Vintage kit fuzz - 1.5 to 9v conversion advice?

[Halkbi]

I stumbled upon this circuit today and thought that it would be a fun little build since it's a different take on the Fuzz Face that I haven't seen before that looks interesting enough. Behold, the GU-15 Single Guitar Booster 



The thing is though, this is supposed to be powered by a single 1.5v battery. Given my previous experience with such units - and the fact that this is just another Fuzz Face variant - I would prefer to run it at 9v for increased headroom and output. I have no idea on how to change resistor values for proportional bias or even if that's the way to do it. Original transistors are listed as having a gain of 125+, so I think I'll be using BC108s.

How would you approach something like this? Please explain thoroughly, it's been close to ten years since I was into eletronics.

[anotherjim]

Get a solderless breadboard & try it? All the resistors except the input R1 might be about x6 higher value if go from 1.5v to 9v supply. The capacitors can probably stay the same.
Both transistors are being used without thought for clean & linear operation (not having emitter resistors is a clue to that), so it won't be all that fussy. "Gatey" or "Blatty" sound might be the worst outcome but adjusting R6 should take care of that.

[iainpunk]

its not really like a fuzz face at all.
it doesn't have the feedback bias, it doesn't have the output attenuator, Q2 is also grounded etc...
this thing might have more volume than a fuzz face has, despite being only 1.5v, since it does omit the output attenuator!

i recommend R3 and R5 be trim pots, 10k, with a 5k6 series resistor, and bias them both until they sound good.

cheers, Iain

[PRR]

(as Iain says) It is not a fuzzface and it is NOT clean for any reasonable input.

Q2 will overload at 30mV INput. Q1 has gain near 50 so the box overloads at about 1mV from the guitar. Which is very small guitar. It will be fuzzed-out 99% of the time.

The battery affects the OUTout level. Note that a Fuzzface typically has 10:1 divider off Q2 Collector, plus usually a pot. A 1.5V battery with little or no divider can make the same output.

Go ahead and over-volt it if you wish. But I think you are wasting electrons.

[Mark Hammer]

I just use a zener diode to ground from the V+ terminal to reduce the supply voltage to whatever I need.  For a supply voltage this low (1.5V), one can simply use three silicon diodes in series to provide the required voltage.

Normally, one might see a single silicon diode used for protection against use of a wallwart with the wrong polarity at the plug.  That drops the "wrong" voltage down to around a half-volt, which won't likely damage the circuit.  Using three diodes in series makes the supply reduced to whatever the forward voltage of the three diodes is. Note that the tolerances of typical 1N914/1N4148/1N4001 means you might end up with 1.8-1.9V, so measure and pick the diodes that get you the desired voltage drop.

[Rob Strand]

The root problem is you want to change something but perhaps not change the tone.
This is where it changes from a mod to a redesign.

You also want to increase the headroom.   The problem with that is, if the sound depends on the second stage clipping
you might not be able to achieve all your goals.

From what I can see the original design really does need hFE = 100 to 125 devices to bias properly.  If you use BC108's and they have a hFE of 200 then it's not going to work as intended.   Because of this it is inevitable you will need to adjust the collector resistor on Q1 and Q2.

Option 1:   Simplest change.   Make a 1.5V diode regulator and power the whole circuit from that.


Use circuit (a) to make 1.8V from 9V.  Perhaps use 4.7k instead of 1k.  You  might need to add 10uF to 100uF across the diodes.
The voltage across the diode is 1.8V and acts like a battery.


Option 2:  Aim preserve original clipping of first stage but increase output.

Power the first stage using the diode regulator.  That means putting the diode regulator between R4 and R6.

Next make the following changes:
- *Add* 220k from base of Q2 to ground
- Change R4 to 2.2MEG
- Change R5 to 10k.

These changes preserve a number of features of the second stage.   Good enough?  maybe maybe not.

If necessary adjust R5 or R4 to get 4.5 to 6V on the collector of Q2.


Because the transistor gain might not be 100 to 125 be prepared to adjust R3 to get a good sound over the range of the trimpot.


Option 3: Start from scratch.   Really this is only an option if you don't want to preserve the existing tone.


Option 4:  If you don't have a baseline tone in mind for the existing circuit then why build it!  Just find another circuit with two transistors which operates from 9V.

[antonis]

You can try this:

[iainpunk]

you could put 10 1N4148's in series, from the 9v to the circuit, but i can't guarantee that it won't oscillate due to the series resistance of the diodes...

cheers, Iain

[Halkbi]

Thanks for all the helpful feedback! I didn't know that it was so easy to set up a voltage regulator using diodes, but it does make a lot of sense. I assume mixing germanium and silicon diodes won't be an issue here? One 1N4001 in series with a high-reading 1N34A should be the ticket for something close enough to 1.5v. Hopefully I'll get time to put one together for evaluation later today. If it does indeed have enough juice at 1.5v I wont bother making a true 9v version. This also opens up a whole new world of low voltage fuzz builds 

I'll add a 250k pot wired as a series resistance before C3 (much like the gain control of a MKIII Tonebender) to see if it's any useful in taming the gain without sacrificing it's stock sound. Checking my inventory, I found that I do have some BC107As in the 120 range that should work out well here. Thinking about it, I also have some ge trannies in the 140+ range that I havent found a suitable application for. Wouldn't hurt to try them out here as it's not too hard to turn this into a positive ground affair

The schematic also lists a 47k volume pot. I'll be going for a 100k one along with an output cap blend for added versatility.

[antonis]

Apologize for misunderstanding but original query was how to form a 1.5V power supply for the original circuit..??

[Rob Strand]

1N4001 in series with a high-reading 1N34A should be the ticket for something close enough to 1.5v.

You could also use an LED, try a RED one first.


Forgot to mention full batteries are more than 1.5V.
"The effective zero-load voltage of a non discharged alkaline battery, however, varies from 1.50 to 1.65 V,"
https://en.wikipedia.org/wiki/Alkaline_battery

[iainpunk]

Apologize for misunderstanding but original query was how to form a 1.5V power supply for the original circuit..??

no, but a 1.5v power supply is a more elegant solution than re-designing the circuit!

you could also use a simple variable voltage regulator for changing the tone/timbre of the fuzz:

that 18v becomes your 9v, and the BD139 becomes a BC108 or other smal signal NPN, Hfe matters only a little here.
the pot changes the voltages, and thus the sound of the fuzz (down side it the changing output level.)

cheers, Iain

[Rob Strand]

you could also use a simple variable voltage regulator for changing the tone/timbre of the fuzz:

The shunt regulators will work out better.  They actually regulate (and remove ripple).

An adjustable shunt regulator is a VBE multiplier

[kaycee]

I've built several 3 and 1.5 volt fuzzers, I just use batteries to run them, they last ages and are usually plenty loud enough. I add a power switch so I can leave them plugged in on a board and not drain the battery. I don't see the point in stepping down 9 volts from a a PSU to run them....

[anotherjim]

I would prefer to run it at 9v for increased headroom and output

Folks, how does dropping it back to 1.5v achieve this aim stated in the OP?

[Rob Strand]

Folks, how does dropping it back to 1.5v achieve this aim stated in the OP?

Like a lot of these threads it's hard to know if the requirements are mandatory or desirable.

On the other end of the scale is a complete redesign which runs from 9V but sounds nothing like the original!

It boils down to where to set the dial between the two extremes and how much dev time the OP wants to put in.

[Halkbi]

Alright folks, I've gotten the circuit going with a voltage regulator and spent the day evaluating it and trying to mod it to the best of my abilities. First off I'll answer some questions though.

My interest in this circuit is mainly about me liking the odd stuff. I was surprised by the fact that despite being swedish, I hadn't heard of something like a 1.5v fuzz that was sold as kits in Scandinavia in the late 70's, and while doing some forgotten fuzz archeology I figured it might be a good opportunity to try and make it useful. For me and vintage effects that usually involves two things: decent output and headroom. I'm a very heavy handed player and many fuzzes fart out on me in a way that I don't particularly like and few vintage fuzzes seem to have the juice to push my amp without the need of too much gain. For the longest time I was using an original 1.5v Heathkit TA-28 but as much as I loved the sound it gave me (with it's tonestack bypassed to give it decent output) there were a couple of things that bugged me: it was very noisy, it's usable gain range was too narrow for me and I just couldn't deal with the fact that it didn't have an LED-indicator (which was also something that couldn't be implemented since it operates on 1.5v). So finding this circuit while just getting into building on tagboard seemed like a good place to learn a thing or two, hence the question in the original post. You've been really helpful so far, and the amount of stuff I've learnt by lurking around here is nothing but incredible. It showcases one of the better, these days rarely seen, sides of the internet.

Alright, enough of my rant. So what are my findings?

I did the voltage regulator using the schematic posted by Rob Strand. The first one using a 1k resistor in series with a 5k trimpot and three 1N4148s to ground. Also added a 47u cap over them for filtering. It gave an output of about 1.7v with the trimpot cranked, so I removed the third diode which left me with an output voltage of about 1.4v with the trimpot at zero. I figured that would be closer to a real life battery situation so I went for that one. I didn't notice any difference between the two in how the fuzz performed.

How did it sound then? Pretty cool! Heaps of gain and plenty loud as expected by some of you, but the 100k pot wired as a variable resistor between Q1 and Q2 did pull it back a little bit while subtly darkening the sound (again, much like the MKIII Tonebender). It was still far too much for me though and it exhibited unintentional gating when I was hammering riffs on the low strings, much the same way that the Heathkit does if I give it slightly too much gain. They also both behave similarly in that it would only be an issue when playing on the bridge pickup; neck was absolutely wonderful (if still a bit gainy). It did clean up with the volume knob so I thought increasing the input resistance would be my ticket to happiness. I had to increase it to a whopping 47k before it would stand up to my heavy garage psych pounding. It sounded pretty decent (albeit into my shitty home amp) and compared favorably to the Heathkit. I also tried wiring the pot between Q1 and Q2 as a voltage divider and it gave a real cool almost-Fuzzrite kind of vibe. Could have been an artifact of it messing with Q1s bias since I wired it before the coupling cap between the trannies. I will try and see what its like with another cap one added before it but I will probably make the gain pot's ground switchable for two different sounds. Oh and the 5k trim does add some gating/velcro which is a nice addition. I recommend putting it on a 5kC pot as most of the action happens towards the end of the turn.

My final mod was the output cap blend, and here's where I hit the wall. The limiter resistor needed to keep the bass side from getting louder than the treble side means that the once healthy amount of volume on tap now has such a hard time hitting unity that the first half of the gain range is unusable (and at 12 o clock, its still barely loud enough). Not surprised to end up here, but I'm here to learn I would want some kind of low end filtering on this. I believe I have two paths to choose from here:

First option is to reconfigure the circuit for more headroom and output. Could try out the suggestion of running Q2 on 9V and adjust it's bias accordingly. Might be a fun thing to try out as I've never heard anything quite like it but it will likely change the character of the fuzz (which could be for better or for worse).

Second option would be to add a gain recovery stage. Since I have 9v on tap, I could just implement something like the LPB-1 or the last stage of a Big Muff at the end of the circuit. An easy fix, but my worry is that it might not stay clean on high fuzz settings (which is still plenty loud).

What are your thoughts? I might try both, as it would be fun to see where those two paths may lead. Plus it would mean a good reason to practice my vero layout and soldering chops.

[Rob Strand]

These low voltage circuit are very touchy because it's hard to set the bias just right.    I think that's one reason that sound a bit gated.

Have you measured the collector voltage on Q1 and Q2?    From what I can see Q1 will pretty much sit close 0V for the entire range of the 5K trimpot.

I'm not 100% clear on where you are setting the 5k trimpot.  Is it near 5k or near short?

It's probably a good idea to experiment adjusting the collector resistors, especially R3.

The docs you posted come form the original April 1974 catalog.     The circuit drives a 47k level pot.   The specs were quoted as x100 gain (40dB) and 0.1mA current draw; the current draw doesn't look right.

I found these pages which look like the supplied build instructions for the kit.   You will note the transistors are BC170, which are lower gain that the April 1974 docs and might bias better; adjusting the collector resistors is a way of compensating for different transistor gains.  The trimpot value is 470k which looks wrong.   Notice they show a voltage divider to drop the voltage from 27V 270V down to 1.5V.





From a web-source,  this spec summary has 0.5mA current draw, which looks more realistic,


I tried to translate the instructions which explained how to adjust the trimpot R6 but from what I can see it doesn't actually say much.

Nonetheless,  your experimental approach has a lot going for it for these circuits.   You can try playing with the cap values to see if it pulls the tone close to what you want.  In particular reducing the input cap, or the cap between the two transistors.  This might help prevent the pedal farting out too much with strong bass signals.   A very worthwhile experiment.

[anotherjim]

Love the Josty transistor symbols. If I didn't know any electronics, I'd see a non-return butterfly valve!
And I now know my Indgangs from my Udgangs!

We actually have Swedish Chainsaw type designs on this site, although it's going to be little more complicated.
https://www.diystompboxes.com/smfforum/index.php?topic=122928.0

[iainpunk]

Love the Josty transistor symbols. If I didn't know any electronics, I'd see a non-return butterfly valve!
And I now know my Indgangs from my Udgangs!

We actually have Swedish Chainsaw type designs on this site, although it's going to be little more complicated.
https://www.diystompboxes.com/smfforum/index.php?topic=122928.0

i have a non-boxed 'mini sawmill' on my finished board pile, its a cool pedal, with heavy as ballzz distortion, but not a swedish chain saw, since it misses the grind-full 1kHz peak and fat 100Hz peak, its more of a hardcore punk sound than a death metal sound.
[shameless plug] if you are in to the chainsaw, i posted a similar sounding pedal the Embryo Pink. the crossover diodes are LED's in my pedal, but 4148's are probably a better idea for most guitars (my guitar has insane output volume)
https://www.diystompboxes.com/smfforum/index.php?topic=122851.msg1159848#msg1159848
biggest downside is current draw through the first gain stage, about 20mA. [/shameless plug]

cheers, Iain