Comment 1

Original : http://www.blueguitar.org/new/schem/_ss/smokey.txt
From:

Dave Stork
Date:
8/5/2001 5:19 PM
Subject:
Observations on the Smokey circuit



Just for fun, I built one of these during my lunch break the other day. I built it
inside an Altoids can and used the circuit exactly as shown. Here are my
observations after a couple of afternoons of playing with the circuit.

As drawn, the circuit wants to oscillate badly and is very sensitive to component
placement and the length of the battery leads. The supply pin (pin 6) should be
bypassed to ground, as close to the chip as possible, using at least 0.1uf (TODO!). This
took care of the audible oscillation, but there were still some spurs that
manifested themselves as spiky "farting" on otherwise clean tones. This might
account for the complaints in various reviews that I've read that the Smokey is
incapable of producing clean tones. Adding a 10 ohm resistor and .022 cap in
series from pin 5 (output) to ground stabilized the amp and restored the clean
tones. (TODO!) It did not have a negative effect on the distorted tones. You can try values
from .01 to .1uF if you like. This R-C network should also be located close to
the chip.

The 47uF output coupling cap shown on the schematic will start rolling off bass
around 400Hz. This may have been done on purpose to keep the small speaker
from farting out. With a "real" guitar speaker connected, you'll get better results
with a larger value cap. Try any value you like, up to 250uF for guitar. Even with
the bigger cap, the little transistor radio speaker I was using didn't sound any
worse than it did before. (the one on pin 7 I guess ?)

Incidentally, those little speakers are usually optimized for good treble response.
I don't know of any tiny speakers designed for guitar use. (Are you reading this,
Ted?) You can use a variety of mechanical means of damping the cone to get
rid of the "buzzies" and make the speaker sound a little better. This can be as
simple as putting plastic sheet or cloth in front of the cone, or perhaps even a
little thin cloth on the cone. If you can make the cone a little more dense, it won't
respond to high frequencies as well.

Some reverse polarity protection wouldn't hurt, especially if the amp might ever
be used with an external power supply. A silicon rectifier diode between the
supply and ground (cathode to supply, anode to ground) makes a crude but
effective "crowbar."

I don't know if Zinky did this by design, by accident, or just to save the cost of a
part, but one really unusual thing about the design as drawn is that pins 1 and 8
are connected together directly instead of through a cap. Not only does this
bypass the emitter of the differential input pair for higher gain, but it also affects
the DC operating point of the internal transistors, which must have some effect
on the tone. You could replace this shorting jumper with a 10K pot and see how
the tone and gain of the circuit changes as you turn it up and down. (ok, I have a 10kOhm on version 2)

A very useful addition is a pot on the output to allow you to control the volume.
A 25 ohm pot would work well whether you're using the unit to drive a speaker
or the input of another amp. When driving an 8 ohm speaker, the minimum
resistance seen by the amp will be 6 ohms, well within permissible limits.

The little amp sounds fairly dark, which might be due to the guitar pickup being
loaded by the 50K input impedance of the LM386. This might also explain the
fairly low value of output coupling cap Zinky chose... He may have been trying
to balance out the resulting muddiness. An input buffer seems a good idea here.
A buffer would also allow you to insert gain and tone controls at the input to the
386, with the gain of the buffer being chosen to make up for the insertion loss of
the controls, plus maybe a little more gain for more distortion if you like. The
possibilities of the basic concept are endless.

I was surprised by the performance of this circuit, and I can see why so many
Smokeys have been sold. The distortion is really dynamic and responds well to
pick attack; it's also surprisingly smooth for a solid state overdrive sound. With
some simple mods as outlined above, it cleans up very nicely when you back off
your guitar volume or lighten up your pick attack. It drives a "real" guitar speaker
(in my case, the 10" alnico eminence in my Pro Jr.) to surprisingly loud levels;
and it makes a great stompbox, if you add an output level control. There are
many improvements and refinements that can be made, and I look forward to
experimenting with this circuit in my spare time.

I never gave much thought to the LM386 for guitar use--I always considered it a
little "junk" chip for transistor radios and tape decks--but Zinky's simple, clever
concept really has me intrigued. This is one of the best little solid-state practice
amps I've ever used, even with its shortcomings.

START NEW THREAD
REPLY
PREVIOUS
LIST

Read 211 times

From:

Bruce (MissionAmp@aol.com)
Date:
8/5/2001 9:47 PM
Subject:
Re: Observations on the Smokey circuit



Jeeze Dave, I didn't know we were allowed to have lunch let alone do all that
during lunch!
damn bro'...your the "man"!
Now let's stick a padded down high zed FET in front with a volume pot with
tweed type tone stack between it and chip input, hook the output up to one of
Ted's inexpensive 8" signature speakers... all in a little 10" x 12' x 6" cab.
We can have a Storknose amp.
Bruce

START NEW THREAD
REPLY
PREVIOUS
LIST

Read 184 times

From:

Dave Stork
Date:
8/6/2001 1:44 AM
Subject:
Re: Observations on the Smokey circuit



Bruce wrote:

Jeeze Dave, I didn't know we were allowed to have lunch let alone do all
that during lunch!
damn bro'...your the "man"!

Lunch was an unknown concept when I was a full-time amp tech. But now I'm
an engineer for Uncle Sam, and lunch is mandatory. Gotta love it!

Luckily for me, my office is a workshop well-stocked with parts, tools and test
equipment. Along with the stack of empty Altoids cans I've been accumulating
on my desk, I had everything I needed to build my "Smokey" in no time flat.

Now let's stick a padded down high zed FET in front with a volume pot
with tweed type tone stack between it and chip input, hook the output up to
one of Ted's inexpensive 8" signature speakers... all in a little 10" x 12' x
6" cab.

Yep, that's basically what I had in mind as a next step... either a JFET buffer or
an op-amp, and a simple tone and gain control at the input to the 386. That's a
good suggestion about the 8" Weber; I hadn't thought of that. I know Ted was
making Champ speakers, but I haven't looked at his product line lately.

We can have a Storknose amp.

Would the volume knob be shaped like a bill?

Speaking of the Pignose, I bought one a while back and discovered the only
schematic available on the web was full of errors. So I carefully traced mine out
and created a corrected schematic. I can CAD it and post it if there's any
interest.

Small amps have captured my interest lately because I live in an apartment and
have been getting back into playing lately, after a long hiatus.

START NEW THREAD
REPLY
PREVIOUS
LIST

Read 175 times

From:

Tim C. (tcowles@cableone.net)
Date:
8/6/2001 2:01 AM
Subject:
Re: Observations on the Smokey circuit



Dave,

Shoot! Now you have given me another project to work on. I didn't need that!


Really, thanks for taking the time to post all that valuable information.

By the way, where did you get the speaker and what were the specs of it?

Thanks,
Tim C.

START NEW THREAD
REPLY
PREVIOUS
LIST

Read 182 times

From:

Dave Stork
Date:
8/6/2001 2:06 AM
Subject:
Re: Observations on the Smokey circuit



By the way, where did you get the speaker and what were the specs of it?

The speaker? Oh geez... It's just a little crap 8 ohm square thing, alnico magnet,
about 3" wide. I bought a dozen of them from Jameco to replace dead speakers
in the intercom system in our TV studio.

Like I said, the little amp sounds much better through a "real" guitar speaker.
But as I wrote earlier, you can kludge the little junk transistor radio speakers to
sound slightly better for guitar, at the expense of lowered output. The main thing
is to cut the treble response.

START NEW THREAD
REPLY
PREVIOUS
LIST

Read 108 times

From:

Dave Stork
Date:
8/10/2001 5:01 PM
Subject:
The circuit mutates...



So, here's where I'm at with the LM386 amp...

Common-source JFET amp stage driving tweed-type tone and volume controls
into inverting input of 386, series R-C from output to ground to stabilize amp as
per previous post, 220uF output coupling cap, master volume control to
speaker/line out.

I am very, very pleased with the tone and response of this little monster. It has
much more oomph than the original Smokey circuit, and is definitelty less prone
to oscillate I'll post a schematic when my experiments are done.

START NEW THREAD
REPLY
PREVIOUS
LIST

Read 73 times

From:

Bruce (MissionAmp@aol.com)
Date:
8/10/2001 9:25 PM
Subject:
Re: The circuit mutates...



Common-source JFET amp stage driving tweed-type tone and
volume controls into inverting input of 386, series R-C from
output to ground to stabilize amp as per previous post, 220uF
output coupling cap, master volume control to speaker/line
out.

I am very, very pleased with the tone and response of this little
monster.

HA ha ha.. thank you, I thought for sure I'd be the one to have to do it!
Should be kinda tubey sounding with a little RC "sizzle" control circuit in there.
Bruce

START NEW THREAD
REPLY
PREVIOUS
LIST

Read 18 times

From:

Dave Stork
Date:
8/11/2001 5:27 PM
Subject:
Re: The circuit mutates...



Should be kinda tubey sounding with a little RC "sizzle" control circuit in
there.

It does sound and play surprisingly "tubey"... not much of the usual solid-state
nastiness going on at all. But the discrete JFET amp is, by its very nature, rather
low gain and high output impedance and is loaded down somewhat even by the
simple volume-tone control arrangement I used. I had to use fairly low-value
pots because the following impedance (the input to the LM386) is pretty low
(50K). It actually sounds a bit more ballsy when the JFET drain is coupled (via
.022uF) right to the 386 input. With that arrangement, it makes a good "no
controls" circuit in the spirit of the original Smokey.

For volume and tone controls, though, I'm starting to think that a better
approach might be to use an opamp input stage or perhaps two JFET stages.
More experiments to take place this afternoon...