marshall 2021 fuzz pedal project (help very much appreciated)

[Allanp]

Hi guys, I have decided that I want to build my first effects pedal. The end goal is the buzz bee violin type sound got by Justin Hayward on such albums like the blue jays and seventh sojourn. For some examples listen to this from the 2:43 mark

https://www.youtube.com/watch?v=iF2__-K8qFo

or this from around 2:08

https://www.youtube.com/watch?v=ITAf6ZgMPMo

That's the sound I want, and it was originally done with a 335 (which I have) through a very rare marshall 2021 fuzz/reverb unit, sometimes straight into the desk. The reverb part was never used, the fuzz part is very similar to what is found in a marshall supa fuzz pedal, schematic found here (I think this is correct, could be wrong as I am completely new to all this):

http://forum.musikding.de/cpg/albums/userpics/16602/SupaFuzzSchem.png

and an interior pic from a pedal with a near identical circuit board to the 2021 unit is here:

http://i.ytimg.com/vi/FhardaiHBek/maxresdefault.jpg

I have a picture of inside the 2021 unit, and the fuzz circuit board looks identical, save for possibly a larger capacitor for the 50uf. Can't see what is written on it, I guess it could be 50uF but I can't find any 50uF/12volt caps that large. Just looks a bit big. So, I have listened to a few supa fuzzes and a 2021 on youtube, all of different ages and to my ears they all sound a bit different to each other and all of them a bit different to Justins sound. The part that I want is in that difference I think. All the others sound like raunchy 70's stuff, where as Justins sounds like an electric cello or violin or something! I am hoping you guys with well trained ears can help me out in this respect.

So far I have ordered 3 oc75s, various types of resistor to try (though I doubt from what I've read they'll make a difference, I'll try metal films first), the mustard cap and the disk type cap and some 0.15 pitch vero board. So my question is regarding the remaining caps, the 50uF and the pair of 10uF. In the original, these caps were all 12 volt which is a hard voltage to come by in caps nowadays it seems. So, expert ears guys, not looking for necessarily the "best" caps, but which caps or type of caps do you think would give a sound most like what you can hear in the videos? I don't know how good your ears are, but can you tell if the large 50uF cap is actually 50uF or is it bigger, like 160uF? Would it make a difference like make it a smoother sound like what Justin has? What about the othet 10uF caps, would these be just electrolytic or could I use something else? I can't find 10uF at 12 volt at all. I know it's safe to go higher but what would I get to get me closer to that particular sound? Like I say, I am new to this, and I know there are many topics asking which is the "best" cap to use in effects, but I am not looking for "the best". I am looking for this particular sound which was made at a time when there was so much variation, you could make one with the exact same parts and it'll sound quite different. I think they stumbled on a happy accident with that particular sound and I want to recreate it. I may just breadboard it and tune it by swapping parts or something, but I wouldn't know where to begin with that.

Any help you guys can give would be much appreciated.

[Kipper4]

The 50uf cap is a power supply smoothing cap
so it smooths out ripple from the power supply.

you can use a 47uf rated at 16v for this.
it is not a part of the audio path. you can use any cap that's voltage is rated above the power supply, so seeing as you will be using a 9v wallwart/battery.
Do not use any caps that are voltage rated below 9v.
beware the orientation since this is a -9v circuit.
And welcome aboard buddy

[Allanp]

Thank you very much for the info. Ok, that helps cos 47uF 16v is easy to get. Uhm, I have some noob questions 

So as long as that power smoothing cap is the right values it can be any type or does it need to be polarised, electrolytic or whatever?

As it's not part of the signal path, it's quality does not effect the sound?

Why would the battery powered version still need the smoothing cap?

Are the 2 10uF caps part of the signal path and if so, which ones would be best to get that particular warm fuzzy tone of Justin Hayward, or does it even make a difference as long as they are the right values?

And finally, out of all the components in this particular circuit, which ones do you think are most responsible/critical for a)unwanted noise and b)that particular warm cello like quality to the fuzz tone that seems to be quite unique to JH?

[Kipper4]

Thank you very much for the info. Ok, that helps cos 47uF 16v is easy to get. Uhm, I have some noob questions 

So as long as that power smoothing cap is the right values it can be any type or does it need to be polarised, electrolytic or whatever?
Doesnt matter electrolytics are cheap so thats what i use

As it's not part of the signal path, it's quality does not effect the sound?
Not as far as i can tell

Why would the battery powered version still need the smoothing cap?
Shouldn't be any need if battery powered only.

Are the 2 10uF caps part of the signal path and if so, which ones would be best to get that particular warm fuzzy tone of Justin Hayward, or does it even make a difference as long as they are the right values?
Sorry cant help there I dont know the circuit. Ill take a look later and see if i can figure it out.

And finally, out of all the components in this particular circuit, which ones do you think are most responsible/critical for a)unwanted noise and b)that particular warm cello like quality to the fuzz tone that seems to be quite unique to JH?

Again Sorry cant help
Maybe JH himself has something to do with that sound.

[Kipper4]

Are the 2 10uF caps part of the signal path and if so, which ones would be best to get that particular warm fuzzy tone of Justin Hayward, or does it even make a difference as long as they are the right values?
Yes they are part of the signal path and to be honest in a FUZZ i wouldnt think you would hear much differance between a cheap Tayda electro and a hifi Wima.
Its a distortion pedal not a high end hifi amp.


And finally, out of all the components in this particular circuit, which ones do you think are most responsible/critical for a)unwanted noise and b)that particular warm cello like quality to the fuzz tone
I still have no idea but i guess the transistors and the two 100k would have something to do with it. Take that with a pinch.
I've never built it and I'm not much for fuzz mungering.

[Ben Lyman]

I'm just a novice myself but I've learned so much from the generous folks here.
I say definitely breadboard it. I also say, don't sweat it too much, it probably won't be the last pedal you build. Maybe even start off with something simpler, you can get very close to those tones with a Fuzz Face. I think a big part of the smooth violin tone is using the neck pickup in combo w/the tone knob on the guitar, a 335 should get you there easily. Also remember, a SupaFuzz is a Tone Bender and i found an easy to make Si TB MKII schematic online. Took about $3 in parts and I had it on the breadboard rockin' out in about 15 minutes, neck PU & guitar tone knob down=violin w/endless sustain. I didn't box it up but I learned how to make it and have since moved on to more expensive NOS Ge tranny's, feeling ready to box up some real pedals soon.
Enjoy & welcome aboard!

[idy]

Yes, read threads about tonebenders. The thing that accounts for the differing personalities of various individual units is the transistors, especially their gain. You bought three, were they a measured/matched set for tonebenders? Three random OC75s are unlikely to give you any exact sound you may be seeking. At least you would need to try switching them around.
I've built TBs and Buzzarounds and other Ge fuzzes. They always work when you get transistors in the "accepted" gain ranges. And sometimes when they are out. And leakage may not be such a deal breaker.

And yes, a silicon model will probably make you happy and will eliminate much confusion. Just for testing you can put low fain PNP silicone in your 2021. Make sure if works before you go crazy worrying about transistors.

[Allanp]

Thank you very much for any help you can give guys.

@ Kipper

I'm sure JH has a lot to do with how good it sounds, but there are some things that make me wonder. JH himself has said that he hasn't ever found another effect unit like it, and couldn't use that effect unit live because it was too noisey. As a result, he never had that tone live. Still a fantastic and highly under rated guitarist for sure, but never has he had that sound live, despite it being the same JH with the same 335. So whilst he's putting the heart, soul and melody into that sound, I am pretty sure the sound effect itself is coming from that 2021 unit. Many artists at the time used marshall supa fuzz pedals and I am sure marshall made many 2021 units (which has basically the same fuzz circuit) and yet I have never heard another recording from anywhere else with that sound, and JH never saw fit to use a supa fuzz, so I don't know what it is with that 2021 unit, maybe it was "faulty" or some components weren't up to spec which lead it to have that sound. With the help of you guys (maybe some of you on here can actually hear where there might be an issue with the fuzz circuitry, like I say I don't know how good your ears are!) I plan to make a copy of the original circuitry and if it doesn't sound "right" I can alter it from there on the breadboard. I am a maintenance engineer by trade so I am probably just over thinking everything as usual. But yes, with the help of this forum I will bottle my tone holy grail (which may or may not have been brought about by accident of 60's component inconsistency), then after I've done that I'll bring about world peace, which will probably be easier!  But hey, nothing good came from always taking the easy road 

@Ben

So many replies already, this forum really is generous  I am now feeling much more encouraged  and yes, breadboard then mess about till happy is likely the way to go.

@idy

According to the ebay seller they were a matched set of 3 for tonebenders that I have ordered! I have an electrically savvy friend who can double check them for me. But that is very useful info, so it's the transistors and their gain that's the big thing, interesting. What do you mean by "silicone model" and "low fain PNP silicone"? Do you mean cheap silicone transistors to use just for testing to make sure it works before I ruin some nice OC75s?

[Ben Lyman]

Alan, I believe that is "low GAIN silicon" transistors. Ge (germanium) transistors are all over the place when it comes to gain and leakage, also VERY temp. sensitive, you will need to read up a bit on how to check the gain (HFE) but then again you have a matched set on the way and they will probably be marked, Q1, Q2, & Q3.
Si (Silicon) on the other hand is very stable and easier to work with, i.e.: you won't have to handle them with tweezers in a 70° room to measure gains, set the bias, etc.
Whether Ge or Si, they come either PNP or NPN.
PNP, like you ordered for the TB will be grounded by +9v
NPN, like a lot of modern circuits use the -9v as ground.

[stringsthings]

But that is very useful info, so it's the transistors and their gain that's the big thing, interesting. What do you mean by "silicone model" and "low fain PNP silicone"? Do you mean cheap silicone transistors to use just for testing to make sure it works before I ruin some nice OC75s?

Ge transistors ( like the OC75's ) were used in many old fuzz boxes because Si transistors weren't around.  When Si transistors were invented, they could be made much cheaper than Ge ones, and they were more consistent.  Over time, Si transistors replaced Ge ones and the Ge transistors gradually became harder and harder to find.  That's why they're so expensive these days compared to the Si ones.  Ironically, the Si ones work "better", especially in modern circuits. 

But for those old fuzz box circuits, the Ge ones work better. ( as long as the gain and leakage is in spec to what the original circuits used )

There are others here in the forum that are much more knowledgeable on this subject, but that's a brief history.

[smallbearelec]

Hi--

A three-stage tone bender is Not what I would have recommended as a first build; too many parts, too finicky about transistor gains and biasing. But you are clearly determined, so I hope you have a breadboard

http://smallbear-electronics.mybigcommerce.com/breadboard-strip-with-power-bus/

and were planning to set up the circuit on it before committing to solder. Make it work on the breadboard first, then find a vetted layout to build.

[Allanp]

Is it complex? I didn't know that this would be a tricky one to start with, lets hope I'm a fast learner then!

Yeah, I do plan on breadboarding it first. As for layout, I have some 0.15 pitch vero board on the way and I'll lay it out exactly as done in the original marshall fuzz 2021 unit/supa fuzz. I have a shot of the underside of the supa fuzz circuit which should help. As for the transistor biasing, this is the first time I have even heard of that. My assumption was that I could get all the same parts at the same values, put it into the same vero board with the same lay out and wiring and all would be well. Is this my nooby ways showing through?

[samhay]

>Is this my nooby ways showing through?
yes, but you are not the first to have made similar assumptions.

These old circuits will generally work a lot better if you spend some time tweaking them. Making part substituions to perf and vero is not much fun unless you use a lot of sockets, so if you can spare the time and mental energy, buy yourself a breadboard - as Steve suggested - a pack of a few common value components (resistors, caps and perhaps a few Si transistors) and a cheap digital multimeter (so you can measure voltages).
Once you have come to grips with a breadboard - there are some nice tutorials on the smallbear site - build the circuit on this (it is all non-destructive/reversible) and post voltages.

And welcome to the forum.

[antonis]

So as long as that power smoothing cap is the right values it can be any type or does it need to be polarised, electrolytic or whatever?

Whatever..!!

Caps at this capacitance and voltage rating have to be electro or else you'll need much more money and space... 

[Electric Warrior]

As you're using OC75s, you shouldn't be far off. The circuit is already biased for these. But it usually helps to have more than just three transistors to swap them around until it sounds and behaves right. I usually socket the transistors using inline sockets before I make the commitment of soldering them in.

This vero layout may be helpful:



And here are voltages from my vintage unit:

Battery: 9.67V
Q1 C -9.02V B -0.03V E 0V
Q2 C -0.17V B -0.08V E 0V
Q3 C -8.44V B -0.17V E -0.11V

Q1C measures rather high in this one. They usually measure closer to -8.5V there, but there's a bit of a range..

[Allanp]

Ah, thank you. I wonder why they swapped the 10uF for a 64uF. Your point about possibly needing to try other OC75s and swapping them round is duly noted, thank you.

I have acquired a breadboard, multimeter, every value of resistor I could need and most of the parts have arrived, just waiting on the important bits. Once I get those I'll get to building something a keep you guys updated.

[Electric Warrior]

No idea about the 64µF. It's a rather uncommon variant.

Good luck with your build!

[Allanp]

Ok, so went to a friends house, he's into electronics, and we made it on a breadboard with all modern parts, and modern pnp silicone transistors. He explained to me the very basics of biasing, not enough to do it myself but enough to know it's a big deal. We put it on an oscilloscope and compared what was going in (300hZ sine wave @ 1v) to what was coming out, and the circuit seems to work. At full volume the output was very slightly louder than the input (very slightly larger amplitude) and the wave form was nice and jagged, and could be smoothed and altered quite alot with the attack control knob. We decided to listen to the output through an amp (not a guitar amp, just regular hi fi kind of amp) and although it was only a sine wave going in (I thought he had a guitar but he didn't, would have been helpful!) it sounded like a fuzz kinda sound. Obviously not expecting greatness with a signal generated sine wave and the so on, but the circuit seems to work.

Nows here's the thing. I tried it with my guitar this morning, and I could only get a fraction of a second of sound, like a click and only when strumming really hard. My assumption is that the modern transistors are needing something like 0.9 volt to switch, that's why my friend turned up the signal generator to 1 volt, at which point it worked as expected. lower than that and it didn't give out much. When turning it up, my friend said something about accounting for the bias, and that OC75s only need 0.2 volts to switch. I am also assuming that only at the loudest part of a really hard strum on the guitar is giving me the needed 0.9 volts to switch the transistors. So I am also assuming that with OC75s it should work as the circuit was designed (and biased?) to work with those. Also that it might be a good idea to put a compressor before this fuzz unit in the signal path. Are any of my assumptions correct?

[Electric Warrior]

This circuit needs leakage for Q1 to bias. You can't just substitute the transistors with silicon.

[Allanp]

Ok, so all the parts and oc75s arrived and I breadboarded it and it works fine, but the sound is a little off.

1) there is a lot of crackling, like a scratchy record.

2) if I turn the attack down, I get closer to the smoother sound I am after, but it seems as though I need to be able to turn it down further so that it's less harsh. This thing sounds like a beast!

3) not sure how to describe this, but it sounds quite closed and almost muffled. Like, if a neck pickup is closed and almost muffled sounding and the bridge pickup sounds more open, I would like this to sound more open.

What I have is close enough to confirm that it's a fuzz like this that he was using, but Justins tone is so warm and full, not harsh at all, really smooth and fully open sounding. Any ideas how to solve these problems?  BTW I am using a historic re-issue 1963 es335 with has stock classic 57 PUs straight into the desk if that's of any use to anyone.