Cherrybomb doubts to be undone

[bordonbert]

You haven't reused electrolytics which you have had connected the wrong way around and just reversed them have you?  That would be a "bad thing".  Once used with the wrong polarity they are severely stressed and compromised.  At a few pence each they should be binned!

If it's in a case, is the case grounded at one point and one point only?

Check what the noise sounds like when you play with the tone controls.  Does the character of the noise change in tune with the tone control setting?  If so then the noise source is likely (not definitely) in front of them in the input stage.  Also how does the gain control affect the noise level and character?

[garfo]

All the capacitors are new and have been wired the right way. At first I had the circuit grounded at input only. I changed the 47uf cap to ground with a 100uf and have grounded the pedal on both sides, in and out. Now I have much less noise when playing, but If I turn the volume knob down, the floor noise rises. With the pedals tone controls I don't hear much noise. mostly when I mess with the treble. Noise increases when the Gain is set near it's final stage. And yeah, the noise has the same frequency character from the bass and treble frequencies when messing with those.
Nevermind, I found the noise source. It was my Computers charger.

[garfo]

Anyway, the pedal is ok. The only thing I can't get read of is the splat sound it comes when getting into overdrive territory.It's just annoying. I've tried an original and it behaves differently from the two clones I've made. His was fatter sounding, less Gain, way less gain, and no splatyness.
The only differences I saw was an extra transistor as I mentioned before and instead of the bc184 or bc469, it had bc182L on Q1 and Q3 and bc184L on Q2. Q4 appeared to be some sort of Fet.

[bordonbert]

A warning.  Different transistors will make no difference to this circuit, (they rarely do), don't fixate on that!

The circuit uses very basic standard elements to perform each task.  The first two transistor stage has been around since Adam and the final stage is a classic single transistor amplifier.  They both employ local negative feedback, (it's a GOOD thing), within them.  One of the jobs negative feedback is used to do is to make the circuit immune to differences in the gain of the transistors.  For all modern transistor types like the ones you  listed it works.  'Nuff said?

[garfo]

I understand. But let me tell you, I tried different transistors, and the ones with higher hífen also brought more volume to the effect. Appart from that, I can't understand How is it that with my build I pickup so much noise interference and with the original pedal that didn't happen.
Ok, i did day before that once I unplugged my computers transformer the noise went away, but with the original overdriver, even with the PC transformer plugged in I didn't experience those noise levels.
So, my final question is: Is there anything I could do to filter some of the noise that gets picked up by the circuit? I've changed the 47uf CAP with a 100uf CAP, but it wasn't enough.

[Cortexturizer]

I have no noise with this, no spatter, no nothing, just a juicy sound. I used some shitty maybe even fake BC109Cs. Everybody is experiencing fizzy decay and shit, not me. I made three attempts on this before making it right, all of those on vero before having it done on madbean pcb. The versions on vero were just not right for me, sounded bad and sometimes the bass control wouldn't work (it's subtle anyway though) and all sorts of problems. But now it sounds glorious. I posted my demo somewhere on the forum just recently. Here it is again - https://soundcloud.com/kuato-design-stompboxes/colorsound-overdriver-demo
First the guitar volume is on half, then on 75% and then on full. The gain control on the pedal is on max all the time. You can hear how little noise it has, and it was on my testing rig at the time, not even boxed.

It is very unusual that a lot of people are struggling with this build, myself included. I came on top but I thought of quitting so many times...

[bordonbert]

    

That's the danger of fixating on things that you just "know" or that "my tech mate" has told me MUST be the problem.  It stops you looking at the issue logically.  This type of circuitry is so well established, understood  and documented in electronic terms it's a yawn.  It's so predictable yet people want to elevate it into a magic black art.  And there are those who have a vested interest in doing so.

In the country of the blind the one eyed man is king.

Cortexturizer has got it right in his post when he says:

It is very unusual that a lot of people are struggling with this build, myself included.

and it's telling when he points out:

I made three attempts on this before making it right, all of those on vero before having it done on madbean pcb.

That's not unusual.  So does building it on Veroboard cause strange effects in the circuit action?  Of course not!  The only problem you should be likely to get with a Vero build in an audio circuit is instability - unless you have used an incorrect component, fitted one incorrectly, or your layout is wrong.  And none of those are the fault of the Vero or the circuit.  Vero itself cannot produce a fault like "sometimes the bass control wouldn't work".  Us putting components into the Vero has produced the problem.  And then he got it right and lo and behold the circuit is fine.

You mention that the circuit is oddly noisy.  I would suggest you either have a grounding problem or you have instability due to the layout.  This happens to all of us and more easily with Veroboard.  I have just built a 3 stage JFET SRPP type overdrive circuit of my own and suffered from this to start with.  It was my own PCB design too and it had to be made with off-board pots, 5 of them, and the flying leads to these were a problem.  It oscillated.  There were slight squeaks and high noise at high gain settings.  Fortunately I have a scope to see this, I would assume you don't have that luxury.  Once you acknowledge and accept the problem you can come up with a way to solve it.  Just poking the circuit without having any idea of why you are doing what you are doing is like trying to fix your car backfiring by starting at the starter motor and......

It's not the problem that's the problem it's the way you see the problem that's the problem.

My point is that, if you have tried different transistors and they increased the volume that does not show that higher gain transistors are the solution.  It shows that, knowing what we know about that circuit design and knowing that others have built it with any old transistors and it still worked perfectly, then either your original transistors were ridiculously bad quality out of spec possibly damaged, or far more likely there is something wrong elsewhere in your build to allow them to affect the volume.  As I said:

For all modern transistor types like the ones you  listed it works.

and I meant:

it - just - works!

Half decent transistors cannot make that difference in an otherwise working model of this circuit.  It is basically immune to transistor gains within any normally accepted band, it is designed to be so.  Over the years I've designed dozens of circuits with that exact same input gain block without any of them having an issue of the type you have.  This one is bog standard.  As a tool it's on a par with a spanner in a workshop.

Please tell us:

What transistor types are you using?

Are you building with Veroboard?  If so can you post a decent clear closeup picture of both sides of your board?

If you can we can have a look and try to spot any other issues.  My gut and experience tell me there must be at least one.

[Cortexturizer]

))

I did hear actually that vero-style adds some parasitic capacitance to the game, but nothing major.
All of the other stuff you said is spot on of course hehe.

[bordonbert]

It does that, hence the instability problems and my inclusion of "in an audio circuit", i.e. not at higher frequencies.   

[bordonbert]

This is from Wikipedia's entry on the "Common Emitter" transistor amplifier which is our output gain stage.  It also basically applies to the input pair topology circuit we have which is derived from that.

One common way of alleviating these issues is with the use of negative feedback, which is usually implemented with emitter degeneration. Emitter degeneration refers to the addition of a small resistor (or any impedance) between the emitter and the common signal source (e.g., the ground reference or a power supply rail). This impedance Re reduces the overall transconductance Gm = gm of the circuit by a factor of gm(Re+1) which makes the voltage gain

    Av = vout/vin = -gm RC/gm*(Re+1) = -RC/Re  (where gm*Re >> 1 and RC is the collector load resistor).

So the voltage gain depends almost exclusively on the ratio of the resistors RC/RE rather than the transistor's intrinsic and unpredictable characteristics. The distortion and stability characteristics of the circuit are thus improved at the expense of a reduction in gain.

That's a lot of maths to a non technical guy but the gist is that this circuit is designed to remove transistor gain from its workings.  It is set up so that the gain that is left after the feedback is applied is under our control with the resistors in the circuit, and is exactly what we need to give the effect we want.  It must be true, Wikipedia says so! 

[garfo]

I didn't use vero, I've etched my own board.It is well etched and everything is right in place. Transistors are Q1 and Q2 BC547B and Q3 2N3904. Voltages are all within spec. (Don't have a scope). If I run the pedal at clean settings it is fine, the pedal works as it should. It's only on max gain that I get some noise that is picked up from my television and computer and specially when I role the guitar volume down to zero that I get some hum.
I've added a 100ohm resistor followed by 220uf(instead of 47uf) cap to the 9+, and also added a 330ohm resistor in series with the signal that goes to the board and then a 220pf cap from there to ground. That has significantly bring down the noise.
The splatiness only happens when I go to slight overdriven territory. If I max the Gain knob, it doesn't happen, it's just heavy fuzz/distortion and it sounds good actually.
My only issue is that I've A/B'ed it with an original(reissue) and they behave different. The original sounds way less aggressive and there's no splatiness on any part of the gain setting.

[garfo]

By the way, I was comparing schematics and found something different.
On this schematic, which appears to be the one used for the cherrybomb, the 12k resistor is connected to the negative side of C5, but on other schematics as the one attached it is connected to the positive side. Is this on purpose? only the 9volts version shows it facing the negative side.

That's a strong claim in need of proof.

You're quite right there Kothoma, but I would maintain the proof is in the simple circuit action.  All of this circuitry has been around in exactly this form for 40 years!  You are doubtless right that the Baxendall tone circuit is used by named guitar electronics manufacturers like Ampeg, that is domain knowledge which I do not possess, but in its passive mode as they usually use it it is ground referenced.  The difference is that here it utilises feedback from the collector of the third transistor.  It is not in any way in passive mode, it is an active Baxendall control setup exactly as it is used in hifi.

Check out here:  http://makearadio.com/tech/tone.htm  They confirm,

Independent adjustment of bass and treble frequencies in high fidelity  audio amplifiers is usually accomplished utilizing specially designed tone-control networks. There are versions for these tone controls based only on passive components, such as the ground-referenced  (my bold underlining) James network shown in fig. 1. Among those versions using active devices we must make mention of P.J. Baxandall´s proposal, in which the tone control was devised as a feedback amplifier. 

This is irrelevant to my claims about the DC conditions for the capacitor polarity issue however.  That is without question.  If you get a clear enough version of the circuit it even shows DC voltages posted on it.  Here is a copy for reference:



Anyone with experience of designing with transistors, (yes I go back that far  ) will be able to see what the DC voltages would be.  They aren't even close to being questionable.

The emitter of Q1 is low, it has to be. Notice Q1 DC reference is the emitter of Q2 and its emitter has to be about 0.6V lower than that.  Q2 emitter must be significantly lower than its own collector to allow it headroom to work and prevent it from bottoming out on signal swings.  So Q2's output capacitor DEFINITELY should have its +ve to C2 collector just as the schematic shows.

Q3 is set up similarly.  It's base DC point is defined by the two resistors at around 1.63V.  It's collector will then be around 5.1V.  This capacitor also has to have its +ve terminal to the collectors.

I'm sure someone else with design experience will come in to either confirm this or tell me I'm a dipstick.  (That may definitely be, but I don't think so in this case  ).

[Bret608]

Where did you find that one on the bottom? It reminds me of Electric Warrior's work (he's usually at DIYSB or FSB). I think Madbean probably used the version available at Fuzz Central, although the Cherrybomb part values look pretty close to the Overdriver version on the bottom, except that the positive side of the electro just before Q3 is facing the collector as we are now pretty sure it should be.

It's interesting that none of the versions in this helpful schematic you've posted match the original factory version exactly. Such is the cross we DIY-ers must bear! 

[bordonbert]

Odd looking versions! I don't like the look of the first 2 in the sense that the feedback via the 12k resistor from Q2 collector to Q1 emitter is taken from inside the output cap.  Look at the original schematic and you will see that it is connected outside so that it's setting a DC reference level in the tone area to a low value which means the coupling caps MUST be that way around. (I'm afraid it's not even "pretty sure", it's absolutely sure!) With the 12k connected straight to the collector there is nothing defining a DC level within the tone control circuitry.  It will float to anywhere the leakages of the caps dictate and the idea of which way around the caps should be connected is a shambles, it could be either way.  This is not a "design eccentricity" or a "novel approach", it is an "error".  Either the designer of the circuit made a mistake or the drafter of the schematic did but someone has goofed.

It's there to define the AC gain but it's also affecting the DC conditions within the first stage as well as any signal. That does not look wholesome at first glance. I'd strongly suggest that may be a typo in both. Notice in the third version it is outside the cap again.