which diodes sound like what?

Started by vseriesamps, October 12, 2004, 12:24:40 AM

Previous topic - Next topic

WGTP

I always enjoy this subject.  It is easy to experiment and tweak on and has cool results.  

Op Amp to Diode Clipper Distortion Ratio = THD

In messing with the Muff Fuzz using diodes in the 2nd stage feedback loop, or to ground, you can go from 2 GE's for the most compressed, distorted tone, thru 2 SI's to 2 LED's and finally no diodes at all with plenty of distortion at all points.  You can also use a single diode to clip one side and allow the op amp to clip on the other (thanks ALF) instead of matched pairs, or you can use mixed pairs.  IMHO "most" of the changes you hear are from the clipping thresholds changing.  Usually the 2 GE or SI diodes sound too flat and undynamic for me.  It keeps the distortion from really ripping things, having an edge.  But, I like sustain and so a certain level of compression is desirable.  I do seem to prefer mismatched pairs.   :twisted:  I think they are more responsive and have a thicker sound.  I also think some of the diodes may sound brighter than others, but it may be due to the ratio above.  Different op amps clip at different levels as R.G. pointed out in the IM thread.  One that clips at low levels, (I'm guessing the 4558, LM308, 741, etc.) probably is effected less by clipping diodes than the rail to rail models that are passing larger signals before distortion.   8)
Stomping Out Sparks & Flames

brett

Hi.  Wow, great thread.
QuoteFor sure there is a variation in the character of the "knee". The harder the knee, the more odd-order harmonics generated
Actually, the characteristics of the "knee" are remarkably similar between all diodes (except Scottkys) *on a change in current per change in volts basis*.  

However, we almost always use clipping circuits where the final output voltage level (not the change in voltage) is constant.  e.g. changing Ge to LEDs in a Distortion+ increases the output level, so you turn down the volume.

It is this adjustment of gain that emphasises the characteristics of different diodes.  Per unit of increasing volume, current increase in a Ge diode is much less than in Si, which is less than LEDs.

How much less?  Well, since they are nearly all the same once they are clipping, the difference is mainly due to the LEVEL at which they clip.

Using a 2V LED as a standard, a 0.7V Si diode will be (2/0.7 = 3) times "softer", and a 0.4V Ge diode will be (2/0.3 = 7) times "softer.

I should say that all of this involves taking shortcuts and is a comparison of diode characteristics, NOT anything to do with what a particular circuit does with those characteristics.  I totally agree that diode placement has a large effect on the diode effect (e.g. op-amp feed back vs. grounding).  To me, I really like asymetric 1N4148 feedback diodes in a tubescreamer or bluesbreaker, and asymetric 1N400X and/or 1N4148s in a distortion plus or a bosstone.

cheers
Brett Robinson
Let a hundred flowers bloom, let a hundred schools of thought contend. (Mao Zedong)

jmusser

I don't know exactly what I just read on this thread, but I'm copying it all off so I don't loose it once I do understand this stuff a little more. Right now, I'm trying to tie this in with what Aron wrote in the simple mods and tips about "differences in diode tones". I realize that diodes (nor any other component for that matter) has a tone, but I know what he meant by his post. This is the tonal effect that he has experieced by switching diodes types, or adding diodes matched and mismatched. If I'm understanding the rest of this thread correctly, you who are in the know, are going to try to give specific diode tone characteristics for specific diode layouts in common circuits. Is that about right? If that's what you're saying, it will be envaluable information for this forum!
Homer: "Mr. Burns, you're the richest man I know"            Mr. Burns: Yes Homer It's true... but I'd give it all up today, for a little more".

RDV

I don't know that much, but I know I like the sound of my Distortion + clone with 2 1N914 & 1 Ge diode. It really is very dynamic. I started out with 2 Ge in there and somehow the thing just stopped working. Couldn't figure it out for weeks till I narrowed it down to the diodes. One of them had gone south somehow(it broke when I tried to remove it). I decided to go assymetrical this time and I don't regret it a bit, as it's now a killer!

RDV

jmusser

That just goes to show that you never know about people. I turned my back on you for just a minute, and now you're going assymetrical on me! I have some proper diodes coming to put in your Tone Bender, so I hope that gets it up and running to hear the kazoo effect. BTW, build the Bronx Cheer, it rocks big time!
Homer: "Mr. Burns, you're the richest man I know"            Mr. Burns: Yes Homer It's true... but I'd give it all up today, for a little more".

jmusser

You knew I meant transistors right? I've heard that kazoo sound on several different recordings, and always wondered what effect they were using to get that sound. I've heard Hendrix do it on a recording, and he may have been using a fuzz face for all I know. I think a lot of times, it comes down to what equipment you're using besides the effect. I hope once I do put in the 4401s in, it that the sound will be pronounced. Believe it or not, the Bronx Cheer can do a pretty good "Satisfaction Fuzz". In Tim's sample, he picks the notes pretty hard to get that synthy sort of sound out of it. If you play it with a light lighter attack, and don't let the notes ring to show off the effect, you get a "Satisfaction" type buzz. If you want the synth buzz to be really pronounced, give a harder attack, let the synth come out, and then lift your finger up enough to let the string buzz on the fret. Really cool! If you go through the notes quickly enough so that the synth doesn't come out while you're at the bridge pick up, you can get a pretty good "Whisker Biscuit" smooth fuzz out of it. As you can see, I'm a "Cheer" fan.
Homer: "Mr. Burns, you're the richest man I know"            Mr. Burns: Yes Homer It's true... but I'd give it all up today, for a little more".

WGTP

I don't mean to offend anyone, but IMHO the standard 2/SI diode combination is the worst sounding there is.  It is lifeless, overcompressed, thin, harsh, etc.  Any other combo sounds better, especially mis-matched or assymetrical pairs or using a single diode.  The additional "tube like" dynamics and even ordered harmonics produce a snarl that is much more desireable, at least to me.  The mid-range is thick and rich and the bass is the same.

I think op amps need a little more room to breath with higher and mixed clipping thresholds.  YMMV   I think this is why some people don't like op amp based distortions, and they can sound great if you give them a twist.  :twisted:
Stomping Out Sparks & Flames

petemoore

SAturation control...a pot and maybe a cap on the clippers in a DIST+, adds a nice knob, adds a variable option that 'opens up' the sound...seems to 'dial down' the diode clipping and let the Opamp tone come through.
  Quite the recommendable item for those 'messing about' with clipping thresholds and frequencies.
 My DIST + Satyration knob is rarely set all the way up [maximum diode clip effect]. IIRC it's got a GE --->l  and a Ge + other diode to l<---
to ground...
Convention creates following, following creates convention.

aron

Again, the problem is that everyone's setup sounds different. I've heard many 2 si clippers that sound fine. It depends on the setup and what R.G. has said for a (decade)??

Pre/Post clipping EQ.

gdmart

are the diodes in a octavia used in the same manner as other pedal layouts,or do they do something eles since they are closer to the transformer.What does the transformer do in this pedal??

brett

Hi.  Yes, the diodes in octave pedals aren't primarily for distortion.  They are there to rectify the signal, so that a signal that had positive and negative parts is converted to only positive parts.  

Why?  Because the number of positive and negative parts are approximately equal, and converting the negative parts to positive makes for twice as many positive parts, and the frequency is effectively doubled.  Cool trick.

You don't necessarily need a transformer, but it conveniently makes two signals that are out-of-phase.  The alternative is to tap the emitter and collector of a transistor, which is what an Octavia does.

cheers
Brett Robinson
Let a hundred flowers bloom, let a hundred schools of thought contend. (Mao Zedong)

gdmart

ok--- so in a tycobra octavia the diodes rectify in with the transformer which puts out positive and negative power but then you said---- You don't necessarily need a transformer, but it conveniently makes two signals that are out-of-phase. The alternative is to tap the emitter and collector of a transistor, which is what an Octavia does.  so in the octavia there is a transformer.Is the emitter and collector of only 1 transistor ,or all transistor tapped???
    please educate me :lol: