Noisy Cricket too much... clean!

[Elijah-Baley]

Hello guys!
I need help.
I built the Noisy Cricket using this layout:


Here's the schematic:


Noisy Cricket Analysis: http://www.electrosmash.com/noisy-cricket-analysis.

Somebody want this kind of LM386 amp cleaner, but they can't. I can, but I don't want it. My Noisy Cricket sounds really clean!
I thought this amp was dirty, but with Gain and Volume at max it is stil clean, and the the Grit make really little difference. Tone maybe is a bit subtle, but it works. Even the Bass Boost is a bit subtle, but I think it works.
I was using a LM386N-1, but then I broke one of its pins . I replaced it with a LM386L (thanks IC Socket). No difference. I connected the status led, it works.
I forgot the filter 220uF cap on the 9v, I'm using a 100uF on sockets, but tried a 330uF. The transistor is a MPF102, but I tried a 2N5457 and a J201. Nice clean with them, too.
I power up it with a Boss psu. But tried with battery, too.
Now the Gain pot is got off, I joined the two wires coming from pin 1 and pin 8.
Checked the resistors, caps, wires.
If I take off the 100uF the power filter cap I got a drop of volume, a fuzzy and gated distortion

LM386L voltages:
Pin 1: 1,32v
Pin 2: 0,00v
Pin 3: 0,00v
Pin 4: 0,00v
Pin 5: 2,66v
Pin 6: 8,99v
Pin 7: 4,61v
Pin 8: 1,32v

MPF102 voltages:
D: 8,99v
S: 2,68v
G: 0,00v

Thank you, guys!

Rewritten the voltages, sorry.

[PRR]

It is odd that pin 5, we expect at +4.5V or +5V, is under 3V.

There could be a short screwing-up bias, but I'm not clear where.

And low pin 5 should distort *sooner*, not later?

[Elijah-Baley]

Thank you.
I'm trying to put it on breadboard, I hope I can understand something more.

[Elijah-Baley]

I'm following this breadboard layout.



I connected pin 1 and 8 of the LM386. Gain at max.
I'm using a jumper for the grit switch, I can connect or disconnect it.

I want to replace the tone pot with a resistor (50k?). I want leave out even the volume pot, but with volume setted at max. But I need some instructions for all this connection on the breadboard.
Thank you.

[Gargaman]

Isn't a cap necessary between pin 1 and 8 for maximum gain?
When volume pot is at max. the lug 2 is connected to lug 3. So, for a fixed setup you remove the pot and connect C2 to the next point, i.e. VR3 (or your 50k resistor suggestion) and C6.
I might be wrong.

[Elijah-Baley]

There's no cap between pin 1 and 8 in the schematic nor in the veroboard layout. Probably it isn't necessary.

Anyway, the circuit on my breaboard sounds bad, distorted but I can hear the volume barely and seems much gated. It is a cheap breadboard I can't trust on it. Unless I wrong to connect the jumper and the resistor for volume e tone pot.

I passed the iron solder on the board, but nothing changed.
The wiring: ground from the board, negative of the battery and the negative of the led are joined, and them goes to the output sleeve. The output sleeve is connected to the input sleeve. Input from the board goes on the tip of the input jack, output from the board goes to the tip of the output jack.

[PRR]

> Isn't a cap necessary between pin 1 and 8

Not really.

Worst-case, the input offset DC error is multiplied by gain, 200. With 5mA input error, that is 1V output error. "4.5V" could be 3.5 or 5.5. When the '386 was new, National wanted you to use the cap so your output power measurements didn't fall shy of the spec. But recent '386 should have much lower offset, and we are not measuring for an exact output power.

[Elijah-Baley]

I can't built it again, I have to fix this.
I want to take off everything the amp doesn't need.
Gain pot is already left out. I used to joined the two wires (pin 1 & pin .

I think I can't take off the Grit Switch, and its 100nF cap, between Grit 1 and pin 5. The amp should keep working anyway.

It is odd that pin 5, we expect at +4.5V or +5V, is under 3V.

There could be a short screwing-up bias, but I'm not clear where.

[...]

The only things connected to the pin 5 should be the 10R resistor, the "Grit 100nF cap" and the positive side of the 220uF cap.

I can't explain this:

[...]
If I take off the 100uF the power filter cap I got a drop of volume, a fuzzy and gated distortion
[...]

Why?!

[PRR]

> If I take off the 100uF the power filter cap I got a drop of volume, a fuzzy and gated distortion

The power supply IS a "signal pin". (Especially if you have not heavily bypassed the Bypass pin.)

Output load makes power sag, that gets back in the amplifier, gets re-amplified, it screams or throws itself on the floor (gates).

In general "ALL" amplifiers need good supply supply bypassing. Some more than others. Extra-more here because it is a pretty high current amp.

Find the National LM386 datasheet and try the TESTED plans in there.

[karbomusic]

I can't built it again, I have to fix this.

My first LM386 based circuit was the build that made me reconsider (for a short time) if I was cut out for any of this. I had grabbed a layout and schematic from the net and just couldn't make it work, horrible buzzing etc. no matter what I did, then another problem, then another. I finally said screw it, grabbed the datasheet for the 386 and the datasheet for the MPF102, studied them and the example circuits in them and rolled my own. Both amps I built as a result are still working today and one is my bedroom practice amp.

So the point is stick with it, In my humble, inexperienced opinion that circuit can be tricky enough to drive you mad because you can't know what you don't know - or (even more likely) I was simply that green at the time. Nevertheless I cut my teeth making that damn thing work but what I learned helped with every build after that.

[Fndr8875]

what kind of speaker are you using? what Ohm is it?  if you notice in the schematic there is a 220uf output capacitor, but on the layout it is only 10uf. From my understanding you can play with the power resistor 3.9k that goes to MPF and go up to 10k, How good of shape is your breadboard in. Mine is worn out, and i did a simple build last night that gave me a hugh range of sounds just from randomly wiggling different parts. I know Aron who runs this site has a tutorial to use a 386 as a simple voltage divider to get your 9v, 4.5v, ground for IC build, so from my understanding a 386 that is working properly with output 4.5v, so if your getting 3 could be the chip. IM Still Learning, and cant backup technically any of how or why these things work like they do, but ive done alot with the 386 so these are just observations that MAYBE could help.

[Elijah-Baley]

I could try the tested schematic on the LM386's datasheet, but the Noisy Cricket is very verified, even the tagboard layout. It should work fine.

I built a Smokey Amp, with gain pot (without caps), power filter cap (it needs with the PSU) and master volume. It works great, sparkle clean and good hard rock style crunch.

I tested the circuit with an 8" guitar speaker 8 Ohm.

There's a mistake on the breadboard, I know. My board has a 220uF as output capacitor.
I tried a LM386N-1 (I broke a pin, then andI soldered it, I don't know if it will work again), and now I'm trying with a LM 386L.

Today I have to test it with the LM386L of my Smokey Amp. Then I'll try to take off the Grit switch and its cap. Let's see what happens.

[Elijah-Baley]

I swapped the LM386L with that on my Smokey Amp, no changing.
On my Smokey Amp the IC's voltages are almost the same, except pin 5. My Smokey Amp measured 5,88! The Noisy Cricket 2,66v. The Smokey is distorted and louder.

I'm gonna take off the Grit Switch now.

Edit:
Still clean without Grit Switch and 100nF cap.
I replaced the output cap 220uF with socket pin and tried a 100uF (I had just it), nothing.
The volume pot is Linear, not Log as the tagboard layout would, but I guess it doesn't matter.

Edit 2:
Desoldered the Tone pot. Clean.
Desoldered the Volume pot e soldered the input wire on the strip of pin 2. Clean! Very nice clean, but CLEAN!

Except some parts the big difference between this and the Smokey Amp is the inpunt. Pin 3 for the Smokey Amp with Pin 2 and 4 at the ground, Pin 2 for the Noisy Cricket (and Ruby Amp), with Pin 3 and 4 at the ground.

And now?

[Gargaman]

Looks like you will end up with a good -- CLEAN -- mini amp project
at least 

[Elijah-Baley]

Yes, I'm considering this, but I was building an amp with a gain pot and a Grit Switch and Bass Boost. I get just a bit of break up if I hit hard the strings, at max gain.
Now I got Smoeky Amp with some filters and it still sounds clean.

Should I reverse pin 2 and pin 3? I wanted put input in pin 3, and pin 2 and pin 4 at the ground. Should I get problems with this?

[Gargaman]

Man, I really got lost here..don't know.
My first 386 amp was a Smokey, (not CLEAN at all).

It was almost the same as the datasheet minimum parts aplication suggested.
Then I moved to Ruby (1/2 CLEAN).
I got problems with identifying the inverting and non-inverting input because mine was 386N something that was the inverse (pin 2 and 3) that the schematics I was followingd. I remember that it was then that I meet datasheets..
Whatever, I can't help you but wish you luck

[Elijah-Baley]

MySsmokey Amp sounds almost like that, I use it with a 8" guitar speaker, that make a big difference, I got more definition and a good eq.
I think LM386L or LM386N are pretty similar, maybe the same. L is by UTC, N is by Texas instruments.

[rutabaga bob]

This won't fix any voltage issues, but I've built a couple of Ruby amps - one I use now for a quick practice and effects-testing amp.  The one I took apart to scavenge was a MarshaRuby, which never cleaned up!  It utilizes the NTE458 jfet, which has more drive.

[karbomusic]

If my brief read of the datasheet and posts over on the TI forum are accurate, that 'grit switch' is an exploitation of the chip's design, meaning I'm sure it normally works but I'd have to wonder if it always works in every build and every rendition of that chip - all it is doing is strapping a 100n across pins 1/7 FYI. I'm sure they had a decent reason for it in lieu of just changing the gain structure via pins 1 and 8 but if you can't make it work you can get a gain up to 200 (translation=not clean) based on how you wire a pot/cap across pins 1 and 8. I'm guessing that exploit is to potentially provide a more musical grit than clipping outright.

And for those who actually know what they are talking about - keep me honest please.

[Elijah-Baley]

Thanks for the sharing, guys!

I don't know what exactly do the Grit switch because my issue, it make very subtle difference. But I have excluded, my issue is somewhere else.

Meanwhile, I breadboarded (this word exists?) my simplified version. I mean: LM386 with a 100uF as output cap, the RC 10R - 47nF, the 100nF between pin 6 and ground. I get a squealing sound (happend with the Smokey amp on breadboard, too), and it is distorted. The voltage on pin 5 is 4,10v, my battery is dying,

The LM386 doesn't sound clean, I bypassed the buffer, so there must be something wrong, some hidden mistake in the second half part. I have to find it.