This Week on the Breadboard: BJFe Arctic White Fuzz

Chuck D. Bones

Circuit Wizard
I've been meaning to build this one for a while. A fairly simple FF-based fuzz with the usual Bjorn Juhl magic. He uses a JFET for Q2 which requires some cleverness because with JFETs, the source voltage is higher than the gate. A big part of getting the biasing right is choosing the right JFET. I tried quite a few, including the J113 that BJ designed-in.

This schematic was lifted from FSB. At first I thought the trace had an error because the TONE control does practically nothing. When I listened to a demo, the TONE control is quite subtle. Still, I think the connection from C5 to pin 2 of the TONE pot is a trace error. More on that later.

Q1 is a high-gain, low-noise Silicon BJT. In a regular Fuzz Face, R10 would connect directly to Q1-B. But because Q2's source voltage is higher than Q2's gate, that won't work. BJ added R3 and R4 to reduce the DC voltage going to Q1-B. I think he chose 91K and 9.1K resistors because they have brown & white color bands, colors you'd find in the arctic tundra. R5 helps stabilize the bias. C2 & R6 provide some AC bypass to increase Q1's gain. C4 & R9 perform a similar function for Q2. I've seen BJ do this in other circuits, such as the Folk Fuzz and the Pink Purple Fuzz. The FUZZ control is not as simple as it looks. It mimics the effect of the guitar's Volume control. It interacts with Q1's complex input impedance. At low FUZZ settings, The balance between bass & treble is pretty even. As FUZZ is increased the tone shifts away from bass and becomes very mid-focused. Max gain is about 70dB at Q2-D, higher than your average FF.

So how does it sound? Pretty much as Bearfoot describes it in the adverts: "...purrs like a snow leopard at lower levels and attacks with full vintage growl when pushed." Noise is practically non-existent. Even with FUZZ at zero, you have to back off on the guitar Volume to play completely clean. I've only run it guitar -> AWF -> delay -> amp so far. Probably the smoothest FF I've played. It should stack well with other pedals and doesn't have to be first in the chain like most FFs. If you have a breadboard and the time, I highly recommend building this.

The TONE network is a bit odd. As drawn, it does practically nothing. So I made a little mod. It Is The Way.

BJFe Arctic White Fuzz.png

The TONE Mod
I wanted more range out of the TONE control, so I made a minor mod. I removed the wire from C5 to C6 and the TONE pot. Big difference. I ended up lowering C5 to 15nF and R12 to 3.9K (had to get that white band resistor in there). I also increased the VOL control to A100K. The modded TONE control does what the original does, but it also goes darker and brighter. C6 and R13 load down Q2's output, which not only affects the freq response but also the harmonic content at Q2-D. Breaking the connection between C5 & C6 makes the loading variable; it depends on the TONE setting. This variable loading alters the harmonic balance as TONE is rotated. It's subtle, but it's there.

Picking a JFET
We want something with Vp around 1 to 1.5V. J201 won't work, Vp is too low. MPF4393 and J112 also won't work because Vp is too high. I got good performance from J113, 2N5246 & PF5102. J113 produced the most gain. 2SK193 is another high-gain JFET if you can find one. Every JFET needs to have its bias dialed-in. The best way to do that in this circuit is to tweak R8. I ended up with a 5.90K resistor. You could put a 3.9K and 5K trimmer is series in place of R8.

My breadboard circuit looks like this. Voltages are measured with no input signal. Because R14 limits the power supply current, a shunt reverse protection diode (D1) is viable, so I did it that way. Not better, just different.


Alternate BJTs & JFETs are in the upper left corner of the board. I tried BC549C, 2N5210 & 2N5089 for Q1. They all worked equally well and each one required a different R8 to get the biasing where I wanted it. I'd say the higher the HFE, the better in this circuit. If I'd had a BC550C, I'd have tried that too. Naturally, I had to use a white LED.

Average White Fuzz breadboard 01.jpg
Last edited:
Thanks, Man. Next time that BB is apart, I'll photo the pot adapter boards. I picked up some of those 3-terminal blocks from EG when they had 'em. Stuck 'em on some leftover Vero board. The bottom side of the Vero has a double-row header. Plugs into the breadboard securely.
Your posts have pushed me into the breadboard game. I’ll have to line this up as the next one, I really enjoy your schematics.

Any issue if I used a 1N4001 in D1? The only difference I can see is Peak Inverse Voltage.
Will an MPF102 work? Are they 100% the same as the PF5102 or just a reasonable substitute?

I have a ton of them, and in most circuits, they are too low-gain to be used as a common-source amplifier.
I kinda doubt MPF102 will work unless you find one with Vgs,off (Vp) in the recommended range.
Looking at the MPF102 datasheet, Vgs,off ranges up to -8V. gfs (gain) ranges from 2ms to 7.5ms, but the ones with low Vp will be at the bottom end of the gain range. MPF102 is an RF transistor. Its performance is optimized for high-freq operation. Not a good choice for this circuit because of the low gain and wide Vp spread.
PF5102 is nothing like MPF102. Vgs,off ranges from -0.7V to -1.6V. Gfs is very high, 11ms or more.
J113 is similar to PF5102. Vgs,off is between -0.5V and -3.0V. Gfs is quite high, around 15ms for the Vp range we want.
I don't see any similarities between MPF102 & PF5102.
Chuck has an impressive mind.
Not really, I just notice things. BJ picks some strange resistor values. Lot of his circuits have 14.7K when 15K will work just as well. I have some favorite resistor values based on colors, so I thought he might be doing the same thing.
  • Like
Reactions: fig
Great analysis and really cool circuit.

The Fuzz pot is even more of an oddity when you imagine it connected to your guitar. With the Fuzz pot at max and your Guitar Volume at max, your pickups are going to get the most loaded by Q1 and the Negative Feedback path. Then as you turn the Fuzz pot down, the load is removed from your pickups, but you get all sorts of complex attenuation from the Fuzz pot, R1, Q1 and the Feedback network. Right?
Pretty much, however there is very little feedback back thru R10 because there is very little signal at Q2-S. R9 & C4 shunt almost all of the signal to ground, except at bass frequencies.

Q1 presents a non-linear load to the input circuit and guitar. The higher FUZZ settings expose the guitar or whatever is driving the AWF to that non-linearity. A certain amount of distortion happens right at the input. There is a lot going on for such a simple circuit.
Last edited:
  • Like
Reactions: fig
I built a version of this a few years ago and also felt like the tone control needed more range. Should probably give this another try with your mods. Thanks for the build report!!
This is a case of the marketing making statements we have no way of verifying, or knowing if the analogy is even a good thing. "...purrs like a snow leopard..." Have you heard a snow leopard purr? I haven't. I doubt many people have, including the marketing guru. For all we know a snow leopard's purr could sound like a bit crusher.

This kind of reckless ad copy has to stop.
This is a case of the marketing making statements we have no way of verifying, or knowing if the analogy is even a good thing. "...purrs like a snow leopard..." Have you heard a snow leopard purr? I haven't. I doubt many people have, including the marketing guru. For all we know a snow leopard's purr could sound like a bit crusher.

This kind of reckless ad copy has to stop.
No, but I did once hold and then play with a wee lion cub. It said, “Raaa!”. Thinking back, it was a bit bit crusher-ish sounding. 🤔
Oh you may well laugh but how many lives have been ruined by copy just like this?

I don't know either.