ZONKBOX not zonking

[PAGOON]

I tried all types of transistors with all types of gain and leakage. I’m just wondering what did everybody do when they built this one ..

 

[z2amiller]

Nice soldering, clean board. Do you have an audio probe to see where the music goes to die?

 

[spi]

Like, no sound? or no fuzz?

 

[PAGOON]

Just don’t sound right… I get sound it just is horrible… I noticed if I boost the signal with another fuzz or my rangemaster it sounds better… I just can’t get a good sounding pedal… what’s the gain and leakages needed for this type of pedal … is it normally this picky ?

 

[PAGOON]

No audio probe I never had luck with those. Maybe I will give that another try.. I even went through all the solder points again.

 

[owlexifry]

Just don’t sound right… I get sound it just is horrible… I noticed if I boost the signal with another fuzz or my rangemaster it sounds better… I just can’t get a good sounding pedal… what’s the gain and leakages needed for this type of pedal … is it normally this picky ?

post up some voltage measurements at each transistor pin if you can.
would be interesting and possibly valuable to confirm what the idle bias voltages are. if the idle bias isn't quite right at certain spots, it will not operate correctly.

the zonk is basically a Mk1 tone bender.
they are an absolute c*nt to get right, mostly because all x3 transistors (IME) require a shitload of leakage to bias correctly.

i've found this advice to be fairly on point:

(most of the transistors i've tried max out at about ~250uA leakage. i don't have the experience or range of Ge units to confirm whether a leakage as high as 500uA is that necessary, but i assume it probably helps.
also bear in mind, more leakage = more noise)

 

[PAGOON]

So I just had the best luck with a 2n2614 in QI HFE 188 190ma Q2 and. Q3 I put the highest hfe and highest leakage transistors I could find in there like 500- 600 hfe with 300-400ma and it still needs to be boosted with the rangemaster to sound best… I’m going to try these specs I saw online … with some different transistors
The PNP Zonk Machine (JHS/Hornby Skewes) typically requires a specific combination of low-to-medium gain, high-leakage Germanium transistors to achieve its characteristic gated, "velcro" fuzz sound. The circuit is famously finicky and often relies on leakage to bias the transistors, making transistor selection crucial.
PedalPCB Community Forum +3
General Recommended Transistor Specs (PNP Germanium):

• Q1 (Input): Low gain (
  
  hFE
  60-80), moderate leakage (approx. 200uA).
• Q2 (Fuzz/Gain): Medium gain (
  
  hFE
  80-120), higher leakage allows for better biasing of the fuzz pot.
• Q3 (Output): Low-to-medium gain (
  
  hFE
  50-90), high leakage (up to 800uA+) to bias the stage.

 

[PAGOON]

post up some voltage measurements at each transistor pin if you can.
would be interesting and possibly valuable to confirm what the idle bias voltages are. if the idle bias isn't quite right at certain spots, it will not operate correctly.

the zonk is basically a Mk1 tone bender.
they are an absolute c*nt to get right, mostly because all x3 transistors (IME) require a shitload of leakage to bias correctly.

i've found this advice to be fairly on point:
View attachment 115616
(most of the transistors i've tried max out at about ~250uA leakage. i don't have the experience or range of Ge units to confirm whether a leakage as high as 500uA is that necessary, but i assume it probably helps.
also bear in mind, more leakage = more noise)

WoW what a picky unit… that’s some great info thank you!!!

 

[PAGOON]

I just put some langrex 6lr in there q1 305 180ma q2 235 .315ma and q3 471 594ma and it’s sounds like a pedal now. What a mess I made trying to find the right transistors for this unit. I still I’m going to mix a few of these around and see if it sounds better… the sustain falls out quick after 6-7seconds after I strum. But it sounds better with these what I thought were useless transistors. Thank you for the help…

 

[luizbyrro]

I believe that R6 should be 33k.

I’ve seen some schematics showing 3k3, but that drops the 50k pot down to 3k. With a 33k it ends up closer to 20k.

When I built mine, it barely had any sustain with single coils, and it wasn’t the gated effect.

After some research, I found that several builds use 33k for R6. I also increased the input cap to 2.2nF, which worked better with single coil pickups.

Now it sounds much better!

After finishing the enclosure I realized I had swapped fuzz and swell...

q1 nec 2sa406 (hfe 81 / 0.114)
q2 toshiba 2sb54 (hfe 121 / 0.350ma)
q3 hitachi 2sa15 (hfe 51 / 0.044ma)

 

[owlexifry]

I believe that R6 should be 33k.

I’ve seen some schematics showing 3k3, but that drops the 50k pot down to 3k. With a 33k it ends up closer to 20k.

interesting...
not sure about the zonk specifically, but:

the 'Gary Hurst' version of the MkI has this 33K:


but the 'Sola Sounds' version MkI omits this resistor completely (as does the Tone Vendor MkI)


so yeah I guess it does seem kinda ludicrous for that value to be so low at 3K3.
seems like it'd be a real struggle to get Q2C to rise/drop to that ideal 4-5V target at max fuzz.
(I had enough trouble as it was when I breadboarded the gary hurst MK1 and struggled to find a Q2 leaky enough to get Q2C up to 4-5V)

 

[luizbyrro]

interesting...
not sure about the zonk specifically, but:

the 'Gary Hurst' version of the MkI has this 33K:
View attachment 118267

but the 'Sola Sounds' version MkI omits this resistor completely (as does the Tone Vendor MkI)
View attachment 118268

so yeah I guess it does seem kinda ludicrous for that value to be so low at 3K3.
seems like it'd be a real struggle to get Q2C to rise/drop to that ideal 4-5V target at max fuzz.
(I had enough trouble as it was when I breadboarded the gary hurst MK1 and struggled to find a Q2 leaky enough to get Q2C up to 4-5V)

The geofex document about the Zonk has a 3k3 resistor:
http://www.geofex.com/Article_Folders/zonkmach/zonkbst.pdf

And this blog lists it as 33k or omitted:
https://vero-p2p.blogspot.com/2022/12/zonk-machine-component-values-table.html?m=1

Probably all values are correct and they changed this value according to the batch of transistors they had at the time.

When I have time, I will experiment with the worst of the worst transistors and with the ones that have the lowest leakage, and see what happens.
For this build, I focused on using transistors that match the published specs (up to 0.050ma for q3 and 0.100ma to 0.400ma for q1 and q2).

Last year I didn't have time to build anything, but I somehow managed to buy lots and lots of germanium transistors...

 

[HamishR]

On a side note - for all of you non-Aussies out there when @owlexifry says "they are an absolute c*nt to get right" he is employing specialist Australian engineering jargon. Don't be alarmed.