Boss EH-2 schematic link

[Nick123]

http://forum.musikding.de/yabbse/index.php?topic=3678.45

[Bernardduur]

Can someone with a Musikding.de membership download it and share it here?

[stm]

Wow! At last I have the EH-2. I've been looking for this schematic for ages. This is a cleaned-up version of the schem:



Now that we're talking about this, I think it is interesting to point out a clarification made by Noelch on the other forum (translated to english by Google, then rephrased and expanded by me):

1) An Exciter produces nonlinear distortions which generate harmonics not necessarily present in the original, and then adds them to the original signal, ending up with ORIGINAL_SIGNAL + SYNTHETIC_SIGNAL.  An advantage is that audio content with very poor harmonics will get brighter anyways.  A disadvantage is that some of the synthetisized signals may not be harmonically related to the original frequencies due to intermodulation products, thus producing a less natural, metallic, and/or dissonant characteristic.

An example of this kind of circuit is the Woody (http://ampage.org/hammer/files/Woody.zip). 

2) On the other hand, an Enhancer **dynamically** raises existing portions of the specturm (normally the high frequencies), thus, the output consists in ORIGINAL_SIGNAL + REINFORCED_HIGHS_THAT_ALREADY_EXISTED.  A good thing is that the reinforcement (enhancement) of the high frequencies varies with the input level, so this kind of circuit is less likely to amplify noise or hiss while the input signal is low or nonexistent.  The drawback is that if there are not enough highs in the original signals the effect will be null. (0 times anything is 0!).

The key component here is a VCA (voltage controlled ampifier). Most likely a suitable workalike could be made using other technologies such as LED+Photoresistor and JFETS/MOSFETS.

The FREQUENCY control sets the lower limit for the frequencies that will be enhanced (it sets the cutoff frequency of a second order filter).  The SENS knob controls the input level threshold above which the highs will be reinforced.  Finally, the MIX control establishes how big will the enhancement be.

One interesting thing is that the filter used to select the harmonic content is a state variable filter that can simultaneously generate LOWPASS, BANDPASS, HIPASS and NOTCH outputs.  So far the circuit could be tweaked with a couple of seletor switches to set which part of the spectrum is actually enhanced.  I think if you select the MIDS instead, it would give more punch and "funkiness" to a bass line (just guessing).  Also, selecting the notch some interesting effect would be added to a guitar.

Best regards.

[Bernardduur]

Thanks!

[Mark Hammer]

Interesting....

The distinction between "exciters" and "enhancers" is an interesting and, I think, useful one.  It also makes sense of some Japanese schems I saw that were VERY similar to this one, and puzzled me as to how there could be extra harmonic content when there was very little gain to be seen anywhere, and no apparent way of generating extra harmonics other than the "sliding T" timbre control I posted about a few weeks ago (see "Stupidly wonderful timbre control" thread).  Clearly the modus operandi is different.

Steve Giles sent me an Aphex B Exciter schem some time ago, and that approach does not use all that much gain.  Rather, it still uses a tunable 2-pole highpass to isolate the relevant signal portion, but it uses that to modulate a VCA.  As you may or may not know, if you feed a ring-modulatro with the same signal as both X and Y, what you get out is the signal plus itself and minus itself; in other words double the input frequency and zero.  This makeshift frequency doubler essentially increases the "spectral energy" of middle harmonics and upper ones, by taking lower-order harmonics (which naturally have more amplitude), doubling them, and adding them into the mix.

A related, though different, approach appears to be used here.  The M5207L is a dual VCA chip ( http://207.234.130.169/pdf/M5207L-05.PDF ), with only one section being used here.  The signal to be modulated comes in pin 4 of the M5207, which you can trace back to the hi-pass output of the state variable filter built around op-amps 4a, 6a, and 6b. IC2b takes its unaltered input from the same buffer stage as the filter does.  The signal is boosted by IC2b (gain of up to x2500), and it goes through a classic full-wave recitifier, formed around !C2a and 1b.  That, in turn, has a wee bit of lag/filtering tacked on by C10 (1uf), and feed pin 3 of the M5207.

So, essentially what happens is that upper treble is identified via the filter.  The rectifier/follower identifies when there is a transient peak in the signal, and gooses the level of that identified treble, so that it is like a treble control that "turns on" primarily when notes are struck hard or at the beginning of notes when the peaks typically occur.  The mix control sets how much of that you want in.

One of the chief differences between this "enhancer" circuit and the "exciter" circuit of the Woody (Hey, they know me in Germany!  Cool!) and the Harmonic Sweetener is that the exciter circuits are less sensitive to dynamic differences in their contribution of top end.  The drive/gain used to generate the harmonic distortion via clipping is sufficient that the extra upper treble is produced even during "softer" parts.  That has a good side and a bad one.  The good side is that if you wanted to spruce up brushes, gently swished cymbals, softly strummed acoustic guitar, etc., you could do it perhaps a little more easily (although with that gain of x2500 in the EH-2, I suppose just about anything would show up).  The down side is that since the gain is always up, so is the noise.  In the case of the EH-2 and similar devices, since thetreble content is boosted by the VCA only during the peaks, the inevitable hiss boost is masked by the signal.  Once the peak is over, the treble dies down, and so does the hiss.  In the case of the HS and Woody, once the signal goes home, the hiss stays out to play past its bedtime.

Many many thanks both Oli for posting it originally, and Sebastian for reposting it with translation.  I learned a lot.

[stm]

Mark, your additions are as usual very interesting. In particular, you saved me the work of searching for the dual VCA chip PFD !

Some additional comments:

1) An exciter would benefit from the addition of a VCA to reduce hiss when signal falls below a certain threshold.

2) An exciter generates both harmonics that are musically related to the original signal content, and harmonics that aren't musically related when two or more frequencies are present (intermodulation).  One way to reduce this phenomenon is to split the frequency band of interest into two or more smaller sub-bands (at most one octave wide), then processing each sub-band separately, filtering appropriately, and then adding the harmonics generated from each stage.

For instance, let's say you are interested in enhancing the range from 2.5 kHz to 10 kHz. Then you would use two sub-bands as follows:

a) Use a bandpass filter covering 1.25 to 2.5 kHz.  Square and limit the signal.  Filter with a 2.5 to 5 kHz bandpass.
b) Use a bandpass filter covering 2.5 to 5 kHz. Square and limit the signal. Filter with a 5 to 10 kHz bandpass.
c) Combine the outputs form a) and b).

Handling two sub-bands is within reach. If you want to cover the guitar fundamental range (roughly 80 to 1200 Hz or four octave) then you would need at least four bands. The reward would be the possibility of having a cleaner Green Ringer.

3) The described method is used in DSP (digital signal processors) implementations on some home-oriented audio equipment.  An AIWA stereo equipment I had some years before (besides the karaoke and echo additions) it had an enhancer option which did this artificial harmonic generation. It had four levels of intensity, from totally off, to mild, to strong, to aggressive enhancement.  Used with discretion it did wonders to my old cassete tape recordings that had no frequencies above 4 kHz at all (most of them recorded ten years ago on "normal" tape in a mono radio without any sort of noise reduction).

Regards.

[puretube]

"Mr Exc*ter" (http://www.pat2pdf.org/patents/pat4150253.pdf)
found some inspirations here:
http://www.pat2pdf.org/patents/pat2113976.pdf
http://www.pat2pdf.org/patents/pat2315248.pdf
http://www.pat2pdf.org/patents/pat2315249.pdf

"Frequextender"

[oli]

Hi,

is there any "one to one" replacement for the M5207L?

Regards,
Oliver