Univox Sq 150 vero build, noise and oscillations problems

[Widows]

Hi folks, I've recently built a Univox SQ 150 as per the layout on GuitarFX Layouts:
http://tagboardeffects.blogspot.co.uk/2013/01/univox-sq-150-square-wave.html?showComment=1358733693905#c2992939746505328521

Now it works great when I'm playing, but when I stop there's an incredible amount of signal noise and oscillation. I know a certain amount of noise is going to be inherent in a circuit like this, but this is a bit over the top.

I've subbed a couple of parts, for example Q1 is a 5088, and some of the more odd resistor values were made up by connecting 2 different resistors to add up to as close as I could get to the required value. Power is supplied to the circuit from a Voodoo Labs Pedal Power 2+, at present I haven't boxed it up so the vero is connected to my breakout box/home made beavis board. I know that boxing it will shield the circuit to a point but I'd still like to be able to reduce the amount of noise of possible. 

What can I do to sort this out?

I can provide some clips if needed, they won't sound amazing but you'll get the idea about the noise.

Cheers

[PRR]

A high gain high-impedance amplifer, out in the open in a room, WILL pick-up and amplify all the buzz and hash in the room.

It doesn't take much to shield it. You can put half a small cardboard box over the circuit, wrap aluminim foil over that, and ground the foil with a clip-lead. That "should" help a lot. It is awkward, and some risk of foil slipping and shorting something.

Go ahead and box-it-up in a large old metal box with cover. Even a cookie-tin (don't let the PCB short on the tin). That "should" be about as quiet as it gets. If it passes smoke-test and ear-test, not a lot of work to move the guts to a nice right-size box.

[Widows]

Ok, I'll give it a go. THanks PRR. A friend of mine was building a fuzz face type pedal a while back and had the exact same noise coming out of it when not playing and they found it to be a biasing issue. That said, I've no idea what to do with regard to that so I'll box it up like you say and see what difference it makes.

[Mark Hammer]

The SQ 150 is a Fuzz Face, albeit with diode clipping, and some other component changes.  BUt the basic topography is a FF.  Stock, it is set for max distortion.

To set it for less distortion (which = less radio reception and similar noises), you will want to insert a resistance, preferably variable, between Q2 and the 33uf cap to ground.  A 5k pot would do just nicely.

I pimped out my own unit, which you can see/hear here: http://www.youtube.com/watch?v=mYGjn7Q-F_E
The sound quality is poor, but it gives some idea of the range of tones possible with just a few tweaks.

[Widows]

Cool, I'll give that a bash as soon as I finish my bass Rat off this week :-)

Nice vid by the way, informative. What did you use for the silicon diodes on your switch there? I tried 914s and it was 'orrible!

[Mark Hammer]

They were either 914s or 4148s.  Sometimes, its not the diodes, but rather what you feed them, and how you filter the outcome.

[Widows]

Ok, well Im pretty happy with the sound of the Ge diodes for now to be honest, the other diodes were just for experimentation really although I may put some on a switch at a later date.

You mentioned putting a variable resistance between Q2 and the 33uF cap to ground: So would I tie lug 2 to either lugs 1 or 3 on the pot, and then connect lugs 1 and 3 so the pot is in parallel with the 33uF cap, so the lug connected to the +ve terminal of the cap also being connected to the Signal leg of Q2?

..and thanks again Mark ;-)

[Mark Hammer]

Take a look at the AMZ MosFet booster here: http://www.muzique.com/schem/mosfet.htm

The values are obviously different, but that's essentially what you're going to do: one fixed resistor to ground and a variable resistance in series with the cap.  Note that there is NO difference between situating your variable resistance between the cap and ground or between the transistor and the cap.

[duck_arse]

I've built 2 sq150's, both with boss style buffering and bypassing. the people buying them where interested in the sound because it was so awful.

mark, what fet did you use in yours, and did you do any bias fiddles? I was told my v2 sounded completely dif to my v1, but I don't know how completely. I suppose that the fet chosen would be as critical as in a ff without a bias pot.

[Mark Hammer]

I used a 2N5457, as per the original, and have no idea what I used for Q1.  I initially made one stock, but was rather unsatisfied with it.  It was the custom of the time (late 70's) and the limitations of the packaging/chassis used by Univox, that there could only be two controls.  I gather, to be different that kept the unit at max distortion, and opted to use one control for tone instead of gain.  As you are likely aware, the tone control basically rolls off low end, by providing a bypass cap for the mids and treble to pass unaffected, and a variable parallel resistance for the low end to pass as well.

I suspect there are likely the same sorts of unit-to-unit differences in sound stemming from component tolerances that are also the case in Fuzz Faces.  And, like the FF, there is no "pure" one, merely those a given person likes more and less. 

I'm content with that.  I suppose if a person has no idea how to build or mod, they get obsessed with a "best" version of some pedal.  If you know how to build or mod, then a different tone is but a few solder joints away; no obsession required.  If you're selling pedals, then you need to identify what caused the sound people liked most, so you can replicate it consistently.  BUt if you're only making things for yourself, who cares, right?

[duck_arse]

I can't play guitar, so I never make anything for myself. and no bugger will tell me what they like. I test a box and think "it sounds like fuzz" every time. nearly.

original circuit said 2sk34c. as if. I used bc549c/j201 for the first, and just grabbed a j201 for the second. that looks like too much gain in both spots now.

the boss buffering and bypassing was a fun(?) exercise in creating a 10 transistor monster from a 2 transistor circuit. I think it might be time to have another, low gain go at it, without all the junk.

[Widows]

one fixed resistor to ground and a variable resistance in series with the cap.  Note that there is NO difference between situating your variable resistance between the cap and ground or between the transistor and the cap.

Ok, so if I were to add it in to the circuit using the layout in the link from my original post, could I connect (for example) lug 1 to the rail with the negative terminal of the 33uF cap, and lug 3 to the ground rail?
Or alternatively connect lug 1 to the rail connected to the signal leg of Q2, and lug 3 to ground? having lug 2 connected to one of the other lugs in either example

Thanks

[duck_arse]

lug 1 to 33uF - terminal, lug 2 to lug 3, lug 3 to ground.

or lug 1 to transistor/3k9, lug 2 to lug 3, lug 3 to 33uF + term, 33uF - term to ground.

I hate calling them lugs. I always use ccw, c and cw for counterclockwise, centre and clockwise. then you never (!) get them backwards.

[Widows]

Legend. Cheers dude!

[duck_arse]

let us know how it sounds, if/when

[Widows]

I didn't have any 5k pots kicking about so I used a 10k instead:

At 10k ohms the signal seems to be fully attenuated as i get no sound, but once I reduce the resistance a bit some signal gets through.

Reducing the resistance further from the point the signal begins to become audible to around 2k ohms it sounded pretty good, but as soon as you creep under about 2k that same nasty squealing oscillation I was getting with the 2 knob version appears again. I still only have it hooked up to a breakout box at the moment and am aware that the circuit would be prone to more interference than if it was in a metal case, but this is a lot worse than I've experienced with any other circuit I've had hooked up to a breakout box. It doesn't sound like interference from radio waves as per your classic distortion unit problems, but more like a sound from some issue with the way the signal is being precessed within the circuit.

Would a sound clip help?

Cheers

[Widows]

Now then, finally got round to finishing this little beast. Swapped the 10k dist control out for 5k lin after picking some up recently, re-seated the diodes, and boxed it up. No more horrible oscillations and it sounds great, the signal still disappears completely when I turn the dist control knob down past halfway but frankly I'm not that bothered as I'll probs be using it maxed out anyway heheh. Thanks for all your help folks, very happy with the results