Accidental Oscillation/Sub Octave?

[Take42]

Hey guys, first post here, but I have come here a lot for reference. Only just recently got into the game, so sorry if this is in the wrong forum.

I really only got going for real making pedals a few days ago, and since made a few fuzzes, and mounted one inside a bandmate's bass.
But tonight, being in the middle of the night when my creativity strikes, I decided to experiment a bit more again, this time putting a blend control on a bazz fuss.
A "mini blender" on the bazz fuss that is.
Link: http://seanm.ca/stomp/images/fet-minblend1.png
As I was breadboarding it up my last transistor broke, on the emitter side... I decided to attempt to use it anyway. Just kinda... Pressed it down against the resistor I needed it to touch.
(Don't judge, it's late and I am tired!)
Anyway, as I was wiring I replaced a lot of part values with others, including a few resistor values, used two 3904 transistors instead of what the schematics called for.
I also used one battery rather than two seperates.
Somehow I ended up with the fuzz not quite getting all the way into the signal, but just enough to hear it when struck hard. But the odd side effect... There is a sub octave/oscillation effect that somehow made its way in there and I have no idea where. I am thinking it's from the poor transistor to resistor connection but I don't know.

I am going to try and write out the schematic in a little bit and post it here.
Also a sound recording is probably gonna be requested...
https://soundcloud.com/brandonchutuk/weird-oscillation-sub-octave
(Sorry for the crap quality and lack of midnight playing skills.)

Anyone have any ideas what might have caused this?
It's the standard bazz fuss circuit without master volume, gain, or any other controls, and the mini blender circuit.
Sorry for the rambling quality of my post, and blaming it on the burning of the midnight oil. Long day.

Schematic coming...

Edit: I also must mention that the octave/oscillation effect only affects the lower strings.
Edit 2: Here's a makeshift schematic - http://i.imgur.com/RUe2dji.png
What I learned from this was I should not be allowed to write schematics at almost 3 am. I think I got it all on there.
Sorry it's not too readable. I couldn't quite learn the program I was using fast enough. I'm guessing it's some misbiasing issue? That seems to always answer things.

[amptramp]

From the schematic, it looks like Q1 is the fuzz stage and Q2 is an emitter follower that gives you clean signal.  The key to what you were saying is that the suboctave effect only seems to appear at low frequency inputs.  The fuzz stage looks like it is prone to gulp distortion or blocking distortion where the input capacitor is charged when the input goes positive and pulls current through the base of Q1.  It may charge to a sufficiently high voltage that when the voltage comes down, it goes negative.  The base-emitter junction is reverse biased, so it can't dump the charge.  The only place tha charge can go is R1 and tthe series combination of D1 + R2, but if the voltage goes below a diode drop, D1 is out of the picture and R1 is the only discharge path.  This path has a corner frequency of 7.2 Hz, but what is important is how much of the signal has to be bled off before the fuzz operates normally.

If you have lost the R1 connection, you may have trouble getting the voltage on C1 to come back up - it may stay low for a cycle, causing it to operate every other cycle like a 2-cycle engine that is mistuned so it runs only every other cycle.  This would give you a suboctave over a limited range of input voltage and frequency.  If you have lost the D1 connection, you may have the same thing but with less fuzz because bias depends on collector voltage which would be high and the bias has to go through a reverse-biased diode.

Good luck and good hunting!

[Take42]

Wow. I am still learning and understood maybe a quarter of those words. Thank you for the in depth answer. I'm looking back at the schematic a bit now.  

If I, say, wanted to expand the octave effect further up along the fret board, how would I go about that? With or without fuzz.

Edit:
I also would like assistance with a separate circuit I want to build - one that has a clean buffered blend and a fuzz with a volume.
I drew out this schematic just now and don't have any parts to test it.
I was wondering if it would work, what I am not getting, etc...
http://i.imgur.com/zAQxj4y.png
Also would I be able to run this off one 9v battery, or would I need two seperate ones? Can this even all be connected up this way?
(I didn't want to start a new topic specifically for this and spam up the joint.)
(Looking back I meant to leave out C5...)

Edit2:
I drew a secondary schematic with a passive splitter at the front too
http://i.imgur.com/kU9HX1o.png
Compared to the other... Well, same questions.

[amptramp]

These comments are based on http://i.imgur.com/zAQxj4y.png

1. The TL071 has to be biased near the middle of its range at the input, so you need a ressitor to ground at the non-inverting input as well as the one that pulls up to +9V.  You may also want to do a search for noiseless biasing where a voltage divider is attached to a capacitor to ground and the bias resistor is taken from that point.

2. The transistor bias is derived from a forward-biased diode, so the collector will be clamped at a diode drop above the base which is at a diode drop above ground.  There will be very little swing at the output because it cannot exceed 1.4 volts.

3. Your input impedance is going to be very low and it will vary with the setting of R1.  It is necessary to use a high impedance if you are connecting a guitar to the input.  It is better to use the slider of the pot as its output so for R1, the signal comes into one end of the pot, the other is grounded and the slider goes to C1.  R3 would be unnecessary.

4. Your op amp buffer is non-inverting whereas your fuzz stage is inverting.  You may get cancellation at some pot settings.

5. You don't need two separate batteries although some capacitance from the +9V to ground may be useful.

6. The output impedance is variable depending on the setting of R2.

[Take42]

These comments are based on http://i.imgur.com/zAQxj4y.png

1. The TL071 has to be biased near the middle of its range at the input, so you need a ressitor to ground at the non-inverting input as well as the one that pulls up to +9V.  You may also want to do a search for noiseless biasing where a voltage divider is attached to a capacitor to ground and the bias resistor is taken from that point.

2. The transistor bias is derived from a forward-biased diode, so the collector will be clamped at a diode drop above the base which is at a diode drop above ground.  There will be very little swing at the output because it cannot exceed 1.4 volts.

3. Your input impedance is going to be very low and it will vary with the setting of R1.  It is necessary to use a high impedance if you are connecting a guitar to the input.  It is better to use the slider of the pot as its output so for R1, the signal comes into one end of the pot, the other is grounded and the slider goes to C1.  R3 would be unnecessary.

4. Your op amp buffer is non-inverting whereas your fuzz stage is inverting.  You may get cancellation at some pot settings.

5. You don't need two separate batteries although some capacitance from the +9V to ground may be useful.

6. The output impedance is variable depending on the setting of R2.

Thank you, I am slowly getting the hang of this. Steep learning curve.
I'll work on it a bit more.