What would be the best distortion for aural exciter like harmonics?

Started by erikb1971, August 07, 2011, 07:12:16 AM

Previous topic - Next topic

erikb1971

Hi all

Iam working on a 19" distortion unit in which 3 different distortions play different roles:



What happens is that the clean signal is split using a crossover filter at 450 hz. The low freq band is being distorted by a Rocktron Austin Gold. I love what is does with the pre bass on the lower frequencies. The band above 450 hz will be distorted by a modded austin gold. I have found a mod that will allow me to add a bit more drive, and working on a way to modify the pre bass to work in the higher part of the feq spectrum.

The third band will be a distortion that process the signal after a high pass filter on a fairly high frequency, or a distorted signal will be filtered by a same filter (pre or post distortion filtering). I was wondering if you have any ideas what (kind) of distortion or overdrive would be good for creating those high frequency harmonics.

All three bands will be mixed with the original signal.

Cheers

Erik

Mark Hammer

There are a few ways to generate those harmonics.  One is employed in Jules Ryckebusch's "Harmonic Sweetener", a version tweaked by RG and myself, and kindly drawn by RG is here: http://www.montagar.com/~patj/harmswtn.gif

It works fine, although I think that it might benefit from some kind of dedicated gate, like that found on the Boss CE-1 and A/DA Flanger.  Being able to adjust where the highpass cutoff is can be very useful.  Use of a dual-ganged pot to control one section of the 4-pole filter can make a bigger difference in usability that you'd think.

The form of distortion or diode type probably makes little difference, though Jules used a pair of LEDs.  Thinking about it now, the LEDs kind of make sense in that they don't provide quite as much clipping.  That means you'd get lots of the natural harmonics from the original input, only boosted, plus whatever additional harmonic content the clipping produces.  The LEDs also make sense in that they are being used with a +/-15v supply not 9v.  For 9v, you might want to opt for something a little different, like silicon type.

An other approach is illustrated by this Aphex B schematic, kindly posted by Stephen Giles: http://img.photobucket.com/albums/v317/StephenGiles/aphexB_grey16.gif

Here, a 13700 OTA section is used, much like a ring modulator, only modulating the signal with itself.  You will know that a ring modulator produces the sum and difference of the input and modulating signal.  If the two are the same waveform and frequency, then what you get is frequency doubling (differences cancel out).  So, it's the same proportional harmonic content as the original, only shifted up an octave.  Never knowingly heard one so I can't comment.

soggybag

What about changing the filter to a parametric type. This way you could control the frequency and width where you placed the distortion.

erikb1971

Quote from: Mark Hammer on August 07, 2011, 08:19:44 PM
There are a few ways to generate those harmonics.  One is employed in Jules Ryckebusch's "Harmonic Sweetener", a version tweaked by RG and myself, and kindly drawn by RG is here: http://www.montagar.com/~patj/harmswtn.gif

It works fine, although I think that it might benefit from some kind of dedicated gate, like that found on the Boss CE-1 and A/DA Flanger.  Being able to adjust where the highpass cutoff is can be very useful.  Use of a dual-ganged pot to control one section of the 4-pole filter can make a bigger difference in usability that you'd think.

This looks like an interesting option! Especially with the suggested mods of the adjustable filter and the drive option!  With the dedicated gate you mention, do you mean a filter like in my conceptual schematic?

Thanks a lot!

Erik

erikb1971

Quote from: soggybag on August 08, 2011, 10:49:29 AM
What about changing the filter to a parametric type. This way you could control the frequency and width where you placed the distortion.

you mean the high pass filter that is in my schematic? The idea is to only add "new" things to the signal. I am afraid (but correct me if I am wrong) that changing it in a parametric will add too much overlap with the other high bands...

Thanks for your input anyway!

soggybag

I was envisioning a filter with a frequency and width control to allow you to place the distorted signal in a desired "location". Just thinking out loud.

erikb1971

Quote from: soggybag on August 08, 2011, 06:15:16 PM
I was envisioning a filter with a frequency and width control to allow you to place the distorted signal in a desired "location". Just thinking out loud.
And thank you for that, I get my best ideas from other peoples loud thoughts... Do you think that kind of filtering would be most usefull pre or post distortion?

Mark Hammer

You realize that most of the people who walk into a music store and try out an exciter with the normal types of guitar, settings, and amp that they would use have a hard time telling if the exciter is "on"?  These devices are traditionally for adding a degree of crispness to otherwise acoustic instruments.  The assumption is that there will be very little lower-order harmonic content, such that the higher-order harmonics generated by the circuit will be psychoacoustically attributed to the physical properties of the signal source.  The brushes on the snare will seem to swish more.  Chimes will seem to crinkle more.  Acoustic guitar strings will seem newer.

If you don't have the top end in the first place, or if your rig won't reproduce stuff you'd need a tweeter to hear, you simply won't hear it.  Anyone attempting to run humbuckers through an exciter circuit and into a 4 x 12 cab will invariably wonder why the hell they wasted money on it.

So, by all means, make yourself a blend box, but you'll want to use whatever kind of exciter circuit you decide one in conjunction with a clean signal, not a distorted one.  It'll sound good on underwound single-coils going into the board, or low-impedance bass pickups.  Beyond that, not much call for exciters in the guitar world.

erikb1971

Hey Mark

thank you for your elaborate reply. I completely agree with you. Two things might not have been completely clear from my explanation:

1. I do not intend to use this device dedicated for guitar. It will be used on a variety of sources in the mixing stage, and I would not be surprised if drums were the main use of it.
2. The concept incorporates the blending of all distortion bands with the original clean signal:


I could imagine that the harmonic sweetener (exciter, high band distortion) will be used to add just a little bit of extra on snare drums, percussion, overhead, who knows....

Cheers

Erik

Mark Hammer

Well, then this "Pope" gives you his blessings!   :icon_lol:

The use of a bipolar 15v supply renders things a little more complicated, but the Ryckebusch Harmonic Sweetener sounded pretty decent to my ears.  I'll repeat that the risk of unpleasant hiss added (it does apply lots of gain to the high end, so it works like a hiss booster if you don't watch it)m points to the benefits of some sort of simple gate applied between the "exciter" portion and the mixing node where original and added content are combined.  I suppose the absence of that source of hiss was what prompted Aphex to use the frequency-doubling approach found in the 'B' model.

erikb1971

Quote from: Mark Hammer on August 09, 2011, 08:32:06 AM
Well, then this "Pope" gives you his blessings!   :icon_lol:
And that, my dearly beloved, is just what I am doing things for! hallelujah :icon_evil:!

Quote from: Mark Hammer on August 09, 2011, 08:32:06 AM
The use of a bipolar 15v supply renders things a little more complicated, but the Ryckebusch Harmonic Sweetener sounded pretty decent to my ears.  I'll repeat that the risk of unpleasant hiss added (it does apply lots of gain to the high end, so it works like a hiss booster if you don't watch it)
I think that for all three distortion bands applies that to get real useful sounds, easydoesit is the key word

Quote from: Mark Hammer on August 09, 2011, 08:32:06 AM
m points to the benefits of some sort of simple gate applied between the "exciter" portion and the mixing node where original and added content are combined.  I suppose the absence of that source of hiss was what prompted Aphex to use the frequency-doubling approach found in the 'B' model.
I still have no clue what this gate would do....  :icon_redface:

So if anyone would bne willing to shine some light on this mysterious gate thingie... please!

Mark Hammer

Quote from: erikb1971 on August 09, 2011, 08:43:51 AM
I still have no clue what this gate would do....  :icon_redface:

So if anyone would bne willing to shine some light on this mysterious gate thingie... please!
Look here at the Boss CE-1: http://www.freeinfosociety.com/electronics/schemview.php?id=119

Coming off of the IC1 input stage, you'll see a path leading to IC3.  IC3 forms an envelope follower.  The envelope follower output feeds the gate of Q12.  When there is an audio signal, Q12 is turned off, such that there is a very high drain-source resistance formed by Q12 and the combination of R48 (2k2) and Q12 provide negligible attenuation of the delay signal.  When there is NO audio signal, the envelope follower section turns Q12 on, such that R48/Q12 form a voltage divider that quickly drops the level of the delay signal.  The strategy here is that if the delay path can be relied on to always provide a source of hiss and clock noise, then when ALL it has to offer is hiss and clock, you don't want to hear it.  When it has delay signal that masks the hiss and clock, though, you're happy to hear it.

Similarly, if one is going to goose the daylights out of that segment of the spectrum where hiss has its mailing address, you won't mind that hiss leaking through to the mixing stage (to be combined with normal signal) if there is lots of musical harmonic content in there.  But when you stop playing, and ALL it contributes at the mixing node is hiss, you'd rather gate it out.  capice?

The CE-1 does not provide any means to adjust the level of the delay signal, but the Harmonic Sweetener does provide a harmonic level control.  This is actually handy.  You'll see that the mixing stage combines the clean and harmonics signal via 10k resistors.  If you use a vactrol, with the LDR placed in parallel with that 10k resistor, you can use a very simple envelope follower that, when there is signal, lights up the LED, making the LDR resistance lower and mixing in more harmonics content.  You may have to play around with the resistor value, but it would be a simple one-op-amp mod to nicely clean up the hiss when not playing.

erikb1971

oooohhh.... why didn't you say so in the first place? :-)

Wouldn't a gate like that cause problems when processing drum loops with big dynamic differences?

The other thing that crossed my mind was: if I am mixing in distorted bands with the original anyway, wouldn't it be better to leave out the mixing stage and just add the harmonics?

Mark Hammer

Quote from: erikb1971 on August 09, 2011, 01:09:53 PM
oooohhh.... why didn't you say so in the first place? :-)

Wouldn't a gate like that cause problems when processing drum loops with big dynamic differences?

If you anticipate difficulties in discriminating between signal and noise, then you simply modify the sensitivity of the envelope detector, or its time constants, or just defeat it.  One of the nice things about acoustically generated sounds is that they have most of their harmonic content (which is what you would be using to generate additional harmonic content) right at the initial attack.  And since you would be using any exciter as a source of upper harmonic content blended in with straight signal, having the excited part gated out prematurely wouldn't affect the tail of the drum strike note strum or whatever, since the gate would not be applied to the straight signal.
QuoteThe other thing that crossed my mind was: if I am mixing in distorted bands with the original anyway, wouldn't it be better to leave out the mixing stage and just add the harmonics?
It is the mixing of straight signal and additional upper harmonic content that allows for exciters to have that "more live than live, more crisp than crisp" sound.

erikb1971

Quote from: Mark Hammer on August 09, 2011, 03:39:58 PM

It is the mixing of straight signal and additional upper harmonic content that allows for exciters to have that "more live than live, more crisp than crisp" sound.

yeah but, no but, yeah but, no but.... Iam already mixing the three bands with the original clean signal....


Mark Hammer


erikb1971

*sigh of relief*

But seriously, would i need the clean signal from the harmonic sweetener if I mix in the effect with the clean anyway?

erikb1971

Since I have never build a pedal before, and the Austin old seems a lot more complicated, could you guys in this thread help me from going from the drawing provided by Mark Hammer earlier in this thread to an actual working pcb, including the mods suggested in the drawing (dual gang pot for a adjustable filter and the pot for drive) and the earlier mentioned and finally understood gate? Or should I start a new thread for this?

erikb1971

Ok I started with putting the original drawing in eagle:



I have two questions:
1. in the original drawing, there is an arrow pointing at (in what I call in my schedule) R11. Is that a variable resistor? edit: it must be a 10k pot to mix in the distorted with the clean signal right?
2. do you guys see any flaws in this schedule?

Cheers

Erik