Diode sound differences
Screaming Skull
them switchable what two sets would give the most noticeable change in tone?
LED or germanium, germanium or silicon, etc or etc.....
any suggestions are welcome. thanks!
Tim Escobedo
won't necessarily be in tone but in volume. Germanium diodes clip at
the lowest levels and LEDs at the highest.
Presuming just the diodes are changed in a circuit, the Germanium
would probably give the most distortion, but also have the least
volume output, causing you to open the output knob, meaning more
noise.
LEDs would probably have the least amount of *diode induced*
distortion. But some circuits begin to clip themselves before the LED
threshold is reached, so some of the distortion heard may actually be
part of the circuit itself, not the LED clippers. LEDs also allow the
most volume to pass, neaning you may need to cut back the volume knob
and a better signal to noise ratio. This higher output may also
distort any following stages that may be present.
Silicon diodes clip at a level between LEDs and germanium.
There isn't any one kind of better clipping diode. It all depends on a
number of factors and ultimately, your personal taste.
Steve Daniels
self-destruction. There are an infinite variety of possible
combinations, and any one may or may not be the one that makes your
hair stand on end.
Buy a solderless breadboard from Radio Shack, and then find one of the
many circuits out there for a basic IC- or transistor-based
boost/distortion. Make it work, and then start fiddling with the diode
loop and the coupling caps.
There is a simple and inexpensive one, the Tweak-O, on my site:
Http://home.netcom.com/~smallbearelec
and I am pretty sure Jack Orman has an FET-based one at Muzique.
You can find the parts at Radio Shack or on my stock list.
Regards
Steve Daniels
Small Bear Electronics
Howard Davis
think the question was not about relative levels but about TONE.
Germanium diodes (and transitors, when driven into distortion) have a
slightly "softer" saturation "knee" in their non-linear
(distortion-producing) voltage vs. current characteristics. The distortion
they produce therefore has a smoother, somewhat more musically harmonious
sound. The sound has more of the even distortion-generated harmonics, as
compared to the harsher odd harmonics generated by such "hard clipping"
devices as silicon diodes and LEDs.
The circuitry these devices are employed with also affects the tonal
characteristics, but in general, go with germanium if you want a more
tube-like sound. If you want a biting or edgy fuzz, silicon is your thing.
Tim Escobedo
> What I've read above in this thread is generally true and informative, but I
> think the question was not about relative levels but about TONE.
My point is that it's sometimes difficult to really assess tone when
there is such a dramatic difference in volume. After compensating to
volume changes, it can sometimes lead to misleading presumtions about
the tone associated with a particular type of diode. For example,
assume a circuit is feeding a pair of diode clippers a signal +-1Volt
in amplitude. LEDs would be doing some rounding off of the peaks but
probably not any hard clipping, as the clipping threshold would not be
reached. Silicon diodes, with foward voltages at less than 1V, might
be cutting off just the peaks abruptly and deforming the waveform as
it reaches that peak. This would probably sound harder than the LEDs.
Germanium diodes might transform the waveform to resemble a square
wave the most, having the lowest clipping threshold, and would
probably have the hardest distortion sound.
> Germanium diodes (and transitors, when driven into distortion) have a
> slightly "softer" saturation "knee" in their non-linear
> (distortion-producing) voltage vs. current characteristics. The distortion
> they produce therefore has a smoother, somewhat more musically harmonious
> sound. The sound has more of the even distortion-generated harmonics, as
> compared to the harsher odd harmonics generated by such "hard clipping"
> devices as silicon diodes and LEDs.
I'm not sure this always translates well to real circuits. A germanium
diode will "hard clip" just as much as any other diode depending on
the amplitude of the signal being fed. The clipping qualities may
depend more on other factors such as opamp/transistor circuit than the
diodes. Also, the power of suggestion can be a bigger factor
determining the "musical" qualities of certain clipping methods. I've
heard well respected folks extoll the virtues of LED clippers,
claiming they create more even-numbered harmonics. The same goes for
asymmetrical clipping methods. And to further confuse things, take a
look at popular fuzz/distortion/overdrive boxes. Some "overdrive"
boxes use silicon diodes, LEDs, or none at all. Some "fuzz" boxes use
germanium diodes, silicon, or none at all. There's certainly no
consensus over which overdrive sounds most natural or which fuzz
sounds fuzziest. And the clipping methods used run the spectrum.
> The circuitry these devices are employed with also affects the tonal
> characteristics, but in general, go with germanium if you want a more
> tube-like sound. If you want a biting or edgy fuzz, silicon is your thing.
Again, tube like sounds or biting, edgy sounds are not clearly
associated with any particular type of diode. The WHOLE circuit needs
to be designed for any particular performance. The most dramatic
differences among diode types will be in clipping levels (the amount
of clipping) and the associated tonal qualities associated with the
amount of clipping.
Serious Shawn
Can you name a few? I'd be curious to try them out.
Also, you mentioned other parts of the curcuit clipping. What other
parts of the circuit can clip, and what do they tend to do to the
signal? Do you remember off the top of your head any common devices
that clip like that? I'm always interested in any unusual forms of
distortion.
Shawn
Daniel R. Haney
Screaming Skull wrote:
>
> If I were to put two different sets of diodes in a distortion pedal and make
> them switchable what two sets would give the most noticeable change in tone?
> LED or germanium, germanium or silicon, etc or etc.....
For diodes in a feedback loop clipper, the important traits are
turn-on voltage, turn-on speed, and conduction curve below
turn-on voltage.
The turn-on voltage determines the maximum signal output.
Ge diodes (.4 volt drop) limit the amplitude to +/- .4Volt.
Since their turn-on time is slower, they produce fewer
and less high order harmonics, unlike silicon switch diodes
such as the 1N4148.
The turn-on speed determines how much fizz and RF noise
you will need to equalize out.
The conduction curve below turn-on voltage isn't horizontal.
or flat. It's non-linear and determines a great deal about
the clipper's sound below the hard turn-on (where all clippers
sound more alike than different).
Germanium diodes sound "dark" or "soft" while silicon diodes
sound bright and fizzy. LED's turn on at 1.3 volts and above,
have a longer non-linear resistance region. FET and MOSFET
transistors wired as diodes are very dark sounding until they
turn on very hard.
You can extend the conduction curve by putting diodes in series.
The older Marshall Bluesbreaker pedal does exactly that to achieve
its sound.
-drh
--
Tim Escobedo
> I haven't seen a commercially available device that does LED clipping.
> Can you name a few? I'd be curious to try them out.
Marshall Drive Master, Dano Pastrami also reputed to use LEDs.
> Also, you mentioned other parts of the curcuit clipping. What other
> parts of the circuit can clip, and what do they tend to do to the
> signal? Do you remember off the top of your head any common devices
> that clip like that? I'm always interested in any unusual forms of
> distortion.
Transistors and opamps themselves can clip when providing more gain
than parameters will allow. Fuzz Face uses no diodes and Boss OD-2(?)
Turbo Overdrive has a non-diode overdrive "channel". Just about every
fuzz/OD/dist box uses an amplifier circuit. But amplifiers can only
amplify a signal depending of the limits imposed by supply voltage and
semiconductor characteristics. So if you crank the gain up enough,
it'll clip even without diodes. This is sometimes some of the hardest
clipping around, when the circuit reaches its' limits, it really
reaches the limit. Viewing waveforms clipped this way show very abrupt
clipping of the peaks. It can sound good or bad depending on the
circuit.
Howard Davis
news:3819137c.0202...@posting.google.com...
> "Howard Davis" <hdavis.i...@rcn.com> wrote in message
news:<a3scnb$7sa$1...@bob.news.rcn.net>...
> > What I've read above in this thread is generally true and informative,
but I
> > think the question was not about relative levels but about TONE.
>
> My point is that it's sometimes difficult to really assess tone when
> there is such a dramatic difference in volume. After compensating to
> volume changes, it can sometimes lead to misleading presumtions about
> the tone associated with a particular type of diode. For example,
> assume a circuit is feeding a pair of diode clippers a signal +-1Volt
> in amplitude. LEDs would be doing some rounding off of the peaks but
> probably not any hard clipping, as the clipping threshold would not be
> reached. Silicon diodes, with foward voltages at less than 1V, might
> be cutting off just the peaks abruptly and deforming the waveform as
> it reaches that peak. This would probably sound harder than the LEDs.
> Germanium diodes might transform the waveform to resemble a square
> wave the most, having the lowest clipping threshold, and would
> probably have the hardest distortion sound.
I meant a comparison under the same relative drive conditions, same
circuitry, and the differing output levels compensated for. LEDs have a
sharp knee characteristic like silicon diodes, and there would probably be
no audible difference between them.
> > Germanium diodes (and transitors, when driven into distortion) have a
> > slightly "softer" saturation "knee" in their non-linear
> > (distortion-producing) voltage vs. current characteristics. The
distortion
> > they produce therefore has a smoother, somewhat more musically
harmonious
> > sound. The sound has more of the even distortion-generated harmonics, as
> > compared to the harsher odd harmonics generated by such "hard clipping"
> > devices as silicon diodes and LEDs.
>
> I'm not sure this always translates well to real circuits. A germanium
> diode will "hard clip" just as much as any other diode depending on
> the amplitude of the signal being fed. The clipping qualities may
> depend more on other factors such as opamp/transistor circuit than the
> diodes. Also, the power of suggestion can be a bigger factor
> determining the "musical" qualities of certain clipping methods. I've
> heard well respected folks extoll the virtues of LED clippers,
> claiming they create more even-numbered harmonics. The same goes for
> asymmetrical clipping methods. And to further confuse things, take a
> look at popular fuzz/distortion/overdrive boxes. Some "overdrive"
> boxes use silicon diodes, LEDs, or none at all. Some "fuzz" boxes use
> germanium diodes, silicon, or none at all. There's certainly no
> consensus over which overdrive sounds most natural or which fuzz
> sounds fuzziest. And the clipping methods used run the spectrum.
Asymmetrical distortion (rectangular waveform, triangle wave, sawtooth,
etc.) contains both odd and even harmonics. A square wave has odd harmonics
only, and therefore sounds harsher, less sweet, less tube-like, etc. And
yes, the circuitry outside of the non-linear devices themselves makes a big
difference in sound, often more of a difference than the non-linear devices
themselves do.
> > The circuitry these devices are employed with also affects the tonal
> > characteristics, but in general, go with germanium if you want a more
> > tube-like sound. If you want a biting or edgy fuzz, silicon is your
thing.
>
> Again, tube like sounds or biting, edgy sounds are not clearly
> associated with any particular type of diode. The WHOLE circuit needs
> to be designed for any particular performance. The most dramatic
> differences among diode types will be in clipping levels (the amount
> of clipping) and the associated tonal qualities associated with the
> amount of clipping.
True.
Gary
>
> There isn't any one kind of better clipping diode. It all depends on a
> number of factors and ultimately, your personal taste.
I think someone should experiment with two different types of diodes
to generate more even harmonic content.
Also there's a Schottky diode which only uses a "half" junction
(rather than P and N type semiconductor, it's just one type and the
other side is a metal connection)... as a result the forward drop
voltage can be 0.3 V instead of 0.6 - 0.7 typical of silicon diodes.
But they tend to be more expensive so it's unclear if anyone would
bother on a commercial fuzz device.
BUT THE HOME EXPERIMENTER ON THE OTHER HAND...
Tim Escobedo
>
> Also there's a Schottky diode which only uses a "half" junction
> (rather than P and N type semiconductor, it's just one type and the
> other side is a metal connection)... as a result the forward drop
> voltage can be 0.3 V instead of 0.6 - 0.7 typical of silicon diodes.
> But they tend to be more expensive so it's unclear if anyone would
> bother on a commercial fuzz device.
>
> BUT THE HOME EXPERIMENTER ON THE OTHER HAND...
Good points about the Schottky diodes. it's easier to compare
Schottkys with germanium diodes since the forward voltage drop for
both types seem to be in about the same range. Though more expensive
than average silicon diodes, they seem to be cheaper than germanium
(which seems to be disappearing in these parts). Also, the point about
uneven clipping is well taken, using dissimilar diodes or using an odd
number of diodes to clip a waveform asymmetrically. Depending on the
circuit, I like doing this to get hard clipping while slightly
retaining some dynamics of the guitar. The possibilities are nearly
endless.
Smallbear will hook you up with the parts...
Kate Ebneter
>
> I haven't seen a commercially available device that does LED clipping.
> Can you name a few? I'd be curious to try them out.
One of the modes on the TS9DX uses LEDs.
Kate Ebneter
Collector of Noise Toys
Daniel R. Haney
> Good points about the Schottky diodes. it's easier to compare
> Schottkys with germanium diodes since the forward voltage drop for
> both types seem to be in about the same range.
The Schottky's aren't as non-linear or conductive as germaniums
before hard turn-on, but they turn on harder and faster. You
get all the fizz of a silicon switch diode but at a lower clip
voltage.
> Though more expensive
> than average silicon diodes, they seem to be cheaper than germanium
> (which seems to be disappearing in these parts).
After purchase by MicroSemi, BKC did their final run on germanium
diodes back in Fall 2000. They are responsible for the light blue
1N34A Ge diodes you saw at Radio Shack a few years back. I know of
no other source in the world and believe me, I looked. All germanium
diodes are now NOS.
-drh
--
Screaming Skull
which made it sound better with obviously lower output though. Many cheapy
practice amps use them as well. I had a Fender frontman and a Dano that
used them in the built in OD circuit.
Thanks to all that contributed to this thread. I now have my head straight
on this
Steve Daniels
> >
> > Also there's a Schottky diode which only uses a "half" junction
> > (rather than P and N type semiconductor, it's just one type and the
> > other side is a metal connection)... as a result the forward drop
> > voltage can be 0.3 V instead of 0.6 - 0.7 typical of silicon diodes.
> > But they tend to be more expensive so it's unclear if anyone would
> > bother on a commercial fuzz device.
> >
Tim--Any suggestion for a part number? I will be happy to put it on
the stock list in spring.
Tim Escobedo
> the stock list in spring.
I've been using 1N5817s for my Schottkys. Seem to work well.
Tim