Help me out with an op amp fuzz circuit i made [Schematic/Soundclip included]

[Se7en_Costanza]

Hey guys,
I recently made a fuzz pedal with a TL072 op amp, not following any schematic just playing around, and it sounds great, very punchy and almost big muff like.
The only problem im facing with it is that its very gated and crackly when the note sustains long enough, or if you turn down the guitar knobs volume and play, it sounds horrendous, super gated and crackly. Ive heard its something to do with not biasing the op amp correctly?
So i was hoping someone could give me a hand as im still a bit of a newbie, and also any other advice to improve this circuit would  be much appreciated! especially if there is something missing that is crucial.
(Ive left out having a tone filter section as i didnt want to filter the circuit so much as i like the raw sound of the op amp's gain.)

Here is a schematic/layout i draw of the circuit:
https://imgur.com/gallery/I7w0G

Here is a soundclip of the sustain of the pedal to demonstrate the sound i want to try and get rid of:
https://clyp.it/ovcd1gff

[thermionix]

4.5v?

[Se7en_Costanza]

For the positive supply voltage pin on the op amp you mean?

[thermionix]

Sorry, I was being intentionally vague because I'm a noob at this stuff and I really don't know what I'm talking about.  But don't you need Vref, 4.5v, bias, whatever?

[Se7en_Costanza]

Ah right yeah im not sure either, still a noob also, i know they use that on tubescreamers and other op amp effects but wasnt sure how to add it to my circuit. i tried supplying the op amps pin 8 plus voltage to 4.5 but it made it a lot worse. not sure if im meant to add the 4.5v's somewhere else in the circuit..

[stonerbox]

Could you measure the pins? We need to re-bias that thing to stop the gating and maybe (not very sure about this) add a buffer at the input. That is IF you loose any amount of fuzz after the bias is set correctly. Hopefully not.

Also tying the unused amp to ground is not a good thing. The none inverted input (3) needs to see 1/2 VCC (4,5v) and the inverted input (4) goes to the output (7). Check the second (a) image.

http://www.electronicproducts.com/Analog_Mixed_Signal_ICs/Amplifiers/Properly_terminating_an_unused_op_amp.aspx

[duck_arse]

my advice - draw the circuit. what you have is a layout diagram, which makes the underlying operation and overlying mistakes much more difficult to spot. draw the electrical connections instead of the mechanical.

you have no feedback resistor on the first stage, you need that. you have no bias on either stage, you'll need that and something to generate it, then some capacitor at the input to isolate the signal source from the DC.

I'm not sure any of the advice at imgur was much use, frankly.

[TejfolvonDanone]

ummm....
1) None of the op-amps ha got feedback so they are acting like comparators: when the voltage on the positive terminal is higher than on the negative the output goes to (almost) 9V and when not the output goes to near 0V. (the actual voltages are around 8V and 1V)
2) Giving a not rail-to-rail input op-amp a rail input (GND on the + pin of the first op-amp) won't give the result you'd like. Also you drive the op-amp's input lower than the negative rail which is out of the absolute maximum ratings of the op-amp and can cause damage to the IC.
3) Because they are JFET input op-amps and the gain pot configured as a variable resistor the left part of it has no use.
4) The right part of the gain pot and the volume pot is somewhat redundant.
5) The input is DC coupled you need a cap to make it AC coupled and not let any DC on the input influence the whole circuit.
6) The first resistor just damps the input increasing the gating.

Can you tell us what are you trying to achieve?

[GibsonGM]

I'd suggest you learn to draw it out as a schematic, as Duck said.  This will help you FOREVER, once you can do it you will always have that skill.
Look at the Tube Screamer for some ideas!  http://www.geofex.com/article_folders/tstech/tsxtech.htm

Next (or maybe first), you'd probably like to learn a bit about how opamps work.  There are millions of tutorials online, here is a start: https://www.youtube.com/watch?v=K03Rom3Cs28    Watch Dave at EEV blog on this subject, too! His vid will be to the right of that one, in suggested videos...

You do NOT need to know all of the physics up-front...you need to know inverting, non-inverting, and how their gains are figured out.....and probably summing.  Comparator action comes later.   Really you wanna know how to set the chip up for inverting or non-inverting amplification, and where/how to place those 2 or 3 resistors that do this...this is what Tef is getting at.   

When you see "Vref" or "Vr", or "ref" or "bias", "1/2VCC", "4.5v"..., this is 'applying bias to the circuit'.    Since a sine wave (guitar signal is a complex one...) goes from a positive to a negative voltage (it crosses zero...) - and our power supply is typically just a positive voltage like 9v or 18v....we must insert an OFFSET to raise the signal up enough so that the lower portion can be worked with, too.     AC signals can have a voltage applied to them, and then REMOVED after we work on them!  This is what those input and output capacitorss do - they block the bias voltage we are using to manipulate the AC signal (among other things).   So you will need them after you sort out bias. 

Look at the Tube Screamer, my friend..it's all there for you.... 

[thermionix]

When you see "Vref" or "Vr", or "ref" or "bias", "1/2VCC", "4.5v"..., this is 'applying bias to the circuit'.    Since a sine wave (guitar signal is a complex one...) goes from a positive to a negative voltage (it crosses zero...) - and our power supply is typically just a positive voltage like 9v or 18v....we must insert an OFFSET to raise the signal up enough so that the lower portion can be worked with, too.     AC signals can have a voltage applied to them, and then REMOVED after we work on them!  This is what those input and output capacitorss do - they block the bias voltage we are using to manipulate the AC signal (among other things).   So you will need them after you sort out bias.

This is what I was trying to say, I don't have all the terminology down and it sounded idiotic when I started trying to explain myself.  But yeah maybe the gatey sound is partly because the signal can't swing below 0v, it's bottoming out without a bias voltage to ride on (see, doesn't that sound idiotic?).  Obviously there's more to it than that, as the more informed folks above have mentioned.

[GibsonGM]

Yeah, it's tough when you're starting out.  Learning to read/write schematics makes it a LOT easier.  Then we could tell him where to apply bias, which would be easy enough to do.  Gotta add 2 caps and 2 resistor (plus the bias network/cap)...it's hard when you're not really sure how these things work, or why they're set up this way! 

Bottoming out doesn't sound idiotic; technically it's being clipped on the negative excursion of the wave, but same thing.  Nowhere to go, it ceases to be part of the signal, resulting in distortion and 'missing data'.     We can add a DC voltage to an AC signal, and then take it back again (!) after, so we elevate the signal to where we can work with it.   By the time it gets thru the whole circuit (or the portion of it we want bias in), we block that DC with a cap and you have your AC, crossing zero, once again.     

Another concept to grab, like Ohm's Law!   

[Se7en_Costanza]

Ill draw this layout again in schematic form, to make it a bit easier, im just not the best with drawing schematics so i just did a visual diagram of what the circuit is. I'll give the pins a measure and post back, i do that with the negative lead of my multimeter on ground and the positive lead on each pin correct? still a noob at that also haha.

What im actually trying to achieve with this pedal is trying to make my own circuit really, as i did this by accident and it sounded awesome im just trying to get it to be overdrive/fuzz pedal, so i know its not there at the moment but when the gain is maxed at the moment it sounds very fuzzy which i like, id like it to be less fuzzy though and more dirt-overdrivey only when the gain pot is down lower. Kind of my inspiration is the DBA intersteller overdriver, although with an op amp and not with transistors, although im happy to use some transistors throughout my circuit if need be. But Im not trying to be exactly like that pedal but its just a thought i had when making this circuit how drastic the gain changes in that pedal from overdrive to fuzz.

Obviously i have much to do to it so the pedal is cleaner when the gain pot is down, probably have to change a lot to it as when i just turn the gain pot down at the moment it doesnt do much just has the same fuzz sound but at a lower volume as i know ive set the whole thing up wrong anyway haha. If someone could give me some advice on how to do that, ill most likely have to completely change the whole set up of the pots haha.
Also i originally had a 100nf cap at the start of the circuit, it wasnt polarized as when i used electrolytic it made the cutoff of the notes even more gated and a lot more bassy then it was (probably just a bias issue though), it was also more bassier with the 100nf non-polarized cap also so i got rid of it, although its probably what my pedal needs haha.

Anyways ill redraw the schematic in actual schematic form and repost it, as well as measure my pins and get back to you..

[Se7en_Costanza]

Redrawn Schematic in proper format, sorry if it looks a bit wonky, not used to drawing them
https://imgur.com/a/HUboz

[PRR]

Beauty.

Putting image in-thread for quick reference. Also put part-values next to parts.



The opamps are powered with zero and +9V. The guts inside the opamp can not swing this far, say +2V to +7V. You have inputs at zero. The guts of the opamp input are out-of-bounds. You are not supposed to do that. What you get is uncertain.

From experience, what you get is that big signals do push the inputs to a happy-place, and because there is no limit on the gain, the output is slammed. But when the input signal fades-off, the inputs cut-off again, "gated".

thermionix knows (almost) what the "right" answer is. Cap-couple the input and use a large resistor to pull the input into the middle of its happy-zone, around 4.5V out of zero-9V supply. But if you do that, it is likely the thing will just make horrible noises, amplifying its own hiss and sneakage to the max, except when an input signal overwhelms its internal garbage.

You could try cap-coupling into R1, and returning R2 to a trimmer. A few tenths Volt above zero, or a Volt+/- below +9V, may take it to the edge of "working", so that much smaller signals will kick it into life, while very-small signals don't and the self-racket is muted. But it may be insanely fussy.

[Se7en_Costanza]

No worries man, still new to forum haha.

Okay no worries so i add a cap-couple before R1, what value would your recommend to try? also does a 'cap-couple' just mean 1 polarised cap like an electrolytic so it stops voltage going back into the input of the guitar?

Also with the 4.5v where should i be reading that? on a certain pin of the op amp i should do that by dividing the 9v's supply or what? im still very vague to what that means haha.

[FiveseveN]

GibsonGM's first reply covered the steps you're missing:
1. Learn how op amps work.
2. Observe how they are used in other circuits. http://www.electrosmash.com/ Also has some in-depth circuit descriptions.
Otherwise you're just painting by numbers on a picture without lines. Or numbers.

[TejfolvonDanone]

this is what Tef is getting at.

Yeah, sort of... 

[GibsonGM]

Second pic down on left "clipping amp", shows what you need:  http://www.geofex.com/article_folders/tstech/tsxtech.htm

"From input buffer" shows an INPUT CAP next to it.  Value is chosen to taste, based on how it works with following resistances and impedances (more reading, for now say .1uF to 10uF would be fine, other values too, will work but may not sound the way you like).

The 10k resistor with "4.5v" on it is coming from the voltage divider bias network, and is applying bias to the opamp input.  The resistor shown is NOT part of the bias network, it is in addition to it.  The input cap (and COUPLING CAP which will be over by "to tone and volume controls") block this DC voltage from following stages. 

Get the first stage to work, then worry about the second one!   You can "plug in" and listen after 1st stage by adding a COUPLING CAP at the opamp output.  Same values as input (again subject to change based on what you want to hear WRT bass/treble...).

The complete bias network is shown on the left side of the final pic at bottom of page, where you see "9V" and "4.5V" with the 2 resistors and 2 47u caps, plus 9V battery.   I'd actually suggest you BUILD the 'son of screamer' pictured there just to get up and running, so you can play with part values!

[bluebunny]

1. Learn how op amps work.

Have a read of this for a bit of theory, and then check out this video tutorial (it's brilliant).

[samhay]

As an alternative suggestion, you could just replace the TL072 with an op-amp that will work down to the negative rail.
An LM358 would fit the bill and should be easy enough to find.