Low Gain, neutral dist/OD

[Killthepopular]

What collector voltage are you operating at?  It can affect the sound quite a bit.

tranny is 2n3904. using standard 9v batt.

[PRR]

What collector voltage are you operating at?

[Killthepopular]

What collector voltage are you operating at?

Is that the measurement between pin 3 and ground? it says 2.18v. I've now swapped r1 2m2 with a 10m. This reading is now at 4.7v. Is this more or less correect?

Based on my tests so far I find that germanium seems to have a more natural, valve-like clip than silicon or LED.
Silicon seems to give a bigger and deeper tone than LED and has more natural clipping, but silicon also adds some slightly ugly harmonics which make things a bit shrill. LED clipping sounds weirder than silicon but it doesn't seem to add shrillness in the same way silicon does.
Asymmetric clipping is maybe a bit bigger, more dynamic, less compressed sounding than symmetric.

[Killthepopular]

I'm finding that Asymmetric can start sounding subtly distorted so you aren't sure if the sound is clean or distorted. It sounds like a mix of clean and dirty. Whereas Symmetric clipping seems to instantly go from very clean to very distorted with a quick turn of the gain knob. It's less good at sounding slightly distorted.

I've also found that headroom seems to make a big difference. When I compare 2 silicon diodes with 4 silicon diodes (one on either side vs 2 in parallel on either side) I find that the 2 diodes give me a smaller, thinner, harsher tone with less bass. 4 diodes however gives me a much warmer, bassier, darker, looser, less tight sort of sound. I'm adjusting the gain and volume pots to give me the same amount of gain and level in both instances.

What am I hearing here? Is this a standard consequence of higher headroom? Or am I hearing something else? Maybe turning down the output means that the impedance that my amp receives changes and so it sounds deeper and darker because high end is being lost? Is it normal for higher headroom to yield a darker, warmer tone?

[Mark Hammer]

Use a pair of germanium or schottky, but insert a 500R-1k pot between the diodes and ground.  That will provide for some very hard clipping (albeit with low output), but the ability to soften the clipping as that resistance to ground is increased.

[Rob Strand]

I'm finding that Asymmetric can start sounding subtly distorted so you aren't sure if the sound is clean or distorted. It sounds like a mix of clean and dirty. Whereas Symmetric clipping seems to instantly go from very clean to very distorted with a quick turn of the gain knob. It's less good at sounding slightly distorted.

The Electra circuit isn't 100% symmetrical itself.  When the circuit clips the output impedance isn't constant like and opamp. In the positive direction you have up to 68k and in the negative direction you have 0 to 10k depending on you trimpot value. 

The 100k level pot is actually pretty low compared to the 68k.  IIRC the original Electra circuit used 4k7 collector resistor which helps reduce the output impedance.  There's many variants over the web, I know I've seen a 10k version as well. You will need to adjust the other resistor values to compensate for changes in the collector resistor.



You can put resistor in series with the output cap to make the impedance more symmetric but it's probably only worthwhile if you have a lower collector resistor.

The Electra circuit isn't exactly 100% transparent since it loads the pickup.  The design which uses the 68k resistor will load the pickup less and the design with 4k7 will load the pickup more.  So there's a bit a of a balancing act keeping all the non-ideal behaviour in check.

[Killthepopular]

I think I'll definitely go with Asymmetric clipping. It does a very good job of slowly getting subtly dirtier (Which is what I want for this project), whereas symmetric clipping seems more prone to being either totally clean or totally gritty.

I Like 1n4001 and 1n34a diodes the most. I might use a combination of those or maybe just 3 1n4001s. I've also ordered some 1n5817s because someone suggested trying a schottky.

What happens when I have two different types of diodes in series? Say I have an LED and a 1n34a. I understand that the forward voltage or headroom will be equivalent to both of them added together, but what about the tone? Will i get something that's halfway between the two tones? I've read that the lower voltage diode (1n34a) will turn on first so you'll get the tone of just that one diode but at a higher headroom. Does that sound correct?

[Mark Hammer]

Anything you do to increase the forward voltage required for the diodes to conduct will have three consequences:
1) It will take a hotter signal to result in any clipping,
2) Any clipping produced will be for a shorter portion of the note's lifespan (since less of the signal will be above threshold),
3) With headroom increased, the maximum output level will be higher.

The "tone" produced by any given type of diode is largely a function of how its' respective forward voltage affects that stuff.  I find there's a lot of stuff that people attribute to magical diode properties that simply results from forward voltage.  Diodes DO differ in many characteristics, but many of those characteristics are more relevant to signal switching at speeds >500khz that to the puny 6khz bandwidth guitars produce.

Following that logic, sticking several Schottky diodes in series, to achieve the same total forward voltage as a silicon type, will get you the same clipping as that silicon type.  It will NOT get you a tone that is any sort of blend of what schottkys or germanium alone would get.  I understand that conduction speed can vary with the current feeding them, but that's typically something you can tweak with whatever resistor is in series with the signal.

So, if you want hard clipping you can either raise the gain so that more of the signal falls above the clipping diodes' forward voltage, or you can reduce the forward voltage of the diodes by making different diode choices.  Conversely, if you want less clipping, you can either lower the gain, or raise the clipping threshold...or both.

[Rob Strand]

What happens when I have two different types of diodes in series? Say I have an LED and a 1n34a. I understand that the forward voltage or headroom will be equivalent to both of them added together, but what about the tone? Will i get something that's halfway between the two tones? I've read that the lower voltage diode (1n34a) will turn on first so you'll get the tone of just that one diode but at a higher headroom. Does that sound correct?

Suppose you have two identical diodes and put them in series.   The intrinsic "tone" of that combined diode is the same a single diode.  However, it becomes difficult to compare apples to apples:   The voltage drop of the combined diode is now double but the voltages in the rest of the circuit have not doubled.  So even though the intrinsic tone of the combined diode hasn't changed the circuit as a whole will sound different.

So suppose we ignore the fact the sound of the *whole circuit* changes due to different voltage drops then we can say something:  The tone of putting two diodes in series is a mix of the two diodes, however it's not as simple has half of one and half of the other.  The tone is weighted towards the diode with the larger voltage drop.   This formula, kind of represents what happens,

      combined diode tone  =   (VD1 * "tone of diode 1"  + VD2 * "tone of diode 2") / (VD1 + VD2)
      where VD1 is the voltage drop of diode 1 and VD2 is the voltage drop of diode 2.

So adding a germanium to an LED is like adding a dash of germanium salt as the LED voltage drop is much larger and dominates the tone.

Notice that formula shows two identical diodes having the same tone.

[Steben]

https://www.facebook.com/ArgenzianoEffetti/videos/1084476765074158/

Check this out... Silicon Fuzz face clean up. Love it!

[Rob Strand]

Check this out... Silicon Fuzz face clean up. Love it!

Ah, I can't see videos on Facebook.
Was there a schematic?   (Adding caps around the transistors can help.)

[Killthepopular]

I built the electra and fiddled with loads of different diodes and settled on an asymmetrical configuration of 1n4001 diodes. I think it's best for what I want. Does a good off clean tone and has a nice warm fuzzy quality.

Moved onto the red llama. Very impressed with it. Doesn't give loads of gain but if you want lots of gain it takes boosts shockingly well and seems to have a ton of output too, so it doesn't have a lot of it's own gain but could easily fit into a high gain setup. Sounds really good at low gain. Very natural and amp-like. None of that on/off, sputtery clipping that you get from standard distortion setups. Doesn't seem to colour the sound and only gets slightly brighter as you push it harder. Very good all round. Probably trumps the electra all in all. Will probably move onto the vox distortion next.

Thanks to the three or four guys who recommended the red llama. I didn't know anything about this circuit but it seems like a really great circuit as well as being a close match to what I wanted. I still have two more circuits to try but I find it hard to imagine that I'll like them more than the Llama.

[Fancy Lime]

You're welcome! That's what this forum is for, among other things.

Cheers,
Andy

[Rob Strand]

Thanks to the three or four guys who recommended the red llama. I didn't know anything about this circuit but it seems like a really great circuit as well as being a close match to what I wanted. I still have two more circuits to try but I find it hard to imagine that I'll like them more than the Llama.

Maybe you should tinker around with variations of those CMOS gate based overdrives.

The Laney amps have used CMOS gate based overdrives for quite some time.  You might get some ideas by looking at the circuits.  Some channels have more gain than others.

For example,
Laney pl50rh
https://elektrotanya.com/laney_pl50rh_schematic.pdf/download.html
Fairly sure cmos devices are running from 5V (trimpots adjusted for 3.1V at outputs).

Some others were Laney hc25, Laney hcm65r.
A lot of the HC and HCM models.

You can play with the supply voltage and also add resistors to shift the DC bias points at the outputs of the gates.   IIRC some of the Laney schematics have trimpots and show the voltages they use.

[Killthepopular]

Maybe you should tinker around with variations of those CMOS gate based overdrives.

The Laney amps have used CMOS gate based overdrives for quite some time. 

I'll test these 2 remaining circuits (vox dist and hotcake) and If the red llama is my fave then I'll mod it a bit before I start thinking about laying it on vero or perf. I might also want to stick a simple pre-overdrive boost (LPB1 or SHO) with a footswitch in the same enclosure too.

My main amp is actually a laney LV100. Do you know if that uses a cmos type overdrive?

[Killthepopular]

I'd suggest the Vox 1901, same circuit, but some different parts values. It needs a reverse logarithmic gain pot, just like the DOD - some layouts falsely show a log taper.

Doh! I remember reading this but I must have forgotten and ordered a log pot. Can I just connect it up backwards and get the same taper? Albeit with a knob that increases gain when you turn it counter-clockwise?

[Rob Strand]

My main amp is actually a laney LV100. Do you know if that uses a cmos type overdrive?

I'm not sure about that model I've got no info on it. 

[roseblood11]

I'd suggest the Vox 1901, same circuit, but some different parts values. It needs a reverse logarithmic gain pot, just like the DOD - some layouts falsely show a log taper.

Doh! I remember reading this but I must have forgotten and ordered a log pot. Can I just connect it up backwards and get the same taper? Albeit with a knob that increases gain when you turn it counter-clockwise?

Yes, no problem.
And that circuit sounds good with a SHO in front of it

[Killthepopular]

So I've tried the 4 circuits. Electra was good but maybe a bit samey sounding. Red Llama was really impressive. Dod 250/vox distortion was awesome but not really a natural amp-like sound. Hotcake was cool but a long way from having amp-like qualities.

Think I'll give the peppermill a go. It sounds pretty bad to me from the clips but a lot of people seem to really like it and I have most of the bits and it looks very easy to breadboard.

http://www.runoffgroove.com/peppermill.html

I don't have 6.8n or 2.2 n but I have 10n and 1n. Also I don't have B250k but I do have B50k or A250k. Will those substitutions work ok? Just for a quick breadboard layout.

[snk]

I built the Red Llama last week.
It didn't sound the way i expected(i wanted it a bit darker and less saturated), so, thanks to the helpful community here i have tweaked it quite a lot and now it's very good.

I have also heard people raving about the Fairfield Barbershop, but i haven't tried it yet.