Purple Plexi 800 (LM386) hiss/squealing noise problem

[Plexi]

Hi forum!
Yesterday I re-build in pcb (not breadboard) this distortion/od based around LM386.



When the 1k Gain knob is maxed, it made a squealing/hiss noise.

As I can read, there's a common problem in this schematic.
I hadn't this hiss when breadboarded it time ago, same LM386N3.

Is there any way to avoid that noise? Any experiencie?

What I'll try:
- Increase R3: increasing the "min" amount of gain before ground
- Increase C3? (From +9v to ground)

Another thing I don't get it; is if I swap to another LM386, N-4 in this case, it sounds really really bad and gated.
Voltage need of this particular design? I know it can works with more (>16)
I used on another pedals with 9v, and worked fine.

[deadastronaut]

have you got a 386-1?

[Plexi]

have you got a 386-1?

I haven't 
Should I try one, before mess around?

[deadastronaut]

yeah i would....they are different

1 = 1/4 watt

3 = 1/2 watt

4 = 1. watt

[Plexi]

Sure.. know about that difference.
I'll try: maybe it's some kind of oscillation due the Watts.

[Cozybuilder]

I haven't studied the 386, so may be off base here. What I know is the gain of the 386 is adjusted by the relationship between pins 1 and 8, shorted gives a gain of 200, open is gain of 20. In addition, the gain set-up with the parallel R4 (470R) and 1K pot adjusts the amount of AC gain through cap C4. R3 gives a minimal resistance for attenuation. I suspect that you might have too high of gain with that particular chip, which suggests at least 2 ways to attack the problem.

1) The gain pot when fully CW gives 0 resistance to ground, thus shorting R4, so the effective RC on pins 1 & 8 is R3-C4 to ground. You might increase R3 slightly, or add a small amount of resistance to the wiper of the pot to ground.

2) Possibly reduce the gain of the LM386 by adding resistance between pins 1 and 8.

[Plexi]

Thanks Rob and Cozy

I was reading about the JRC386D suggested... and it suppose to be the equivalent of the LM386-1: both 1/4w (~250mW).
I think, like Rob supposed, it's a wattage feedback/oscillation.
I'll try the 386-1 before start to tweak around.

As someone says, it's almost a Krank Distortus Maximus.
But, they solve with better, consistent and useful solution: just go from pins 1-8-->100R (instead 10R)--->47uF---->ground
And boost/clean at the input.

If the problem persist with the 386-1... I'll follow that way.

[Plexi]

Update, problem solved!

That's work:

What I'll try:
- Increase R3: increasing the "min" amount of gain before ground

1) The gain pot when fully CW gives 0 resistance to ground, thus shorting R4, so the effective RC on pins 1 & 8 is R3-C4 to ground. You might increase R3 slightly, or add a small amount of resistance to the wiper of the pot to ground.

R3 from 10R to 39R.
That's why I suppose the 100R are a solution on Distortus Maximus... at the end, the LM386 are pushed with the LPB1-kind of the front.

[Plexi]

Messing around LM386's google search, I found this similar version:
http://2.bp.blogspot.com/-xqH5Q9Q0XYw/TzOt99GqJlI/AAAAAAAAA5g/Ux6id6Ditx0/s1600/BlackMagic.bmp

It's strange how it uses 500R pot, and 1R resistor to ground.
If with 10R oscillates/squeals as crazy... with this method It'll squeal as a pork in mating season...
I'll breadboard and see. But, as Rob says,I believe it's the LM386 over saturated.
I'll wait for some LM386-1

[Plexi]

What would happen, theoretically speaking, if I switch C4 to 10uF?
Less overall distortion?

[EBK]

Man, I read the thread topic, saw the author's name, and was worried we were dealing with a medical crisis! 

[Plexi]

Man, I read the thread topic, saw the author's name, and was worried we were dealing with a medical crisis! 

...don't forget the purple-face of the avatar guy.....