scatterbox... npn ge version of the HS Shatterbox

[pinkjimiphoton]

haven't built it ..... YET. probably will tonite.
if so, if it works, i'll post back.
if someone is crazy enough to try and build this before i get to it, please let me know.

that said... here ya go. been over it a bunch of times, it should be good to go.

here's the "original layout"



here's the "vero layout"



here's MY layout



what's different?

well, i made in npn to play nice with pedal board daisy chains.

added power supply filtering so it will play nice with wall warts.

added trimmers to the collectors to make biasing easier

set it up for GERMANIUM npn's. silicon will probably work, if the gain is low enough, but you may have to play with some resistor values.

tried to go with everydays bog standard parts, not esoteric values, as this probably had a 20% part tolerance AT LEAST...

i was gonna add a small r/c filter to e of q1, which i've been doing lately.... as little as 10-47r can make a HUGE deal with stability.
if you do that, add an electro, somewhere between .22 and 22u to help boost the gain up, and give it some sparkle. (just like a cathode bypass cap on a 12a_7 would....thanks to Gus for hipping me to this)

the more i play with transistors, the more parallels i see to tubes.

c would be the plate, e would be the cathode, and b would be the grid. it's a LOT more similar than i think some folks realize...and you can interchange a lot of the same tricks.

anyways, i hope to get on this tonite maybe, so hopefully soon it'll be vero-fied. i went over it a bunch of times trying to refine it (and leave space for top hat ge's) and it should be jake, but if ya find any mistakes, please let me know!!

rock on!! more fuzzes in the works still...

edit: duh!!! also added a master volume, which may or may not be necessary.... so stay tuned!!

[wilrecar77]

Glad to see you're putting those germaniums to good use! I haven't checked most of mine, how are they for leakage?

[pinkjimiphoton]

hey pat!
thanks bro... i'm loving these things!!
i've been meaning to build the leakage test thing, i even bought the 1% resistors...

but... 

i usually chuck 'em in my meter and see what it says. if they look ok, i try 'em.

the meter's reading between about 70-120 most of the time, so i have no idea what their true gain is or the leakage.

but they sound great!

[LucifersTrip]

set it up for GERMANIUM npn's. silicon will probably work, if the gain is low enough, but you may have to play with some resistor values.

shouldn't this statement be the other way around, since it was originally silicon. Electric Warrior posted the schematic...

[pinkjimiphoton]

i know bro, i used that schematic.

it has a mistake. the 500k swell pot is in the wrong place as shown. i built this last nite, and it will fire as shown on the net, but it's most definitely
not right.

this was silicon? i thought this box predated that?

the footswitch wiring that accompanies this on the dam forum is also wrong. soo....

the footswitch wiring is wrong, and the pot is in the wrong place. already fixed on mine, just gotta draw it up.

as shown, this thing is fairly useless.. the fuzz is kinda weak, and barely to unity, and when ya hit the boost, kapow.. about 30db louder. it's ridiculous unuseable and unmusical.

i looked at it long and hard and did some head scratching, and realized if ya look at this, it makes sense.... if memory serves, this was a fuzz with a booster. maybe the fuzz had a 500k output, but i doubt it... i mean the "swell" knob.

it has a .1 cap off the c of q2 going to a 2.2m resistor feeding the wiper of the swell control. it doesn't need a 500k pot there. but if ya want a high impedance, 2.2m suddenly makes more sense (barely, to me at least) if you move the 500k to the other side, on the INPUT of the booster. then the booster works, you can turn it up and down, and the fuzz ain't pathetic and weak, either.

so... i'll work up a corrected schematic. i believe as posted now on the net it's wrong. it doesn't make any sense to put 2.2m resistance just before another 500k resistance.  but you want a high impedance if you wanna keep gain from one stage to another i'm guessing.

seriously... nite and day difference moving that pot in the circuit.

i ended up using npn ge's for q1 and 2, and went with a 3904 for q3 cuz it seemed to sound the best out of the ones i tried.

the vero, consequently, needs amendment as well to match my findings... stay tuned bro.

i think you'll be surprised at the difference when you look at it and hear it..

all it takes is one mistake on a schem or layout, and stuff happens. i believe this is the case here.

too burnt to work up schem/vero tonite, but will post tomorrow, with the corrected footswitching as well.

the footswitching as shown on the dam forum will work, but not RIGHT. the swell pot has gotta be relocated, then suddenly it all makes consonant
sense and works as described and demo'd.

peace

[Electric Warrior]

i know bro, i used that schematic.

it has a mistake. the 500k swell pot is in the wrong place as shown. i built this last nite, and it will fire as shown on the net, but it's most definitely
not right.

Nope, this is how they made them.

this was silicon? i thought this box predated that?

If you had bothered to look up the datasheet, you would have known that the 2n4061 is a PNP silicon device.
And no fuzz box predates silicon transistors. They were introduced in 1954. The Zonk II/Shatterbox dates from around 1969 and certainly wasn't the first silicon fuzz box either.

the footswitch wiring that accompanies this on the dam forum is also wrong. soo....

I've added true bypass switching. What's the problem with that?

as shown, this thing is fairly useless.. the fuzz is kinda weak, and barely to unity, and when ya hit the boost, kapow.. about 30db louder. it's ridiculous unuseable and unmusical.

The Shatterbox has the reputation to be one of the shittiest fuzz boxes ever made. And not in a good way. Despite its reputation everyone who plays one is still surprised by its shittyness when they get to play one - go figure.
The fuzz in my clone hits unity easily. The booster is more of a bass cut. It doesn't boost at all. Vintage units had the same problem. Depending on how you voice the fuzz they work ok together - or not.

The only good reason to build a shatterbox is to nail those early T.Rex tones, really.

i looked at it long and hard and did some head scratching, and realized if ya look at this, it makes sense.... if memory serves, this was a fuzz with a booster. maybe the fuzz had a 500k output, but i doubt it... i mean the "swell" knob.

it has a .1 cap off the c of q2 going to a 2.2m resistor feeding the wiper of the swell control. it doesn't need a 500k pot there. but if ya want a high impedance, 2.2m suddenly makes more sense (barely, to me at least) if you move the 500k to the other side, on the INPUT of the booster. then the booster works, you can turn it up and down, and the fuzz ain't pathetic and weak, either.

so... i'll work up a corrected schematic. i believe as posted now on the net it's wrong. it doesn't make any sense to put 2.2m resistance just before another 500k resistance.  but you want a high impedance if you wanna keep gain from one stage to another i'm guessing.

I can assure you they have a 500k fuzz pot and 2M2 resistor at the output.

Some reference pictures can be found here: http://decadeoffuzz.com/index.php?/project/john-hornby-skewes-pedals/

This is the unit I've based my schematic on:

[pinkjimiphoton]

my apologies, and i stand corrected.

i think maybe i figured out where THEY f'd up then.

gimme a little bit. and yah, you're right, i shoulda looked up the data sheet, etc etc.

moving the switching and the swell pot completely changes it into a very different sounding circuit despite being the same for all intents.

i bet you the peeps creating these in the first place this is where they dropped the ball.

as shown, in pics and your schem and layout, you're one hundred percent right. all the way.

and no offense was intended. i'm thinking the mistake was something that THEY made.

i built this as shown on the schematic you laid out, and tho i made my own vero (don't give a hoot about mojo personally) it matched
yours node for node, and your schematic as well.

i even bet petey twofinger 10 bux it wouldn't fire up when i powered up. not cuzza you, cuzza ME.

it did. had fuzz and a boost right away. the fuzz was a bit weak. i used trimmers so i could tweak the resistors to c's of both transistors.

in this case, a trio of npn germaniums pat had sent me a while back.  i had originally also included a volume control, simply spliced in between the output of the board and  the return of the footswitches. tried 100k, didn't work. the swell knob didn't do much of anything, other than kill the pedal when turned down... if you turned down the boost, you killed the fuzz.

dialed it in a little, and started thinking about the circuit and it just plain makes more sense to have the swell pot at the input to the .01u coupling cap going to the b of q3 coming from the node at the end of the 2.2m resistor. this also means the wiring of the whole rats nest needed to be reexamined.

what i found was to move the pot i had to change the wiring to the footswitch to accomodate the pot getting spliced in. it is almost the exact same wiring, for all intents it's the same goddamn circuit... but by moving that pot to the input of the boost instead of the output of the fuzz, and switching between them, you have a loud strong high impedance signal feeding a lower impedance signal that is still more than high enough to drive the final stage with plenty of fidelity for a FUZZ box.

so i did it. and it works. and i mean really works.

electric warrior, i totally believe you traced it and traced it right. with all my heart and soul, bud, you did it exactly right.

what i'm saying is whoever originally built these things i'd be willing to bet screwed 'em up, figured, wtf, it's a fuzz box, it sounds horrible, the kids will love it. or they didnt even realize their mistake... or maybe they figured it would make it easier for production.

beats me. but i bet when they drew it up (did they true bypass this?) they made the mistake i believe i have found.

while one may argue and rightly so that the tone of the box may be different because i prefer germanium in my fuzzes and that's a completley different discussion, but i'm talking from a standpoint of use by actual musicians, and based on function and tone above and beyond worrying about whether it is bone stock vintage correct.

in this case, the function would expect is restored. the fuzz and boost are independent and footswitchable easily. you have control over fuzz level with the fuzz knob, and boost level with the swell knob. it plain doesn't work right stock, and again, i believe that was their mistake, not yours.

when i get a chance i will draw up what i'm talking about , but you can do it too just as easily. if you move the pot it changes the circuit just barely, and the shitty sounding shatterbox becomes a very cool sounding scatterbox.

i'll post video thru a real amp even, and do it before i box it so you guys can see the circuit, and give ya the vero i worked up and the schematic so you can see for yourself.

further... electric warrior... without you, bud, this never would have happened in the first place, so big time props bud.

i guess i was obtuse at 1am or whatever..

i didn't know who drew this up originally... and it may be right, but its not right right.

you drew this up from an actual unit, working, and sounding like all the other ones i'd listened to.

the guy who built the circuit, or designed it is the one i am saying made the mistake, and i believe i can prove that...

or if not, at least humbly offer a way to make these things not suck.

btw... i did not use ge in the booster. i used a 3904, after trying various other npn's or different gain ranges there, and it had a good balance and decent tone. bc549 was close, but the 3904 just sounded better.

i will tinker with si transistors in it, but i in general will go ge even in si circuits when possible.

anyways... no offense intended or taken, man, sorry if i came off like a vinegar and water monger rather than a fuzzaholic.

stay tuned..

[Electric Warrior]

No prob  

I don't think that's where things went wrong. The booster wasn't meant to have a volume control. It was meant to be part of the fuzz circuit. The standalone Zonk II and JHS Treble Booster used the same circuits.

Both parts of the circuit were documented before I traced this one. What I did was mostly confirming things (like the booster booster having a 0.001µF on the output with several different value input caps used over the years) and documenting this early variant (with 270k on input and no 3k9 across the fuzz pot).

The move from germanium to silicon must have been the problem. The original germanium version of the boost supposedly sounds much, much better than the silicon circuit (it actually boosts  ). There was also a germanium version of the Zonk II, which hasn't been properly traced yet.

That said, the Zonk II can be made to sound ok on its own with stock values. It's not the most useable fuzz, but no volume issues here:

In the Shatterbox you'd probably want to tune it to have less treble than that, so it doesn't get too ear piercing with the booster switched on. I believe both owners of vero board Shatterboxes I know told me the fuzz was rather bland in theirs..

[pinkjimiphoton]

thanks for being so cool bro.


here's the thing... as originally built, yah, it sounds like the youtube videos... but when ya move that one part, suddenly it becomes a completely different animal. i don't think the impedance is hight enough at the base of q3 to do much after all the losses from the switching and stuff.

the original zonk machine had fuzz and volume, right? i'm thinking maybe when they tried to add the boost, they figured it would be better to just tack on the boost to the end of the circuit. it certainly looks good on paper, until ya try it... then .... not so great.

also... the wiring as shown on the dam forum is perfectly correct, if you have the switching after the swell pot.

but when you move the pot, which i bet they didn't think to do cuz the lsd25 was flowing strong in them days.... it becomes a different animal.

i'm finding more and more that correct to the unit schematics often also contain mistakes made by the original producers... like when lucifer's trip found the "fix" for the umi buzztone.... i'm betting our dave is right on the money, and the people that traced the originals like leper and rg and them were right, too... tracing things from actual units... but i bet these units were BUILT with mistakes. ludwig pII is another one, that schematic had all kinds of issues we discovered. i'm humbly suggesting this may be the case as well.

if i can get off the net for a little bit, i'll try and get the schematic drawn up... i still kinda suck with drawing switches, but will do my best!!

stay tuned bro...  worst case, you helped me develop a "new" "vintage" fuzz. best case, we figured out where HS's imagineer did too much acid... or too little.

thanks for the info on the history of this. i'll post what i find over to the captain's forum too, which is where i found your schematic and layout in the first place.

rock on!

[kingswayguitar]

i like that zonk II vid way better than that t rex example.

[pinkjimiphoton]

just played with this for a couple hours, messing with things, swapping transistors, including silicon. which sound pretty much just like the demos posted.
but germanium just plain sounds better... in the fuzz. in the booster, germanium is pretty much just a treble boost.
with a 2n3904 in there, it becomes anything from a treble boost to a significant volume boost with the mod done to the swell placement.

i think the swell control is where they screwed it up. simple mistake!! i've done worse!!!

until i get schematic/vero done up... here's deets.

b+: 9.31volts

Q1&2 : heathkit ge npn house number#:R609980243 metal top hats nice gold leads.
q3: bog standard 2n3904

q1
hfe: 79

50k trimmer set to 13k with q1 out of socket, read from input to c socket

c: 2.85
b  .10
e  0.0


q2 hfe: 90

10k trimmer set to 9.9k, same as above

c: 6.19
b  2.88
e  2.74

q3 hfe: 209

c: 2.37
b: 1.28
e  .68

in this case, a 10k trimmer from + to c.... i haven't read the resistance yet, as this seems to be where i can set the boost's gain to get a decent boost and a bit of control over the tone.

also... looking at it... the 100k resistor across the output jack (!!) is probably a dodge to try and bring down the volume to a useable level... again, cuz they put didn't move the swell control as they should have..  i actually cut the lead on it so i could try both ways. it does have a minimal effect on the tone. i didn't put it across the output jack tho, i put it from the output of the .001 cap to ground. to me, there's no point using tb and having a resistor across the jack that's always in the circuit!!  again... this is how the original is, and one more reason i think it was a dodge.

so... the mod....

thinking of the numbering of the pins (right or wrong, no idea, but to explain it) going counterclockwise from where pin one would be on an ic, with pin one top left, i number it 1-6 for both switches.

input and output, jumpers and connection between pin 5 on both switches is fine.

node at end of 2.2m resistor connect to switch 1 #6
input to board, node of 10u and 270k resistor , connect to switch 1 #1

switch 2 #1 to output of booster,  node .001 cap
switch 2 #6 to wiper of swell control, top of swell pot to node at end of .01 cap,  bottom of swell control
to ground.

that's it. totally different beast!!

more later...

[pinkjimiphoton]

here is what i believe is actually what they intended, but screwed up back in the 60's.
this is built, verified, and working as expected.
adjusting tr3 allows you to set the over all volume of the boost from below unity to pretty well above.
the actual transistors i used are in ge npn's from an old heathkit organ, house # r609980243,
q1 hfe 80, q2 hfe 90, and q3 is a bog standard 3904, hfe 210. a ge transistor in q3 gives more of a treble and distortion boost than volume boost.
the footswitch wiring is correct and modified to allow moving the pot to where i believe the proper position in the circuit should have been.
thanks to electric warrior for tracing an original, and leading to this possible discovery.
the one i built, once biased to the "sweet spots" for the transistors i used is a very nice dynamic fuzz, and a good solid boost completely independent.
by moving the swell control after the bypass switch, it's actually useful.
i believe this was the intention of the EE who developed this pedal.. you want a high inpedance input signal...2.2m is good and high.

by moving the 500k swell pot to the input node of the booster instead of the output node of the fuzz, i believe the 2.2m gives a high enough impedance to allow enough signal to get thru the switch, and hit at a high enough impedance at the input of the now moved swell pot to hit the booster section a lot harder... think of it like a tube amp, where you need a grid load resistor going to allow the tube to function.

i'm not savvy enough electronically to explain it any better than this, sorry.

all i can say is try it. moving that pot takes this thing from crappy relatively useless curiosity to an actually useable, toneful pedal that acts as you'd expect a fuzz into a boost to act.

anyways... first attempt with DIYLC, not the best schem to be sure, but i am 100% sure it's right.

consider r7 as optional... i believe it was a dodge HS used to try and compensate for dropping the ball during the design process.
i can see no other reason why on earth anyone would bleed signal across the OUTPUT JACK with a 100k resistor!!
i tried it with and without.. it does smooth the signal just slightly. i wired it not across the jack, but the output node from the .001 cap.

anyways... schematic. stupid pedal trick video and updated veroboard layout to come soon.

peace.

[kingswayguitar]

happy sunday
love that font jimi
peace!

[mac]

i can see no other reason why on earth anyone would bleed signal across the OUTPUT JACK with a 100k resistor!!

anti-pop 

mac

[Electric Warrior]

I think it's meant to be part of a high pass filter at the booster's output. The treble booster always had it. The Selectatone (which is basically the treble and bass boosters switchable in one enclosure) only had it in the Treble boost path.

[pinkjimiphoton]

I think it's meant to be part of a high pass filter at the booster's output. The treble booster always had it. The Selectatone (which is basically the treble and bass boosters switchable in one enclosure) only had it in the Treble boost path.

REAL good call bro, cuz without it, it's a little jagged. doesn't make all that much diff, but it is noticeable.
but it's still making me think more and more it was a dodge to try to compensate for the ridiculous boost the way they had it..
i mean, it just seems like if they wanted it to limit the current or act as a filter, or even an anti pop, why not put it on the board itself, before the switch? after the switch, it affects the tone even in bypass.

i know i'm monday morning quarterbacking 50 years later, but i recognize some of the dodges as mistakes i've made myself and tried... 

it was too late last nite to mess with it, and today's my girlfriend's birthday and i think the last thing she will want is a stupid pedal trick... 

i'm thinking, the swell control made perfect sense on the zonk machine where it was. but i believe they made a mistake, i can almost hear the guy going... "it's a FUZZBOX... if it sounds like poo, the kids will love it!!" and everyone assumed it was the way it was supposed to be.

but i mean, from a simple standpoint, looking at it, if ya wanna booster, you want the volume at the beginning of it if you want boost, or at the end of it if you want dirt... not before the thing even starts, it's the equivalent of turning your guitar down at that point... and then turning the switch off.

i'll stop raving now.

btw... the font is called magical mystery tour, and i believe it's freeware. there's two versions, one is dark letters with a light shadow, the other the opposite.

E.W. i'm gonna look at that last schematic you posted a bit closer once i have some coffee and a doobie in me.
thanks bro!!!


i gotta ask, for my own (in)sanity.... do you guys think i may be right on this?

i can't wait for you to hear it.

also, i realized i made a mistake on mine, which may account for the volume problem i was having, i used a 10 k instead of 50k trimmer on Q2... so i think i'm limiting how much gain i can dial in... the board is so small i dunno if i can get it apart again to change it.

thanks gents... onwards and upwards...

[pinkjimiphoton]

I think it's meant to be part of a high pass filter at the booster's output. The treble booster always had it. The Selectatone (which is basically the treble and bass boosters switchable in one enclosure) only had it in the Treble boost path.

hmmmmm... this too looks like they did something kinda weird.. if you put caps in series, it's like resistors in parallel... why have a dc blocking cap off the collector, and then a switch between a small cap and  a big cap? wouldn't it make more sense to just use the treble and bass caps after the switch, and not before?

i could see the reason to block dc, but the other two caps are gonna block dc anyways!!

thanks for this schematic... it's making me wonder more and more if the peeps that built these things actually followed the original designs, or just thought they did!!

just found this on the link you posted bro, (thanks, never saw that page before!) :

"He carefully examined several of my Zonks (and the others) and said that although he might have had an advisory role at the very beginning - it does bear resemblance to a modified Mk1 Tonebender - he does not recall actually building them anything - BUT THEN as he pointed out - he was working flat out in that 1966-1968 period and it is possible that another Macari employee could have had a hand in it. Gary did seem to think that a builder he knew "up in Birmingham" just may have built the earliest ones - he said it clearly wasn't his work as there were instabilities inherent in the design which he'd never have allowed to stay in there (a little hard earnt self pride coming in there methinks!)"


sounds to me like gary hurst had the same issues i found, and that hornsby skewes didn't wanna fix it.

instabilities.... like a useless knob, perhaps?  

edit: sorry... too little sleep, too many bingers and not enough coffee yet...

[Electric Warrior]

He's talking about the earliest Zonk Machines (Zonk I), which are similar to the MKI Tone Bender and Maestro FZ-1 and sound really good. But those had the 2.2M and 500k pot, too. Go figure:

The cap at the selectatone's collector was part of the bass booster's filter - also used in the standalone bass booster:


It's probably a 0.1µF, so it hardly makes any difference to the 0.001µF value when put in series with it.

[pinkjimiphoton]

ah, but on the earlier one, it looks like they had the swell knob right... at the end of the circuit.

it looks like they tried to slap a switch in between that model's second and third stage, and figured they'd put the volume there, after the fuzz. seems reasonable. hell, i've done stuff like that... and had the exact problem i ran into here!!
i had that issue on the lunchbox amp i built for dick wagner, drove me crazy.

i bet thats what they did... and then after modded it slightly into the boost it ended up being, probably cuz they killed the tone in the first model.

maybe i did too much lds at berkley.... i dunno, but it sounds and looks plausible the more stuff i'm seeing. and from the results of my tinkering, blowing stuff up, etc.

thanks for the info bro.... what's your name/preferred handle?

gonna look at tis more... and i'm gonna definitely try building this up as the earlier model. i bet it will rock if done right... maybe add the boost at the end of the original, but with the swell control moved? in that case it may well need a volume control somewhere on the fuzz, for sure!!

[pinkjimiphoton]

ok, vero cuts n jumpers



as seen from the "rails" side



and here's a verified vero, the thing in the video:



stupid pedal trick:

http://www.diystompboxes.com/smfforum/index.php?topic=105438.0