Companion Fuzz with contour & blend controls

[duck_arse]

dth - what type number transistors are you using?

this page is getting pretty long now - please! don't make me scroll all the way past those pics - repost the Q1 volts each time, if you would be so kind. thank you. also, what voltages do you have at - R2/R3 junction, R3/R5 junction?

isolate the Q1 and Q2 circuit from all the other junk. lift, say, C9, C10, C8 and R13, then listen/monitor [LOUD] at the R7/C7 junction. does it oscillate?

also - lift C11 at the R14/R15 end, and inject signal, monitor at the level wiper. any sign of oscillations? [monitor at R12 - is that section oscillating?]

[Mark Hammer]

If I haven't already beaten this corpse to a bloody pulp with a stick, this thread illustrates why I far prefer perfboard to stripboard.  It often lets you place components in a manner that conforms more closely to the drawing, which in turn makes checking things and troubleshooting easier.  At least it does for me.

I've perfed FY-2 circuits into a board the size of a dollar coin, using 1/8w resistors and little ceramic disc caps.  I could tell exactly where any issues that cropped up were occurring, because the board flowed exactly like the drawing.

Not bragging, or criticizing.  I just think that perf and its cousin, pad-per-hole, are sorely overlooked, amidst the many sites showing vero layouts, and software aids that both support and encourage use of vero.

In any event, enough complaining.  Back to work.

[dthurstan]

Thanks Mark, I've also used vero. I actually want to try PCB at some point as it makes sense for more complex circuits. I also though perf/vero where interchangable terms & there was just pad-per-hole as the alternative.

Thanks duck_arse, sorry for the delay getting back to you. I'll post the voltages again.
Q1
c 4.00
b 0.65
e 0

Q2 (either side of the balance pot)
c 1.04
b 0.6
e 0

R2/R3 9.0
R3/R5 3.6

I'll isolate the sections and get back to you. I'll also see if isolating it changes the Q2 collector.

Dave

[dthurstan]

Hi duck_arse

I've isolated the fuzz section now. I lifted C9 (still connected to C1) and lifted R13. I probed the output from R7/C7 junction. Works fine no oscillation. Contour pot still connected that works as it should. No changes in voltages with movement of Fuzz or contour pots.
Voltages are:
Q1
c 4.00
b 0.64
e 0

Q2
c 1.06
b 0.6
e 0

Isolated Q4 gain stage and injected signal into it. No oscillation there.

So I guess the blend/balance pot is creating a feedback loop which the fuzz section doesn't like?
I know I keep saying this but it did work fine on the breadboard, I'm wondering if a higher value pot might help, say 25k?

Thanks Dave

[dthurstan]

I've been thinking about how the balance control operates. It changes the attenuation of either signal. So in the centre position both sides are attenuated by 9.5 dB. At either side of the control one signal is shorted to ground the other the attenuation reduces to 6.1 dB.
Therefore, increasing the pot value will decrease the attenuation, 5 dB centre, 3 dB either side. So maybe its the resistor values (R12 & R13) I need to change. I'll try increasing R12 instead? Does anyone have any thoughts on this.

Thanks
Dave

[duck_arse]

your Q1 and 2 voltages are now fine. DON'T CHANGE THEM! and you have proved things to be not the problem, which is allays good. the balance pot, to one side, connects that side to ground. see the wiper connection? nothing gets past that. I'm no good on those type balance/mixers, but there is threads and threads on them hereabouts.

[dthurstan]

Thanks for all your help duck_arse. And everyone else too.

I've been researching the blend/balance/crossfade/mixer section, to try and work out the best approach to deal with the oscillation which I think is due to bleed through. R12-R15 together with the balance pot form two T pad attenuator or a H pad attenuator.
At the moment I've removed dry gain stage (Q3,R8-R11 & C10) and increased R10 to 22k. This does tame the oscillation somewhat but obviously at the cost of reducing the dry signal.

I'm beginning to think this is an inherent part of trying to mix a loud fuzzy signal with a dry signal in this way. It was just tamed to such a degree I didn't notice it when I breadboarded it previously.

[antonis]

Try to modify Q3 into Emitter follower..

[dthurstan]

Try to modify Q3 into Emitter follower..

Yep that worked. Thanks Antonis.
I'll see if I can post a video, showing the pot operations as it's not perfect but I think you can get some useable sounds out of it.

I tried playing a bit earlier but my son burst in saying "stop playing guitar. I don't like the noise, put it away."

I'm not allowed to do anything 

[dthurstan]

Hello

Here is the finished article;



And a schematic showing the changes I made.


A quick video of it working;

Not sure about the audio on this think my phone is doing some noise cancelling so it is maybe picking up the physical guitar (as I'm sat closer to the phone than the amp is).
The mixing isn't perfect but I wasn't excepting it to be, more to try and bring in a bit of dry signal so things don't get lost in a mix/band. I was thinking this could be used on bass as radiohead used a Shin-ei on one or two songs.

[Mark Hammer]

Not quite there yet, but well on its way.

The controls on the fuzz side seem to be working as intended, although I'm puzzled by the manner in which the fuzz seems to "retreat" at times.  I don't know if that's an artifact of how you recorded things, or something inherent to your modifications or build.

That aside, I think you can see why I backed away from the initial recommendation of a 50k contour pot.  Clearly, it doesn't take very much resistance between the 100nf cap and ground to eliminate any audible scoop.

[duck_arse]

on your last circuit dia, C10 seems to be listening to the +9V line. not much to hear there, you would hope.

[dthurstan]

Not quite there yet, but well on its way.

The controls on the fuzz side seem to be working as intended, although I'm puzzled by the manner in which the fuzz seems to "retreat" at times.  I don't know if that's an artifact of how you recorded things, or something inherent to your modifications or build.

That aside, I think you can see why I backed away from the initial recommendation of a 50k contour pot.  Clearly, it doesn't take very much resistance between the 100nf cap and ground to eliminate any audible scoop.

That fuzz retreating is the recording not the pedal. Sorry, not very helpful video I'll see if I can do it again.

[dthurstan]

on your last circuit dia, C10 seems to be listening to the +9V line. not much to hear there, you would hope.

Yep very true, it should be coming off the emitter & c10 is 10uF.

[antonis]

C10 could be 10 times smaller (1μF) with no audible effect but it definately should come out of Emitter..!!

[dthurstan]

Sorry, it's been awhile.

Here is another video, not sure if this helps I tried playing quieter and away from the camera so it would focus on the amp. Phone doesn't give the best audio, not really set up to record anything better with a dedicated mic then edit audio and video etc.

[antonis]

Your videos help..
Your schematics don't.. 

[dthurstan]

Your videos help..
Your schematics don't.. 

Refresh, the schematic should be correct now and smaller. 

[dthurstan]

Your videos help..
Your schematics don't.. 

Refresh, the schematic should be correct now and smaller. 

That didn't work. Here


I'm still cool though. 

[antonis]

Just curious..

What's the gain of Q3..??
(we are still talking about 2N3904, aren't we..??)