Problems with Ruby LM386 amp on breadboard

[deadastronaut]

going by the video, it looks suspiciously like your fet is backwards...

flip it around...

with the flat side facing you it should be

D S G


guitar into G.

[slashandburn]

Maybe not fair for me to have suggested op-amps are "easier" to use on the front end of the amp, btw.  I started off with a Ruby circuit, but didn't have any Jfets and stumbled upon what at the time seemed like the holy grail.

If the trannies aren't working for you and it's all still proving a headache,  strip it right back to a Smokey amp.   Once you get that working on the breadboard, add the input buffer with the JFET (or opamp if you go that route).

Think of them of independent circuits though.   Essentially, the "Smokey" part of the circuit is the amp.   The Jfet bit in front of the 386 chip is a different circuit to ensure the 386 chip see's a good signal from the guitar.   Get the amp running first and then tack on the input buffer later might be a good way to approach this if you're not having fun.

http://www.electrosmash.com/images/schematics/smokey-amp-schematic.gif

[Hydrilla]

going by the video, it looks suspiciously like your fet is backwards...

flip it around...

with the flat side facing you it should be

D S G


guitar into G.

I do have the input going to gate, which is what the diagrams I have been using suggest.  When I reverse it, there is no sound.  The diagrams I am using also have G going to the positive rail, and completing that part of the circuit is what causes the loud BEEEEEEEPing.

[Hydrilla]

Maybe not fair for me to have suggested op-amps are "easier" to use on the front end of the amp, btw.  I started off with a Ruby circuit, but didn't have any Jfets and stumbled upon what at the time seemed like the holy grail.

If the trannies aren't working for you and it's all still proving a headache,  strip it right back to a Smokey amp.   Once you get that working on the breadboard, add the input buffer with the JFET (or opamp if you go that route).

Think of them of independent circuits though.   Essentially, the "Smokey" part of the circuit is the amp.   The Jfet bit in front of the 386 chip is a different circuit to ensure the 386 chip see's a good signal from the guitar.   Get the amp running first and then tack on the input buffer later might be a good way to approach this if you're not having fun.

http://www.electrosmash.com/images/schematics/smokey-amp-schematic.gif

going by the video, it looks suspiciously like your fet is backwards...

flip it around...

with the flat side facing you it should be

D S G


guitar into G.

I do have the input going to gate, which is what the diagrams I have been using suggest.  When I reverse it, there is no sound.  The diagrams I am using also have G going to the positive rail, and completing that part of the circuit is what causes the loud BEEEEEEEPing.

I did get a super basic LM386 circuit working earlier tonight, one that just uses a few components.  It was really distorted but hey it was something.  At least it proves some of my components aren't damaged. 

Here is an screenshot of the .fzz file I have been going off in my most recent half a dozen attempts.  I could be mixing something up I suppose- the input jack and the 10k pot aaren't really obvious where they go.  If someone feels like checking this out for me, I would appreciate it.  If I don't have this figured out soon, I think I'm going to admit I'm a f'n idiot, and go back to building kits.

[slashandburn]

at a quick glance I can't see anything wrong besides you're polarity protection diode. Either way it's not the problem.

Other '386 quirks that might be worth playing with:

try running the input into pin 3 and grounding pin two.      And get that polar filter cap (that goes from 9v to ground) as close to Pin 6 as you can get it.

It was really distorted but hey it was something

That'll be the jumper between pins 1+8.     Remove the jumper and you get less gain.

[Ben Lyman]

don't give up yet!
definitely try without the buffer. Gate should not go to pos rail
check out this lesson on buffers too:
http://www.muzique.com/lab/buffers.htm

[Hydrilla]

 I should have mentioned that I don't have the diode, so mine is like this but without that.

[PRR]

> have G going to the positive rail

That does not sound right.

Usually to Gnd or Vref through a large resistor.

[Ben Lyman]

Looking at that pic, I see what is probably
pin 1 to out cap and 3k3 (to ground)
pin 2 to in cap and 1M (to ground)
pin 3 to +9v,

So its gotta be  (if JFET)
1 S
2 G
3 D

maybe its BJT?
1 E
2 B
3 C

[Hydrilla]

Thanks for that.  You got me thinking.  I really figured the issue must be around the transistor, so I took away the caps and resistors, jumpered the pot to the MPF102, eventually to the drain, moved the input jack around, and now it is making sound.  Oh and I also got rid of the jumper to the positive rail. 

I'm using an unknown 4" speaker from a Goodwill stereo speaker, so I don't know the quality, but the volume seems ok.  I would expect it to be somewhat louder, but it is a lot louder than anything I have stumbled across so far.  With just jumpers, not surprisingly it sounds like a bad fuzz pedal, but hey I'm happy with progress. 

With the flat side of the D of the transistor facing me, what I think is the drain is the left lug.  Right now I don't have anything else going to the other 2 lugs.  Both the guitar input and the pot (I have normal ones, I am just using this little blue 10k one for testing) are on the left lug. 

It is too late for me to think through how to add the other stuff back correctly, but again I am happy with progress.  I don't like plug and pray, but whatever works at this point.

EDIT:  Ben, you replied at the same time I did... from what I have read about the MPF102, it is SGD, but I also read that some manufacturers make them with a different pinout, so I don't know if mine is typical or not. 

[Cozybuilder]

You might as well try wiring it as if it is a standard MPF102 first:

[Hydrilla]

OK so I realized what I've done with what I described above is effectively eliminate the JFET altogether.  I can remove it completely, and the sound doesn't change at all.  Still, progress.

[pinkjimiphoton]

sounds to me like the fet is shorted internally. you may be able to pop a normal q in there, some of them will work. but them voltages are troubling.

a too-low voltage to the 386 can make it oscillate.

but i'm betting the issue is something else.

make sure pin 6 connects directly to ground

make sure your decoupling cap is as close to the chip as possible.

[Hydrilla]

sounds to me like the fet is shorted internally. you may be able to pop a normal q in there, some of them will work. but them voltages are troubling.

a too-low voltage to the 386 can make it oscillate.

but i'm betting the issue is something else.

make sure pin 6 connects directly to ground

make sure your decoupling cap is as close to the chip as possible.

According to all the diagrams I see, pin 6 goes to positive. 

I just set it all up again using this diagram, and I am getting sound versus the beeping, and it is using the MPF102.  I have 1 and 8 jumpered and it sounds kinda crappy, so I'll try a pot in a little while.  Not sure why this worked and the others didn't.  It has a few less components, but has stuff reversed from the previous diagram, where the JFET is concerned. 

[slashandburn]

You're right, 9v goes to pin 6. I'm not 100% sure but I think Jimi meant have a big electrolytic cap going from Pin 6 (as close to pin 6 as possible) directly to ground. Either move the one from your power rail closer to pin 6, or just another on there. On breadboard I sometimes have the cap hover directly over the IC chip. Positive end to pin 6, negative end to pin 4.

Glad you have some sound.

Regarding the pot on pins one and 8, it's "backwards" from what you'd expect. There's an international resistor between these pins. Hampering the pins bypasses this resistor and give you a ton of distortion, removing the jumper removes the distortion and gives you a cleaner sound, via the internal resistance.

Forget the pot until you have it rocking, just remove the jumper it's too distorted.

I don't mean to nitpick but your latest diagram shouldn't work without continuity between your two breadboard ground rails (same with the power rails) but if you have sound I'll assume this a discrepancy on your diagram and that you do in fact have them jumpered.

[pinkjimiphoton]

lol...yeah, my bad!! needed coffee or something. i was thinking b+ and wrote ground. sorry!

[Hydrilla]

First of all, thanks to you guys and the others for taking the time to help me.  I am learning from this, believe it or not. 

If I leave 1 and 8 open, the volume is very low, enough to not be worth building.  Of course jumpering 1 and 8 or adding a tone pot boosts the volume a lot, but it gets pretty muddy.  Even with a pot (I don't have a 1k, only 10k, 100k and 250k), it is either quiet or pretty buzzy, like a cheap fuzz pedal. 

I am sure some of it is my speaker.  It is a ~4" speaker from a 3 way that I bought at Goodwill, so I'm sure that could be a factor.  But I don't think I'm getting enough power to it.  I don't have a cab handy, and have been too lazy to drag my Hot Rod Deluxe in here to test with.

EDIT: Oh yeah and I do have the positive and negative rails jumpered between the left and right sides.

[Hydrilla]

I did what ROG calls the Bassman mods, mainly adding a resistor between 1 and 8.  It added a bit more volume, yet isn't muddy.  It might be time to leave it alone for now and experiment with some speakers.  Here is a quick sound sample, it is difficult to tell much about the volume from an iPhone video.

This is with one of my Danelectro guitars- with my Jaguar, it is much quieter.  So I need to figure out how to get more volume without a bunch of gain.

[pinkjimiphoton]

k, i know what your problem is now.

the 10k pot is way too high for that circuit, and the chip becomes unstable. 1k is actually pretty high for it!

try putting a 1k resistor across the input and center of the gain pot and see if the oscillation goes away.

i just recently had worked up my own version of this circuit (search : ruby doobie) and i tried a 10k gain pot.

the whole thing was just like an oscillator!!



google a parallel resistor calcutator ... or better a tapering resistor calculator. you should be able to strap a resistor across the pot and get it not perfect, but functioning.

[Ben Lyman]

Jimi- I bet you're right.
I also want to reiterate the fact that I had major board probs until I star grounded. It didn't matter how neat and tidy I kept everything or how short I made all the jumpers, etc.
The oscillation instantly went away when I star grounded everything right off the LM386 ground pin. But then of course, mine was set permanently on a gain of 200.
When I laid it out on perf, I also took this into account although I'm not sure it is quite as important when it's actually being built. Makes a world of diff on the BB tho.