A-OK Overdrive - Another One-Knob Fuzz / Overdrive

[samhay]

I was looking for a quick diversion recently and came up with this:
There are a few circuits I have always meant to box up. Among them are a Fetzer valve and a Bazz Fuss. This could be their bastard child - a one-knob fuzz / overdrive that blends the two circuits.
To keep the signals in-phase, the fuzz (Q2) is taken from the drain of the FET (Q1) via the 22u cap (hence the 'original'  Fetzer design rather than the newer version without a bypass cap). Feeding a diode-biased BJT from a low-impedance signal like this works nicely and there is no problem with gating of the signal. Q2 biases up fine with a 10k collector resistor, and this puts enough current through D1 to allow it to work as a signal indicator.
This is a fairly tame version of the Bazz Fuss, but like this these two circuits work nicely together. The output of Q2 is similar to that of Q1, so there is not a lot of volume change across the range of the pot - this differs from most one-knob designs, which usually control either the bias or the volume - and is something I have been trying to achieve for quite some time. I dropped a few transistors into this circuit and it worked fairly predictably. However, Q1 and/or Q2 may need to be auditioned in order to find transistors that give similar output levels.
Q3 buffers the mix pot and keeps the output impedance down, so it plays nice with others. I added a switch so you can use this either at unity gain or boosted. Alternatively, a second volume pot could be added...

I made a quick recording last night, recorded direct with no processing. Excuse the playing.
https://dl.dropboxusercontent.com/u/11996927/A-OK_Overdrive.mp3

Layout:


I knocked up a quick vero layout, which is designed for vertical mounting to the fuzz pot. It is quite tight, but can easily be stretched vertically for more  room if it is to be mounted more conventionally.

[Kipper4]

Thanks Sam looks sweet.
Can't wait to hear it. So small too should easily fit in a 1590bb with a battery too.
Another for the list me thinks.

[samhay]

The layout is narrow enough to comfortably fit into a 1590BB as long as the knob is centred - that's what I'm doing with it anyway.
It should be easy enough to get it into a 1590A with a new layout.

What does it sound like? At the 'clean' end of the pot, it sounds like a FET with a bit of gain - fairly bright and a little crunchy if overdriven. At the other end of the pot you get a lot of compression and a reasonably smooth sounding fuzz. Mixing these together you get the sustain/fuzz bleading through under the signal, with the fuzz blooming as the note decays.

[deadastronaut]

Cool description sam...

look forward to hearing it....nice one. 

[samhay]

The link to a cheap-and-cheerful clip has now been added to the first post.
Note that this is not boxed up yet, so there is a little bit of noise, which should go once it is tidied up.

[samhay]

Had a bit of a boo boo last night and managed to fry Q1, Q2 and D1 - note to self, take more care with alligator clips when circuits are powered.
Anyway, replaced Q1 and Q3, rebiased Q1 and it sounded and behaved the same, so it looks like the circuit should be fairly resiliant to component choices.
Another thought. This will probably work quite nicely on bass as it has a fairly generous bass response and the mix/blend function usually comes in handy in this context.

[deafbutpicky]

With a Rs like this it should work with quite a variety of jfets (due to its degenerative effect on gain, it seems a bit
high for fetzer though).

Is R2 really needed?

But what I really wanted to say: it sounds good, nice demo

[samhay]

Thanks deafbutpicky

I took the Rs value from Figure 1 of the 'closer look at the Fetzer' page at ROG: http://www.runoffgroove.com/fetzervalve.html
I went for the 33k series resistor and 220p cap as found in the later revision - told you it was a bastard child.
Anyway - yes, with this Rs value it should work with most of the usual suspects including the J201.

R2 is a pull-down / anti-pop resistor, which is a good idea as I added a cap to the input - I get nervous DC coupling input/output stages to the outside work.

[deafbutpicky]

about R2:
I was just thinking out loud, as the IN cap should have very small to no charge up (no voltage at the gate, you know)
and I've build some without cap pull-down with no problems...

Yeah I see it the same way, DC coupling is a no go.

[samhay]

You're quite right, and the way I see it, input/output pull-down resistors can generally be considered optional. I guess I should get in the habit of annotating schematics a bit more to make this sort for thing clear.

On a different note, I decided that the fuzz was a bit dark in some cases, so will make C5 switchable - easy to do if the LED is mounted on the panel. I have put 150p on the board and will switch 1n in parallel for a 'warm' setting.

[duck_arse]

hmmm, I got's me a bag of smd R's and C's today, along with my first perfboards, and was wondering what I could first build on/with. maybe I'll re-invent "perface-mount" with this.

[deafbutpicky]

Don't get me wrong sam, I was just in a picky mood. I'll give it a try when I get my breadboard free (probably should just
get another one anyways).   

[samhay]

hmmm, I got's me a bag of smd R's and C's today, along with my first perfboards, and was wondering what I could first build on/with. maybe I'll re-invent "perface-mount" with this.

Sounds like a plan. Despite my initial intentions, I guess this isn't the smallest fuzz circuit out there, but there aren't that many R's and C's involved. Not sure how much of an assortment you have, but most values are not very critical.

[samhay]

Don't get me wrong sam, I was just in a picky mood. I'll give it a try when I get my breadboard free (probably should just
get another one anyways).   

Not at all - they were fair points and I always welcome suggestions/criticism, etc.
I understand the breadboard real estate issues - I built this on a little corner of my main breadboard, which is still mostly covered in my delay/reverb project (which I really must get boxed up next).

[samhay]

The schematic and layout have been updated to show what I have boxed up.
I used an on-off-on DPDT switch to switch in the 1n cap across Q2 for a 'warm' setting and to add the boost. The boost is significant, and might be more useful if the value of R8 is reduced.

[samhay]

And a gut shot.
There was quite a lot of room left so the second board has another stage of power filtering and a MOSFET bypass indicator with the LEDs mounted on the bottom.
The extra filtering isn't necessary - it is quiet either way, but I figured it wouldn't hurt.

[duck_arse]

stroke of genius using the bazz fuss, sam. I threw this at the bb today, had a cloth-eared, 1 chord test listen.

to me, with the jfet biased right down low, roughly 3V for a j201, 2V5 for a 2n5457, it sounded "more". probably not in the spirit of the fetzer, using it to provide an assymetric signal, but what are you going to do? looking at the signal at the fetzer end of the fuzz pot shows some odd little glitches, devils' horns, like crossover distortions or something, don't know if they affect the sound.

I tried 3 j201's, a 2n5457 and an mpf102, so not much of a sample. mpf had less background fizz than the 2n, the j201 probably best sounding, but only when low-biased. bc548b hfe 273 sounded a little less sputter decay/tail than a 2n3904 hfe 156 at Q2, to be expected.

carry on.

[samhay]

I did spend a little time playing with the FET biasing and I think it sounds beter with the drain at about 4.5V vs. 6V. Didn't try much lower, but can pull the bottom off and see.
Only problem is that Q3 is DC coupled to the source, so you will start clipping this if you drop the voltage a whole lot lower.

It is not the typical Fetzer clean with the fuzz pot CCW. The 22u source cap makes it a little prone to clipping and there is a little fizz from the fuzz bleeding through due to the non-negligable output impedance of the FET.

I preferred a higher gain Q2 - 2N5088 sounded slightly better to my ears than a 2N3904.

[duck_arse]

I completely missed the dc coupling aspect. my sig-gen is decommissioned for some mods, I'll have to get at that and look proper.

can I ask about your "source cap"? the 22uF is not meant to be source bypass cap is it? as a dc blocking for Q2, it won't affect the gain of the fet, will it?

[samhay]

The 22u cap is both a DC blocking cap and a bypass cap (hence why I chose a rather large and classic value).
There is no base resistor on Q2 so the input impedance is rather low.