Green Ringer Octaver Help?

[nick12391]

Hi all!
   I've recently breadboarded this circuit and it works perfectly when given a nice sine wave from a function generator. However, when I hook it up to the guitar I don't seem to be getting the octave shift I'm after. Any Ideas what could be going on?
Thanks!

Link to Green Ringer:

http://www.generalguitargadgets.com/effects-projects/octave/dan-armstrong-green-ringer/

Links to oscliioscope waveforms (Yellow is guitar, blue is output):

https://drive.google.com/open?id=0B-kTfmLz7qLKM1pYa3JXckZUUmc

https://drive.google.com/open?id=0B-kTfmLz7qLKVlZPTUg4aFZpUGM

And the waveform when given a sine wave (Although this is a model, it behaved comparably in practice)

https://drive.google.com/open?id=0B-kTfmLz7qLKTGVXUkFpZUpMaGs

Thanks for the help!

[Mark Hammer]

Congratulations.  You now know the difference between a signal generator and a guitar.

The secret to getting an octave doubler circuit like the Green Ringer to work well lies in making the guitar sound as close to a signalgenerator as possible.

That is normally done using the following strategy:

• Use the neck pickup
• roll the tone back on the guitar
• use a compressor before the octaver
• provide appropriate gain for the guitar
• play at the7th fret or higher

The less harmonic content you have, the better.  Fretting above the 7th reduces the harmonic content, because shorter strings are stiffer strings.

[Ben Lyman]

ya, Mark nailed it. I will go one further and say:
Play only the D, G & B strings between the 10th and 14th frets (just for testing purposes)
Also, try it with an OD pedal after it. Before it can be made to work if you adjust the tone properly but to keep it simple, stick the OD after to be safe.

[Mark Hammer]

I will simply add that if you replace the 1N914 diodes with Schottky types, you will find easier production of octaves, and ability to generate them probably down to the 5th fret in some instances.

[Ben Lyman]

Very cool Mark, I am going to try that. You can also put a 5K trimmer in series with one diode and it helps to find that sweet spot

Here's one I haven't tried yet, looks interesting. Mark, I think you mentioned something like this when I was working on my "Form-Inator"

[Mark Hammer]

I have a variable gain stage on the input of mine, and a 3-position filter switch, but I can't recall what I did, and given that it is perfed, I may have some difficulty decoding the mods.

I don't know that the Ringer absolutely requires 3 poles of lowpass, as the diagram above shows, but having at least one is probably a good idea.  The net impact is to reduce the need to roll the treble off completely on the guitar itself, making it possible to play bright and clean, hit a footswitch, and have the pedal performing as if you set everything just right.

The 5k trimmer you note is one of several things one can do to optimize octaving.  Conceptually, you want the two signals coming off the emitter and collector, that get rectified by the diodes, to be as close in amplitude as you can get them.  There are  a variety of things that affect the signal balance.

• the 10k emitter and collector resistors should be pretty close in value
• the emitter/collector caps (C2, C3) should be close in value
• the resistors to ground between each cap+diode pair should be matched
• the forward voltage of the two diodes (regardless of type) should be pretty close

Substantive differences in any of those component pairs would be expected to lessen the likelihood of a robust audible octave.  It won't eliminate it; just make it kinda touch and go.  How much the suggested 5k trimmer would be able to null out discrepancies in everything else is an empirical question.

[Mark Hammer]

And see my latest comments here:  http://www.diystompboxes.com/smfforum/index.php?topic=115951.msg1093281#msg1093281

[nick12391]

Thanks for the input guys! I have a few follow-up questions:

Here's one I haven't tried yet, looks interesting. Mark, I think you mentioned something like this when I was working on my "Form-Inator"

Ben, can you explain the function of this addition? I haven't seen this mod.

I will simply add that if you replace the 1N914 diodes with Schottky types, you will find easier production of octaves, and ability to generate them probably down to the 5th fret in some instances.

Mark, can you explain why shottkys are superior for this application?

• use a compressor before the octaver
• provide appropriate gain for the guitar

Why will compression and gain help? It seems like a higher input amplitude will cause clipping of the output. Or am I incorrectly assuming that you are referring to amplification?

And see my latest comments here:  http://www.diystompboxes.com/smfforum/index.php?topic=115951.msg1093281#msg1093281

What does the diode pair to ground accomplish?

Thanks again for the help I really appreciate it!!

[Ben Lyman]

Nick, the mod is to create 3 different low pass filters and, as Mark said, it's probably 2 more than necessary.
Also as Mark said, you can do the same thing by rolling off tone on your guitar and selecting the neck pup, as Jimi most certainly did. Listen to the area where Jimi plays most of his solos using the octavia, mostly around the middle of the guitar neck. hmm.. middle of the neck... like where the 12th fret octave harmonics lie.. coincidence? What say you Mark?

Check my last schematic in this early creation of mine, with much help from everyone in this forum:
http://www.diystompboxes.com/smfforum/index.php?topic=113381.msg1050204#msg1050204
I put a fuzz circuit going into the green ringer and put two different diodes on a blender pot in the octave portion

[Mark Hammer]

The signal coming off the phase splitter must pass through the diodes in order to be combined and heard.  Those diodes behave almost like a noise gate that allows the signal to pass when its level exceeds the forward voltage of the diodes.  The higher that forward voltage is, the more the input signal has to be strong to pass.  Silicon diodes like the 1N914 or 4148 have a forward voltage of around 500-600mv.  Germanium types are in the neighbourhood of 250-300mv.  Schottky types are generally under 230mv.  So Schottky diodes allow for octaving even when you're picking gently.  The different diode types all do exactly the same thing - separate one half cycle of the signal to combine with the other half cycle - but the Schottkys do it at a lower signal level.

Why is compression helpful?  Each stum/pluck/pick begins with a lot of harmonic content and an unsteady pitch (fretting aritrarily sets the length of the string but picking slightly corrupts  the length).   Shortly after picking, the harmonics are reduced and the fundamental is more obvious.  As well, the string is not bent by picking anymore and has a steadier pitch as it is vibrating freely.  All of that makes it easier for doubling of the pitch of the note to stand out and be more easily heard.  BUT, after picking, the signal the string produces drops a lot.  So, right when conditions are most favourable to be able to hear the octave, the string says "Nope, I'm not going to co-operate".

Compression allows for the level of the string to be maintained after the pick attack, so that the octave produced isn't substantially quieter than the initial pick attack.  This is also why Schottkys are good.  As the string amplitude declines, Schottkys permit even fading signals to pass and be turned into octave doubling.

It's a bit like conversation at a social gathering.  When everyone is there and the music is pumping, "conversation" consists of yelling "WHAT?", and even that gets lost amidst the noise.  After most everybody leaves, the music gets turned down,and there's just you and one or two others, the conversation consists of actual talking, and gets nuanced, easier to understand, more interesting, and more enjoyable.  But all of that depends on you still having the energy left in order to converse.  So the trick is to do what you can to keep the string energy around after the "harmonic crowd" has left,  and keep the circuit open to subtlties so you can talk at a normal level, and feel like the conversation is intimate.

The filter circuit is part of that conspiracy.  It eliminates a lot of the harmonic content that comes in at the beginning.  Of course, after the string settles down, there is less harmonic content, so the filter really only has something to do during the first few hundred milliseconds after you pick.  Still, it allows the fundamental to be  more audible as early on as possible.  And the more audible the fundamental, the more audible the octaving.

And yes Ben,  most historic benchmark solos using octaving tend to take place on unwound strings around the 9th or 10th fret and higher.  But then, those pedals were developed before we really set ourselves the task of producing analog octaving lower and lower on the fretboard.  Still it is very true that unwound strings, played higher up the fretboard, provide the optimum conditions for getting robust octaving: less harmonic content, and less pitch wobble when you pick (i.e., more stable pitch).

[nick12391]

Thanks for the tips, guys!

   I'm going to try resistor/capacitor matching and changing the diodes first, as that sounds easiest. Then if I still need to I'll try adding the filter mod and a compressor.

Also I've read that some people have replaced the PNP with an hfe 150? Have either of you tried this?

Also, Mark, I have little to no experience with dynamic range compression. Do you have any suggestions for the compressor?

Thanks and stay tuned for updates!

[Mark Hammer]

Read through this thread, and you'll see a recommendation to build the "compression" into the Green Ringer itself.  See the last post in particular. http://www.diystompboxes.com/smfforum/index.php?topic=115951.0

[nick12391]

Alright guys so I matched my resistors and caps and swapped the diodes with matched shottkys and that helped a lot, and I changed the pnp from a fry's version to an oldschool hfe150. I tossed in a single LPF between Q1 and Q2 but that actually seemed to hurt more than it helped, so I pulled it back out. I'll try to get a vid up soon so you can hear the sound.