Dead Easy Dirt Issues

[chromesphere]

Hey everyone,
Building the dead easy dirt effect for another 'build a guitar pedal in 10 mins' video, but i have struck a problem. 

http://www.forrestwhitesides.com/node/93

Basically, the output is ridiculously loud (see voltages).  And the dirt effect sounds pretty much like a bluebox.  All glitchy and not very musical. I realise its 5 components / not a refined beast, but i dont think its behaving how it should.  There is 8v on the output which explains the mental volume output, but also adjusting the volume pot does nothing.  When its at max and min the volume turns off completely.
Heres my voltages:

Pin 1 & 8 - 1.36v
Pin 2 3 4 & 5 - 0v
pin 7 - 8.6v

Diodes positive - 8v
output - 8v
ground 0v obviously.

Positive i have wired this up properly...
Anyway, any help would be great
Cheers,
Paul

[Jdansti]

Well, it is a 386 amp capable of 1/4 to 1watt output. The volume pot should work, so check the wiring.  Also, try removing the connection between pins 1 and 8 or place a 1k linear pot between them to give you control of the gain.

[Electron Tornado]

Putting a pot between pins 1 and 8 never really gave a very good sweep range to the gain. However, you could try different pot values, like 5k, and different tapers. You could max the gain on the chip and try a pot in front of the input pin and see if that gives a better "drive" control.

If you max the gain, you won't really need those clipping diodes. The chip will happily distort all on its own.

When using a 386 like this, a 386N3 behaves better than a 386N1. There is also a 386N4, but I've never tried it.

[chromesphere]

Hey thanks for the responses.  Checked the wiring.  Its definitely correct...

Im using this op-amp, not sure what that "L" suffix means though!?

http://www.taydaelectronics.com/lm386-lm386l-audio-power-amplifier-ic.html

Paul

[Jdansti]

I agree about trying the volume pot on the input.

Also, most 386 applications use a 10R resistor and 50-100nF cap in series between pin 5 and ground to prevent oscillations.

[chromesphere]

Well it magically started working.  God knows why...didnt change a thing...must have been a short somewhere.

Anyway, this circuit is a good one to start learning with and building upon i reckon.  Might have to do that at some stage!

Thanks for everyones help!  The problem lied between my chair and the circuit board for this one!

Paul

[Jdansti]

Glad you got it going! 

[slacker]

...must have been a short somewhere.

Probably from pin 6 to the diode side of the output cap, that's about the only way you could get 8 volts on the volume pot.

[chromesphere]

Tried it this morning, doesnt work... 

Good pont slacker!  I was thinking how it would be possible to get that sort of voltage on the output -> DC isnt being decoupled.  There is nothing shorted around the cap though, so im going to replace it. 

[chromesphere]

replaced the cap, no difference...

Im concluding that this circuit has a design flaw and is unstable.

Paul

[ashcat_lt]

I don't know if it's the same with the 386, but with other opamps if there's a DC path to the negative rail in the feedback path, it will push the output as close to the positive rail as the opamp will go.  Basically the thing runs open loop, and even tiny amounts of noise get amplified to as loud as it gets.

[Jdansti]

^Is that why the manufacturer specifies a cap and 10R resistor from pin 5 to ground on the 386?

[chromesphere]

I think im going to have to get my head out of this circuit and start from scratch...i breadboarded another one, same issue.  So i decided, screw it and converted it to a smokey amp.  Same issues.  After a HUGE amount of messing around i think i...may...have narrowed down adding a cap from +9v to ground turns the ic a sounding like its busted, to working...maybe theres something esel going on with my build.  Highly likeyl, theres stuff everywhere at the moment..
Going to take a break from it and return later.  Thanks for everyones responses!
ps, john that out filter / anti-oscillation cap / res did nothing for some reason.  Infact, building the amp they way the datasheet says, still sounds like rubbish...yep.  Going to need to take a break

Confirmed, if i removed everything from the circuit, so there is just a 47uf cap on the output, and i remove the cap (50nf and 2.2uf have been tested with the same result), the output goes from glitchy 'sounds broken' to working normally...

The strange part is, if i add my power filter to the 9v dc input (which has caps to ground, its basically the beavis huminator) it makes no difference.  if i run the 9v dc striaght in, no difference....BUT AS SOON as that cap hits +9v to ground, bang, its alive...

Paul

[Jdansti]

Yah, sounds like there's something funny going on that you haven't put your finger on yet.

Btw, I've seen comments online where others have had similar problems with lm386 amps. One suggestion is to swap pins 2 and 3. In other words, ground pin 3 and connect the input to pin 2. No guarantees, but it's a quick and easy thing to do on a breadboard.

[timd]

Yah, sounds like there's something funny going on that you haven't put your finger on yet.

Btw, I've seen comments online where others have had similar problems with lm386 amps. One suggestion is to swap pins 2 and 3. In other words, ground pin 3 and connect the input to pin 2. No guarantees, but it's a quick and easy thing to do on a breadboard.

I agree 100%. I have messed around with the lm386 so much I should try to be sponsored by somebody. Some circuits work better with a specific input. Also Paul - are all the chips you are using from the same source? This could narrow the issue. Do you have any JRC386's you could try?

[Jdansti]

+1

[chromesphere]

Hey guys,

I got the smokey amp to work.  That cap from plus to ground fixes it.  Petey Twofinger confirmed that this cap needs to be as close the IC as possible which is exactly what i was experiencing.  The power filter didnt fix it, but a cap on the breadboard did.  Thats absolute insanity, but there you go...

I will try the dead easy dirt again, might fix its issues as well.

Edit: I recorded a video of the circuit and the cap issue, sort of vlog style.  Thanks for your help:

Paul

[Jdansti]

Great video!

Have you tried the cap and resistor from pin 5 to ground to see if it still oscillates when you crank up the booster on the front end?

[tca]

If you want dirt you can go beyond the 40dB datasheet value to some higher values up to 72dB!

Cheers.

[pinkjimiphoton]

paul,
i made a version of this too, it's a little weird, but sounds great.
check it out...schem/ vero/ demo video in this thread. it uses a pre-(and passive) tone control that makes it a lot nicer sounding in my opinion.
peace!

http://www.diystompboxes.com/smfforum/index.php?topic=96856.0