Monday, 1 May 2017

Hornby Skewes Zonk Machine I

Well got time for another study break, so it's that time for another layout. Just like the last layout I posted, here's another uber rare fuzz, a gift from the fuzz gods that no worshiper of fuzz should be without IMHO.

One of the more famous Tone Bender MKI influenced fuzz boxes and now almost as legendary as its golden cousin was the John Hornby Skewes Zonk Machine. A little more wacky in the graphics department, but not as heavy duty or robust as the MKI, the Zonk Machine was almost identical in its circuitry. Made between 1965 and 1966 in Leeds, England, the Zonk Machine like the MKI was only made in small numbers before the pedal was updated with a more modern and easier to produce unit.

Layout 1: Original PNP

Layout 2: Original NPN

Some suggested gains and tips I found over at DIYSB:

One suggestion with the leakage:

Values taken from an Original Unit
Q1=TI AO2 650 Hfe= 180, leakage= 0.12mA.
Q2=Mullard OC75 Hfe= 78, leakage= 0.29mA
Q3=OC44 Hfe= 58, leakage= 0

Tips for other builders interested in the Zonk Machine:

1. Socket all transistors because you’re likely to go through a bunch of them before you find some that work well.

2. The Zonk is very bright. Play around with the input cap and try values between .001uF and .01uF. Maybe even set up a three way switch to toggle between different values.

3. Definitely use a log/audio taper “Fuzz” pot. I used linear at first but it greatly limits the useful range of the pot. The name “Fuzz” is a bit of a misnomer as it doesn’t really fulfill most people’s definition of a fuzz control. It essentially adjusts the bias of Q2. The Sola Sound Tonebender Mk1, which is a topographically identical circuit (sans the 33K resistor), called this pot “Attack” which is a bit more appropriate IMO.

4. Set up an SPST switch to bypass the 2.2M resistor for nasty, all-out craziness!

5. Some came with 25k Log Fuzz pot, and 220k resistor instead of 470k.


  1. layout has no output labelled

  2. Couple errors on the layout the 1m restore goes to the other side of the 1n at the base of q1, the 25uf connects to the emitter not the collector of q1 and the 2m2 connects to thw Junction of 47nf and 56k not the ground and swell 3 goes to output

  3. Oop. Put up the wrong layout. I had 2 going, trying to make it how I wanted it. I'll put up the right one when I get home. Nice catches.

  4. Verified PNP version. Took a while to get it tuned. As usual with MK1-esque circuits it took a fair few trannie swaps before it started to sing. Settled (for now) on Q1: OC75 (~80 hfe) / Q2: NKT275 'reissue' (~100 hfe) / Q3: OC75 (~80 hfe). For the sizzling bacon noise that seems to plague these builds try a cap on the input > ground. 1nf removed the artefacts without killing the highs. With this pulldown cap in place transistor swapping can be more forgiving. Plus, the original 1nf input cap resulted in a rather tinny sound so I bumped it up to 10nf and now its beefier with a stronger midrange. Nothing out of the ordinary but it may help out first time builders of this circuit. Thanks Zach.

  5. Looks like Layout 1 = Layout 2 to me. I don't have time to build *anything* at the mo (grr) but this one's gone on my list.

    1. In Layout 2 the transistors have not been updated to NPN type but the layout is correct in regards to polarity (negative ground).

    2. Did not catch that the transistors are mislabeled in the notes. I'll fix it later. Thanks Phil.