treble, mid and bass each at 10% and 90% with the others centered...







Something resembling flat is obtained with the treble at 50%, the mid all the way up and the bass at 20%...


Here's the fetgpre_ac.asc.txt LTspice simulation file, requires the potentiometer_standard model.

Dealing with unit-to-unit variation

JFETs have very loose specs when it comes to gate threshold, for a 1uA threshold the J113 has a specified Vgs from -0.5V to -3.0V. Other units are no better, it's just something the designer has to deal with when using JFETs. Biasing the gate positive and using a big source resistor helps (in that the circuit will usually work) but it does not eliminate the variability, especially when running from 9V as with a guitar pedal.

Here's a J113 model that can be used with LTspice to explore the effect of varying gate thresholds...

.MODEL J113 NJF (VTO=-1.29 BETA=9.25964E-003
+ LAMBDA=3.03839E-002 RD=1.30170 RS=1.30170 IS=9.86870E-016
+ CGS=1.05000E-011 CGD=1.20000E-011 PB=5.04493E-001 FC=0.5)

To use this model select the text icon and select spice directive, paste the above text into the text box and place anywhere on the schematic, then the 2N4393 text can be changed to J113 (right-click and change the part label text itself, it won't show up in the part select dialog). The default VTO is -1.29 but this can be changed to explore the effects of variability.

Here's the simple preamp circuit with the default threshold...

Here's the waveforms with the VTO=-0.9 and with VTO=-1.7 (about +/- 0.4V)...


...yea that's a problem. Both will "work" but will sound different.

Here's the guitar preamp with the J113 model (source resistors changed from 82K to 75K)...


Here's the response with VTO=-0.9V and VTO=-1.7V...



...not as bad because it's running at twice the supply voltage, but still some variation.

So what to do? One thing that can be done is replace the source resistors with trimmers, especially the last stage where the effect is most noticeable. But that adds cost and there's the additional labor step of connecting it to a signal generator and scope to make the adjustment. Another way to compensate is to grade the FETs - Here's a test circuit that replicates the bias of a preamp circuit...

         G O-----------*---2.2M----------.
FET | |
under D O-----------|----*----47K-----*------O +
test | | | 9V battery
S O--47K--. 470K `-> - + <-' .--O -
| | to meter |
`---*---------------------'

Since the source and drain resistors are the same, for proper unloaded response the meter should read about 3V, but typically you'll need a bit more voltage to account for loading effects, 3.1V or so with a 150K load. Regardless go through a bunch of FETs to see what the average is then in your circuit size the top bias resistor (or the source resistor) to produce the desired clip characteristics. Mark all the FETs close to the average say yellow, these are the "good" ones. If the meter reading is less the FET draws less current, paint it green and to optimize it use a smaller resistor for the top bias resistor. If the meter reading is greater paint it red and use a bigger resistor to optimize. Or just use the yellow ones for the last stage and put the reds and greens earlier in the preamp where exact clipping symettry isn't an issue.

Using JFETs for Guitar Overdrive

My preference is real-tube overdrive, but JFETs do sound quite nice when overdriven and because of how they work can be dropped into circuits that resemble actual overdrive amplifier preamps. The following is a recent variation of a circuit I've been using for years (actually about three decades), and other than using JFETs running at low voltage, is fairly similar to the tube amp mods I was doing back in the late '80's and early '90's at Shiloh Music and outlined in my Amp Mods document. Basically it's a 3-stage overdrive with a compensated gain (when low boosts highs, when high cuts lows) between the first and second stage, a fixed shelving low-cut filter between the second and third stage, followed by passive high filtering and a kind of funny 3-way tone network, the volume control and a unity gain buffer. Here's the basic circuit showing waveforms at various points...


Other than the tone control network and stage details, it is similar to the overdrive circuits I was making in the '90's and ever since. Besides using JFETs running at low voltage, the biggest difference between this circuit and my old tube amp mods is biasing the gates somewhat positive to allow using large source resistors to minimize unit-to-unit variation as previously discussed, thus it requires extra coupling capacitors. Some of my older JFET versions had ground-referenced gates but that meant having to use source trimmers and adjusting each stage, that got old.

Omit C1, C6, R4, R12, R19 and the gain-dropping resistors in series with the bypass capacitors (tube amps need more gain to get to the hundreds of volts needed to clip) then the circuit is almost identical to my overdrive tube preamps, which are in turn similar to something like a JCM800 master volume amp with the bypass caps stuffed and a few caps added to tone down the highs. It's all an evolution of tone!

So what makes this circuit good? As a guitar player I'm very picky about my tone - I want "clean" overdrive where if I back off my guitar volume it cleans up, doesn't get harsh when I push it, gives me lots of gain to get that magical harmonic feedback, and not mud out when I play chords... I want to hear every note of the chord no matter how high the gain is cranked. The secret to achieving this is related to physics.. sonically it sounds like all the notes on a guitar are about the same volume, and energy-wise they are. But clipping is about amplitude and that's not energy, to sound the same volume a note that's half the frequency requires twice the amplitude, and that's why many overdrives mud out - a flat preamp will clip the low notes before the high notes and that's no good because then you can't play chords without the lows overpowering the highs. Even worse, if flat after the clipping stage it'll sound harse, even smooth-clipping devices like JFETs and tubes create too many high harmonics. The basic trick isn't that hard - cut the lows and boost the highs before the clip stage, then boost the lows and cut the highs afterwards. That way the low and high notes have roughly the same amplitudes so they clip together instead of clashing, the high cut afterwards gets rid of the mosquito sound, and the low boost afterwards restores the lows that had to be cut before the clip to keep it from mudding out. If done right, when not clipping the frequency response is fairly flat with plenty of highs despite the harmonic filtering, and when driven it's nice and smooth with roughly the same harmonic content.

That's the basic idea but there are other tricks. Guitars (particularly those with single-coil pickups) output a significant amount of signal above the normal hearing range (ultrasonic), this needs to be filtered out or else those inaudible signals will distort together and make ugly sounds. That's what causes "spatter" when picking high notes hard through some overdrives, that static-like sound that cleans up when the guitar's tone control is backed off. Fixing that is usually as simple as adding low-value capacitors to the gain stages to limit their gain at very high frequencies. Another trick is the compensated gain - the amount of low-cut and high-boost depends on how much gain is applied, if the pre-clip EQ was fixed then at low gain it would sound thin and at high gain it would sound harsh and cause high-frequency feedback issues (squeal). So the gain control is designed to cut more lows at higher gains, and boost more highs at lower gains. Basically, it's a medium-value capacitor feeding the gain control, a resistor from the wiper to the ground, and a small-value "treble bleed" capacitor in series with a resistor from the control hot to the wiper. The feed capacitor is chosen so at lower gains there's sufficient low-end response and at higher gains the resistor from the wiper to ground causes more low rolloff. The bleed capacitor value is chosen so that at low gains it produces enough treble boost to counteract the harmonic filters after the clip stage, the resistor in series limits the amount of treble boost. These 4 components define the pre-clip EQ at different gain settings, usually I select them by ear to get the response I want.

The fixed mid/high boost between the 2nd and 3rd stages helps to keep the overdriven tone tight and focused, especially at high gain settings where the 2nd stage also clips. R16 sets the amount of low-frequency attenuation, C10 sets the turnover frequency, C7 sets the lower low-frequency limit under which lows are rolled off without shelving. If desired R16 can be replaced with a "focus" control to vary the amount of low frequency attenuation, a 1M control varies the attenuation from none to about 9db. This is about the same circuit found in the JCM800 master volume amp and other similar designs.

Another undesirable effect is if the signal forward-biases the tube grid or JFET gate it acts like a diode and both shifts the bias point and loads down the previous stage. This isn't as big of an issue with JFETs as with tubes due to the lower voltages but the fix is the same.. put some resistance in series with the gate or grid. That doesn't do much for the bias shift of the stage being driven but it helps control the loading effect, which can be worse (essentially shorting the output above a certain voltage). A bit of bias shift under duress sounds good, gives it that "it's about to blow" sound, vary R10 and R18 as desired or omit. The resistors also serve another purpose - miller effect, which attenuates high frequencies. With tube amps sometimes I have to take additional measures besides the series grid resistors - EL34 amps in particular are quite sensitive to being fed an asymmetrical signal under overload conditions - the wires in the tubes often light up cherry red on one side! to fix that have to add diodes to counteract the bias shift.

Clipping symmetry is controlled by R17, R19 and R21 which determine the 3rd stage bias point, the values are chosen to clip on both sides roughly at the same time for near-symmetrical clipping. I've never been a fan of asymmetrical/uneven clipping, (besides potentially blowing up EL34 tube amps at high volume) to me it just doesn't sound right, like one of the output tubes is blown. But to each their own, harmonica players love uneven clipping and many popular distortion designs use that technique, I suppose because it increases the 2nd harmonic. If that's your thing R17 can be replaced with a 1M linear taper control or trimmer but expect it to make static noise when adjusted, controls don't like DC on them.

With the core overdrive stages tamed, now have to add a post clip filter network to keep the harmonics under control and provide tone controls to balance the tone when switching between clean and dirty. In this circuit the harmonics filtering is done by C13, C21 and C22. Note that capacitors placed directly on a stage output (drain or plate before or after a coupling cap) have a different effect than capacitors with resistance between the stage output and the capacitor - when placed directly on a stage output then the effect of the capacitor is modulated by the high/low output voltage excursion and produce much more filtering on the high part of the waveform than on the low part because when high the output impedance equals the drain resistance and when low the output impedance is near zero (whatever the source gain-setting resistor is). This effect is greater with JFETs than with tubes, which don't pull all the way low. For the very-high-frequency filters on the 1st and 2nd stages this doesn't matter much because those stages aren't clipping as much but definitely makes a difference on the final clip stage. In this design C13 is the one that mostly shows the effect (and is small enough to not matter), C21 is affected but only over a 2-1 frequency range, and C22 and the tones have no asymmetrical filtering effect at all.

The low control provides a low-cut when turned down, and when increased blends in a capacitor to provide bass boost. The high control is similar to a guitar tone control, and the mid control is kind of like a mild notch filter. The controls don't have a great deal of range because I wanted a pedal that sounded OK with just about any setting (I don't like having to "find" the tone) but still had enough range to balance with the bypass sound and compliment the amp tones. There are a huge variety of tone control circuits that can be substituted here, including the 3-band active EQ circuit presented earlier.

Here's an animated GIF of a similar circuit showing transient waveforms and frequency response as the controls are adjusted...


The waveforms are similar to what I get from a tube amp with a similar design.

I reflected a bit before posting this overdrive design as it's similar to an overdrive pedal I've been making at Superior Music. But frankly it's not getting much exposure and love lately so I feel the need to explain why this design rocks, and to do that I need to reference actual circuits or it's mostly just a bunch of marketing hand-waving. This makes it real, it can be built and verified. I do need to record some clips of playing through it but if you've heard me play live or heard one of my amp mods, it sounds kind of like that. However - sound clips don't convey how it feels - with designs like this the player can control how it sounds by how the notes are played.. soften up and it cleans up, closer to the bridge for higher harmonics, play a 5/1 inversion and it generates an octave-down subtone that sounds like a monster chord, use your palm lightly to generate high harmonics or more damped near the bridge to accentuate the bass. This design doesn't sound exactly like my tube amp designs (after all it ain't hot tubes!) but JFETs are similar enough to tubes to be able to use the same general principles to achieve a similar effect.


Some of my amp mods for comparison

Here are a few of the tube amp mods I've done over the years...




The first design is the modified Fender Princeton I've played through for years - since drawing that schematic I added a transistor unity-gain buffer (emitter follower) between the master volume and effects send jack to avoid loosing highs when using with a pedalboard. It doesn't really have clean/dirty switching, in "lead" mode it boosts the gain of the second stage and inserts a level drop and extra filtering after the third stage. The gain control has low-cut compensation but no "treble bleed" because that would make it too bright in clean mode. For the same reason it doesn't have the mid-high boost between the 2nd and 3rd stages, just a simpler low-frequency rolloff. Despite that, the JFET overdrive sounds more like this amp than not.. in fact one of the reasons I made this thing was so that I could get the sound I liked from just about any amp.

The remaining mods more closely resemble the 3-stage JFET OD circuit with variations. For the 75 mod had to figure out what to do with all those pull switches. The Bassman mod doesn't show a treble-bleed network but ended up adding one later then the customer brought it in for adjustments - besides that put the clean gain before the overdrive gain to make it easier to dial in mild overdrive (not everyone likes balls to the walls) and added "speaker emulation" filtering to the line out. The Bandmaster mod is close to the JFET OD design until it gets to the tones. Both the Bandmaster and Bassman mods have interesting tone networks I would mind trying again in a pedal design. The component values for the various elements vary quite a bit, when I do a mod (or a new pedal design) I'll start out with the basic topology then tweak it by ear until it sounds the way I want.


More JFET Overdrive Stuff

Here's an older version of the 3-stage JFET overdrive design, and something I haven't had a chance to make (yet)...

 

The 8-3-16 version is simpler and more closely resembles the tube amp designs from which it came from. I wasn't crazy about how the low control worked - it could only cut lows, not boost them - and it had asymmetrical harmonic filtering so kind of had an edge to it. The latter issue (if it is an issue, tube amps do that all the time) can be addressed by lowering the value of the 1000p on the stage 3 drain and adding another small-value capacitor after the 47K to ground. Fixing the low control is trickier but could increase the value of the 0.01uF coupling capacitor to say 0.033uF to 0.1uF, adding a 100K resistor in series with it, changing the 22K to 47K then adding a 0.0022uF capacitor in parallel with the low control, then it would more resemble the later design.

The dark blue design is... interesting. Starts off with a normal JFET preamp and compensated gain control, a low-cut "focus" control, then it gets a bit different. The stacked JFET's are what is called a "mu-amp", or at least my take on it. The original circuit was published in the '70's in the app note AN-32 by National Semiconductor but it used parts that are no longer available, if made as-is with modern JFETs you get a circuit that draws ~10ma and has poor gain. Crafters soon realized they had to insert a resistor in the bottom JFET source to bias it, then bypass it with a capacitor. With that (obvious) mod the circuit works fairly well and was used by several boost and distortion effects. But there were still a few issues. If the ratio of the resistors biasing the top JFET gate were equal then the clipping was asymmetrical (fixable by making the bottom resistor smaller), it depended on the gate threshold of the top JFET which varies widely (my more normal JFET circuits mostly compensate for this.. there's no compensation here), the clip symmetry varies with battery voltage (which happens with normal JFET circuits to but would like to fix that), and the top and bottom sides of the output waveform varied a bit too much for my liking, both in shape and effective output impedance. Normally varying output impedance isn't much of an issue (pretty much all single-ended preamp circuits do that) but becomes an issue because the output impedance of a mu-amp is very high to begin with.

So I tweaked the circuit under simulation.. made the source circuits of the top and bottom JFETs the same (including the bypass networks), that pretty much fixed the issues I had with the circuit - made it more balanced (including top/bottom output impedance) and insensitive to JFET variation so long as the two stacked JFETs were reasonably matched with one another, and insensitive to the supply voltage. Then I noticed something cool... normally the "bootstrap" cap - the 10uF ceramic connected to the top JFET gate - would be connected to the drain of the bottom JFET. When it is the circuit is mostly perfect.. maybe a bit too perfect. When the bootstrap cap is connected to the top JFET source instead it does an interesting thing - it breaths! When overloaded the initial waveform is clean and symetrical but as the overload continues the clipping symmetry shifts over the course of several hundred milliseconds. The 47uF values determine the time it takes for the shift to occur, and the 10uF bootstrap determines how quickly the top JFET keeps up with what the bottom JFET is doing, so the value determines how pronounced the effect is. With 10uF or more it's quite noticeable, with 1uF it only happens a little, and with 0.1uF there is no shift at all. Does other interesting things when the values of the 47uF capacitors are different. This could be put on a switch but better hide it because if flipped it'll make a loud pop.

The mu-amp stage must be followed by an emitter follower or anything connected to it will load it down. This is a convenient place to put a 12db-octave low-pass filter to attenuate the buzzy clipping harmonics, now made easier thanks to the mods. This is the classic "Sallen-key" active filter circuit, which is usually made with an opamp in follower mode but works the same with an emitter follower. The filter is a bit underdamped to give a bit of a presence boost before taking the dive, for smoother filtering change the 100pF to 220pF. The rest of the circuit is a simple low/high tone control network, the volume and a buffer to drive the output.

Other than playing with the mu-amp stage and tones under simulation, I have not tried this circuit, no idea how it will sound. It will likely require tweaks to the component values around the gain and focus controls and tone controls. It's something I scribbled down a few years ago then lost interest in it after reading stuff about some company laying claim to the circuit (which I can't find any reference to anymore, they must have figured it out). It's not an original idea, if you look around the net there are several others doing similar things, this is my interpretation of it.


A LM386-based Mini-Amplifier with a JFET Preamp

The LM386 is a popular chip for mini-amplifiers, for good reason - it's cheap, available, easy-to-use and puts out over half a watt into 8 ohms when running from a 9V battery. Here's a LM386 project from ElectroSmash that uses a JFET preamp, numerous other LM386-based amplifiers with JFET preamps are on the net. Here's something I threw together recently...


I drew that from memory after the fact (it's at the shop), should be close. It started as a hobby kit that included basically just the bottom half of the circuit with an 1/8" input, volume control and a (rather tiny) speaker. I added the preamp section with gain and tone controls and external speaker and power jacks. I made the amp circuit according to the schematic that came with the kit, it didn't have the 10 ohm resistor and 0.047uF capacitor on the output and the plus input (pin 3) was grounded. Worked like that but probably should add the extra parts - the R/C on the output helps avoid oscillation and the extra 0.1uF cap on the unused input should improve output offset.. the original circuit has an extra ~12.5mV*20=250mV offset from the input resistance, potentially taking away from the output power. Did notice it clipped a bit extra on the low side, will try it with and without the extra 0.1uF cap and use whichever gives the best headroom. Or could omit the caps.. ground one input and wire the other directly to the volume wiper. I noticed no oscillation without the output "Zobel" network but it's on just about every LM386 app schematic and doesn't hurt. The amp was hooked up inverting because that's what the schematic in the kit said and the inverting configuration is more stable (especially with a high gain non-inverting preamp), for a non-inverting power amp reverse pins 2 and 3 but make sure the speaker wires are well away from the input jack wires.

The 1N4001 diode on the power jack protects against reverse polarity by shorting the supply (ouch!) if backwards (but didn't want to waste a diode drop just for that). The LED should be a high-brightness type so a 10K series resistor can be used to minimize current drain. The LM386 draws about 4.3ma at idle, about 0.7ma for the LED and the preamp section draws about 0.5ma, so about 5.5ma idle current - not much but still don't want to forget to turn it off if running from a battery. Current consumption when delivering power is much higher.. roughly idle plus square_root(output_power/load_impedance)/2 [edit, forgot the /2 part, single-ended] plus losses, for 0.6W into 8 ohms that's about 140ma average. Power into 16 ohms is somewhat less, about 0.4W, but that brings the peak current down to around 80ma, probably better if powering from a 9V rectangular battery.

The 2-stage preamp isn't as crunchy as my 3-stage designs but still distorts quite a bit when the gain is cranked. The stages are mostly self-biasing to avoid drastic unit-unit differences but there is still some, vary the 2.2meg high-side gate resistors to adjust the operating point if needed. 47K resistors are used for both the source and drain resistors, so for even clipping at 9V there should be a ~3V drop across the source resistor. The J113 has a gate threshold of -0.5 to -3V, 2.2meg and 470K puts about 1.58V on the gate, for a typical -1.3V to -1.5V threshold that puts close to 3V on the source, thus about 6V on the drain but it's not guaranteed over the entire Vgst range. If building get extra JFETs or be prepared to trim the 2.2meg resistors (reduce to say 1.8M if much less than 3V across the 47K's, increase to say 2.7M if much more than 3V) but this scheme works out better than always having to trim. These calculations are for a 9V supply, the 2.2K resistor drops it a bit but at these low currents not much - at ~0.06ma per fet that's a drop of about 0.26V, not enough to matter much. Compared to other designers I tend to run the JFET's "starved" and close to the threshold, this basically takes Idss out of the equation. The ElectroSmash article details some alternate approaches with names like "The Tillman Amplifier" and "The Fetzer Valve", those "circuits" (different combinations of source and drain resistors with 0V on the gate) are for the J201 JFET which has a much lower typical gate threshold but these days is only available in a SMT package, the biased gate approach works with the more common J113 or if you can find them MPF102 or 2N5457. Note that on the TO92 J113 the gate pin is on the right side looking at the flat spot. Source and drain can be interchanged which I usually do for layout purposes.

The 2.2K resistor in series with the 10uF bypass cap sets the gain of the first stage to roughly 20, if optimally biased for even clipping (about 2Vrms max output) that puts the input overload point at about 100mVrms for a reasonable chance of being able to set for clean. Increase the 2.2K value for less gain and more overload resistance. The 220pF capacitor across the drain resistor limits the gain at very high frequencies and helps avoid "spatter" when picking hard with single-coil pickups. The gain control is compensated so that at low settings it boosts the highs to make up for the high-cut of the tone control and clip filter, and at high settings cuts the bass to keep from mudding out. Mess around with the values of the 470pF and two 220K resistors to adjust the compensation curve, wired these parts off board right across the control for easy changes. The second stage is run flat-out (no resistor in series with the source bypass cap) to maximize gain, not the best for low distortion but it's a half watt baby amp, low distortion is not really a design goal. The drain output of the 2nd stage passes through a 47K/470pF filter to the output emitter follower, which can be just about any small-signal NPN transistor - 2N3904, 2N2222 etc (used the first thing I found don't recall what it was). The tone control is also connected to this point and for the prototype was simply a 0.01uF capacitor in series with a pot, works more like a shelving low/high control than a treble control. Probably can be improved, see the simulations for a circuit that doesn't cut quite as much highs when all the way down. I think I used a 22K emitter resistor and output resistor, the amp only needs about 0.2V for full output so the output resistor drops the signal so that below about 60% it's all JFET distortion and can clean up over 60%. I wired the 1/8" jack input after the volume so that it would also function as a preamp output if needed and also to better mix the guitar signal with a headphone out, vary the 33K resistors as needed.

I made a LTspice simulation of the preamp to play around with...


The stages are about the same but used a somewhat different tone control network.. added a 15K and 2200pF in parallel in series with the existing 0.01uF so that when all the way down it doesn't remove as much of the high-end while still providing a useful low-end boost, and added another 2200pF on the JFET drain for additional options.. can be an extra high control, or if crossed to the other control, a high boost. Note that the effect of the drain cap is asymmetric - because the impedance is high when the wave point is high, as a cut it cuts more of the top side of the wave, and as a boost it boosts more of the bottom side of the wave. Haven't tried this circuit but thinking that the drain high cut and boost resistors (R18 and R24 on the simulation schematic) can be fixed or trimmers to give the main tone (R19) a greater range of control, but still get out of the way (leaving mostly a square wave) when the tone is maxed.

Optimizing the LM386

The mini-amp worked ok but the clipping wasn't symmetrical. Played around with a LM386 simulation I found and it suggested I needed to bias the -in pin slightly positive.. when I got back to the shop and did the mod it made it worse. Took a few measurements and soon discovered why.. with the original circuit (+in grounded, -in cap coupled) the output pin was sitting at 3.79V rather than the expected 4.4V or so (supply was about 8.8V), biasing -in just drove it lower. Adding the capacitor to the +in pin didn't help much, just brought it up to 3.85V (was expecting a larger effect). Biasing was the answer, but had to bias the +in pin instead to bring the voltage up to about mid-supply. Here are the research notes and the present mini-amp circuit...


With biasing got it from 405mW to 605mW before clipping into 8 ohms. No idea if my particular LM386 represents the average, came in a imported kit so wouldn't surprise me if it was a work-alike clone, but for now going with it. When I get around to PCB layout I'll add biasing resistor locations for both the +in and -in pins so no matter what the particular chip characteristics proper biasing can be achieved.

I made some changes to the simulated LM386 to better match what I was seeing...


The first two plots show the output with a 16 ohm load, first with actual (pulsed) current draw, the second with the filter cap very large to show the averaged current draw.. about 80ma when delivering about 400mW. The next two plots show 8 ohm output and current draw.. about 150ma when delivering a bit more than 600mW. Here's the cba386_1c.asc LTspice simulation file.


MOSFET circuits

I haven't done a lot with MOSFETs.. they always made me nervous, some are so sensitive just looking at them might blow them, or worse damage the gate but still work then blow later. Fortunately these days one can get gate-protected MOSFETs that can be handled without precautions. I've made a few power amps with "HEXFET" outputs, this is one I made into an old Champ chassis with a two tube preamp...


...not to be confused with a Fender Super-Champ, that's just what I called it because it was louder than any other Champ I've ever heard (and I'm pretty sure at the time I hadn't run into Fender's original Super-Champ with 2 6V6's or I would have labeled it something else). At least until I let it be used as a monitor amp at a gig and they connected too many speakers.. smoke came out and it ended up getting scrapped for parts. If only I had added a thermal cutout on it.. power MOSFETs are pretty tough, usually (with simple current limiting like the zeners I used, not marked but 6.2V or 7.5V if I recall) they can take a short long enough for the fuses to blow but they can't survive sustained thermal overload, if only I had used a cutout.

My favorite MOSFET power amp was one I made back in '81 or so.. it also used IRF132/IRF9132 but used a single JFET as the driver - a now unobtainium TIS-58, those things could take a 100V or more. Between the output gates I used a TL431 shunt regulator to set the bias with a single resistor to the + supply on the other side. Single-ended power supply, speaker was capacitor-coupled, a trimpot on the JFET source set the bias to put the output at half supply. Perhaps the simplest power amp I've ever made. Despite no negative feedback it sounded great. Replicating that design with modern parts would require using a higher voltage MOSFET for the driver (JFETs top out around 30V these days), but I doubt it would sound as good, MOSFETs usually don't have as nice of a transfer curve compared to a JFET. Just haven't been that interested in MOSFETs until recently when I realized that the CD4007 wasn't just another CMOS logic chip...

CD4007-based Overdrive

The CD4007 is more than just 3 inverters in a package, although the N-channel and P-channel FET pairs have common gate connections, only one pair has the drains connected together and all of the individual sources are brought out to separate pins. Here's a detailed schematic showing the substrate connections (slightly doctored to correct connection errors)...

The first pair's sources also connect to the substrates of the other FETs, so pin 14 should be connected to the highest voltage and pin 7 should be connected to the lowest voltage. Preferably stable because substrate voltage variations affect the characteristics of the other FETs. Otherwise the transistors can be used individually. Thanks to a fellow pedal hacker I was corresponding with over pedal mods, I learned about this neat chip.. noticed that on a DOD FX90 schematic one of the FETs was being used in isolation (most schematics show only logic symbols).. hmmm gears start turning, had to look into it and sure enough it was really a mosfet array with semi-isolated transistors. Of course the first thing I start thinking is overdrive pedals. I simulated a few things that at first had me amazed only to find out my models were inaccurate and when I plopped in more accurate models the circuits no longer worked.. but with tweaks got it working again. Finally I found Lynn Fuller's more accurate (I hope!) level 7 CD4007 models and wired up the simulated circuit to at least partially simulate the parasitic components.

This is one of the circuits I've been playing with in LTspice...


The basic topology is similar to the "Dark Blue" overdrive idea presented earlier, a preamp stage driving a "mu amp" formed by stacking a FET on top of another to serve as a current source, increasing the gain. In this case not as much as the JFET version due to negative feedback but it still has a maximum gain of about 300 according to the simulation. Disconnecting R13 increases it to about 500 but then the gain controls won't work right. For this application I wasn't after lots of gain but was trying to try to linearize the output stage, I like rounded clipping but the usual mosfet inverter roundness is a bit too much. I'm still tweaking but in this circuit but with the present values under simulation the mu amp stage actually has less gain than the inverter preamp, but the clipping is more like... clipping. Here's the 4007mudrive1.asc LTspice file.

The following simulations are from a similar earlier version of the simulation, ignore the "Cascode" in the comments, that was a mistake - a cascode circuit is a common source amp connected in series with a common gate amp.. a mu amp is (more or less) a common source amp with a constant-current load. These simulations were done before realizing that the NRS and NRD properties were interfering with the models (the effect was minor for this circuit but makes much more difference for higher current circuits - more on that in a bit), and before playing around with the tone network.

It starts off rounded then flattens...



The top flattens more than the bottom because of the substrate diodes shorting out the bootstrap, but it's still much better than the clipped output of the more traditional inverter preamp, the gray trace in the last image. The drive controls go below unity so that sound can still be accessed. One of the things I find interesting about this circuit is (like the JFET version) it has dynamics...



As the signal is applied the operating point shifts downward (shown by the brown band showing LED current which is in series with the output buffer source resistor.. the status LED flickers while playing:-) causing the clipping to shift towards the bottom side. Capacitor C2 sets the time constant - these are with 1uF, 0.1uF causes the shift to happen rapidly, 10uF causes the shift to happen more gradually, this can be a 3-way "feel" switch. There looks like there would be recovery effects between notes.. after the signal stops abruptly there's a pretty drastic downward shift as the bias is sucked out (C2 sets how fast it recovers). Simulating using a switched signal showed that while there is an effect it doesn't fully cut out the signal and it fully recovers within a few cycles. The duty cycle shift has a limit, under heavy overdrive it squares and stays square, the dynamic effect is mainly at the transition between clean and clipped.

This thing should be pretty nasty.. anxious to throw together something to see how it really sounds. The tone network is just an initial guess and something for the sim, the final circuit will probably be different.. in an AC response simulation it wasn't that great. I expect some values will have to be adjusted to work with a real CD4007 chip. There's nothing really unique about using the CD4007 as a preamp, once I looked for it I found several examples on the web, mostly overdrive pedals. There are also several mu amps using mosfets on the net but didn't see any using a CD4007, most use 2N7000 or BS170.

Still working on accurate models.. the Lynn Fuller models from the "Introduction to Modeling MOSFETS in SPICE" document seem pretty good, but there's a typo in the properties line - omits the NRD=0.54 property (also it's supposed to be NRS=0.54 not O.54, an OCR error). These determine the source and drain resistance from other factors, defaults are 1. In the above sims I omitted the NRD property in all the mosfets, but adding them back made almost no difference and omitting both NRD and NRS only causes a relatively minor bias shift. However, when I tried to simulate the traditional self-biasing inverter preamp circuit the results weren't so great.. worked but off center and distorted. In this case omitting NRD and NRS permitted the circuit to work mostly as expected, so now wondering how accurate the model is for simulating analog circuits but being a resistance thing it's probably something that only comes into play with higher current circuits. On the other hand, the N-channel and P-channel mosfets in a CD4007 have different characteristics so I wouldn't expect them to self-bias perfectly.

I simulated a circuit similar to the most common preamp application...



The top two plots are with the NRS and NRD properties omitted, looks about right, self-biases close to half supply. The second two plots are with NRS=1 and NRD=1 for all the mosfets which should be identical but quite clearly the default is not 1! Or more likely the default really is 1 as documented but there's a bug in how LTspice calculates those properties. The plots for the recommended Pchan NRS,NRD=0.54 and Nchan NRS,NRD=0.1 are almost equally bad. It gets worse when using with the simplest circuit with the sources grounded and at supply, with NRS NRD omitted versus the recommended values...


Yuck! I don't know for sure until I breadboard the circuit, but I'm leaning heavily that the first plot is a lot closer to what it's supposed to look like, at least that's what I remember from working on Sunn Beta amps which used CMOS inverters as preamps. Using NRS NRD values of 1 and 0 made little difference. Until I get to the bottom of it I'll be omitting those properties. I'm running LTspice under wine which is known to sometimes induce bugs but usually only with the GUI due to differences between the real and fake libraries, never seen it bug out on math. Here's the test1.asc LTspice file.

The following simulations test the N-channel and P-channel mosfets in a very low current common-source amplifier configuration...



...if actual performance is anything like that then that's great! The P-channel MOSFET has a bit more edge on the drain side when hard-clipped but these kinds of circuits are usually done using the N-channel MOSFETs. The N-channel MOSFET does it too but not quite as much. Note that the other side of the source bypass cap for the P-channel MOSFET is grounded to improve supply rejection but otherwise the two stage configurations are mirror images of one another. With the values shown the effective input/node impedance is about 85K, thus stage gain is about 2.2M/(85K+InputR), or about 17 for the first stage and about 7 for the second stage (not counting the effect of the output impedance of the first stage).

The circuit works without the source resistors and their bypass capacitors but then the 2.2meg resistors from gate to source have to be reduced and tweaked, adding the resistors/capacitors provides a convenient way to set the stage bias without messing with the feedback resistors. These plots are with NRS/NRD omitted but with only 41uA current those properties make little difference, just a slight gain reduction. Because of the low operating current these circuits should be a good test of spice model accuracy. Here's the test8.asc LTspice file.

A more practical CD4007 overdrive circuit

I like the strange dynamics aspect of the "mu amp" circuit, but it's a lot of parts and more to go wrong if the actual parts don't match the models, and unlike other mu-amp variations its "mu-ness" didn't increase the stage gain all that much (plus was knocking it back down with negative feedback anyway). Need something simpler to plan for an actual build. This circuit uses the complimentary preamp section and the drive and tone controls of the previous mu-amp design, but the mu-amp itself is replaced by a simple conventional gain stage. I have seen CD4007-based variations of both types of circuits in the wild so this one has a fairly good chance of working...


This schematic shows the input/output/DC jacks, battery and footswitch needed to make a real pedal. The 2.2K resistors set the stage biases, these values produce fairly even clipping under simulation, with a 9V supply (9.6V at DC jack due to the diode) the bias 1 resistor sets pin 12 to about 4.14V, and the bias 2 resistor sets the voltage at pin 8 to about 4.22V - decrease the 2.2K value to lower the voltage, increase to raise. If not in the range with ~1K-5K then the 2.2M resistor on the bottom might need tweaking.. increase to lower the voltage on pin 8, decrease to raise. The drive and tone control caps are basically guesses, once made will tweak by ear if the controls don't do what I want.

When using the CD4007 dealing with the substrate connections can be tricky - note that the unused pins 1-5 are returned to pin 7 (negative substrate) to avoid possible substrate conduction effects - all pins should at or greater than Vss and in this circuit Vss sits at about 0.3V. Unused pin 13 is connected to pin 14 since the drain of that FET is connected to 14, grounding would cause unwanted current flow. The source of the first stage rests at about 1.5V and doesn't dip under 0.9V even with high input level (at least under simulation) so that shouldn't be an issue with this circuit, but in general the FETs should be chosen so that pin 7 is the lowest voltage and pin 14 is the highest voltage. Also keep in mind that any signal present on pin 7 or pin 14 gets fed back to the substrates of the other N and P channel FETs, changing their characteristics and potentially causing a feedback loop.

This isn't a particularly high gain overdrive, with the drives cranked under simulation it starts getting into the dirt with an input of around 6.5mV or so. For more saturation clipping diodes can be connected at the input of the tone network, the schematic shows a couple ideas using a DPDT center off toggle switch to select symmetric, none, or asymmetric clipping. Asymmetric clipping produces DC offset so the circuit includes a capacitor in series with the diode along with a normalizing resistor to minimize the effect. The value of the capacitor affects the dynamics and low frequency clipping response, in the circuit shown connected the capacitor is connected all the time for both symmetric and asymmetric selections, the alt circuit on the bottom connects the capacitor only for asymmetric clipping. Note that when using the diodes the output level is drastically reduced, to avoid surprises the switch should be positioned to minimize the chance of accidentally flipping.

With no diodes connected and the drive controls at minimum (max resistance) the output stage doesn't clip and the input stage starts clipping around 35mV with the classic CMOS inverter roundness. The drive controls are 3meg reverse-audio taper, what was for the old Fender tremolo speed control and readily available from amp parts suppliers. The idea behind two drive controls is the low drive has a lower high-pass frequency, so it can be set lower and the high drive cranked to bypass it more at mid and high frequencies for more bite and to make up for the treble cut after the clipping stage. Will see how that works out.. could use a single "compensated" gain but the dual drives should provide more control over the pre-clip EQ.

Here are some simulations of the circuit without the saturation diodes...


The following simulations show the effect of the extra diodes in asymmetric and symmetric configurations...



In this circuit the difference between symmetrical and asymmetrical diodes is subtle due to the DC shift effect and the way the diodes round the wave, happens mostly at the onset of clipping and how long the offset effect lasts depends on the value of the coupling capacitor to the diodes. Also, the diodes put an additional load on the CMOS drive stage so additional drive causes an asymmetric shift, even more so with the symmetric diode selection. Interesting circuit. Here's the 4007drive2b.asc LTspice file.

It should be possible to made a CD4007-based 3-stage "grunge zone" overdrive (there's another stage left in the chip), but already hitting more than 50db gain across a tiny chip with high impedances, stability is already a minor concern so need to try this lower gain circuit before jacking the gain by another 20+db. Also have no idea how the noise performance will be until I make it. If more gain is needed it would probably be more practical to add a separate JFET preamp and gain control before the circuit, the J113 is fairly quiet and has reasonable overload characteristics, typically >150mV.


Using these circuits

Everything on my web pages that isn't public domain and doesn't belong to someone else is Copyright William Terry Newton. That's the automatic default for US copyright whether marked or not. Permission is granted to use this material for non-commercial purposes, make for yourself, share on forums, tinker with, etc. Have fun. I would prefer that if the material is used substantially as-is for commercial purposes that you get with me (I can help) but to my knowledge nothing here is patented, and I don't care if my ideas inspire other ideas. That's what electronic design is about, especially with musical electronics. A lot of the ideas in my circuits were inspired by high-gain tube amps from the '70's to '90's, particularly the designs with the tone controls after the clipping stage - that's how grunge came about. Just about everyone does it that way now.

All material is presented as-is and without warranty. If you find a mistake please let me know. Warning! Do not attempt to build any circuits that use high-voltage tube circuitry or mains power unless you have experience with such circuitry and know how to do it safely.

The circuits are generally in schematic form only with no construction details, and often omit things like the bypass switch, power supply and other details. If using these circuits it is assumed that you know how to take a schematic and "fill in the blanks" to produce something usable, because other than maybe a few specific things I don't want to make this a hobby construction site. If new to electronics then I suggest that you tackle simpler projects first and get those going before trying more complicated circuits, there are many such projects on the net complete with layouts and construction hints.

Once you get the hang of taking it from a bare circuit diagram to something that actually works then it gets a lot easier.. the more complicated stuff is the same just more of it. A couple of the diagrams show the bypass switch wiring, that's just one way to do it. Don't be afraid to play around with component values and different circuit arrangements, when running from 9V there's not much that can go wrong other than making it sound worse or make no sound at all, that's how learning happens and sometimes you might discover something magical.

For a deeper dive into the circuits I recommend using LTspice or another spice simulator, once you can correlate a waveform to what it probably sounds like using a simulator makes it much easier to play around with component values and circuit variations, no solder needed and you can better see side effects like excessive current drain, mis-biased stages etc. LTspice circuit files are provided for some of the circuits here. While working under LTspice is no guarantee at all that the circuit will sound good, but if it doesn't work under LTspice chances are good it won't work in real life.

I can't promise technical support but feel free to email me with questions comments or improvements, I love to talk shop.

Stay Well People!!! And keep up the good tone.


Last modified January 31, 2021
Terry Newton (wtn90125@yahoo.com)