korg x911 guitar synth svc manual w/schems inside

[digi2t]

The big question is, do people think that guitar synths have hex pickups, or not. For me, I insist that my guitars be playable polyphonically while still outputting a synth waveform (ie HEX pickup).  That makes things a lot more complex.

I'm just a layman, and I only speak for myself; Guitar synth = hex pup.
Anything else is just a digital, or analog, circuit "faking" it. A synth to me is something like my GR-50. I play my guitar, and I hear synthesized strings, horns, drums, etc. An EH microsynth, Korg X911, or anything else in that vein, is just a circuit mangling the guitars original signal to sound like certain synthy sounds. Hell, a fuzz through a Blue Box sounds synthy! But, alas, in my book, a guitar synth it does not make  .

Before anyone jumps all over me, I realize that I'm treading in both fields here, but it's just that there is just a big difference in how the signal is processed (from what I understand). The X911 will never do what a real guitar synth can do, but it doesn't make it good or bad. It's just different. In my case, owning a "real" guitar synth, just makes it impossible for me to justify buying, or building this particular box.

Maybe that's why I built so many fuzz boxes, because there's no Germanium in my Digitech GNX3, or Boss GT Pro  . I love "real" wahs, because I've yet to find a digital processor that can give me "that" sound. I love analog as much as anyone, but everything has to have a purpose at some point, and some things are just better (subjectively) than others.

Sorry if I offended anyone, just my 2 bits.

[Mark Hammer]

I think it will depend on one's playing style in general.  Some players ARE largely monophonic, whether playing in the conventional guitar-into-amp way, or playing through a synth.  Others adopt a style where the harmonies or even dissonances between strings are an integral part of what they play.  Still other players are sort of in between, picking a note at a time, but incorporating open strings and making use of the overlap between notes.  And then there will be players who are generally polyphonic, but revert to mono when it comes to producing synthesized sounds.

I suspect the needs, with respect to conversion and prepping for synthesis will be different for each type of player.

[Mick Bailey]

I owned one of these and in my view it would not be worthwhile replicating for the following reasons;

1. The tracking is inaccurate and glitchy, and relies on the usual 'humbucker, neck position, tone rolled off, well damped single notes' style of playing that afflicts the Blue Box.

2. The synth voices are mainly pretty thin sounding and cheesy.

3. The total envelope duration is that of the plucked note, therefore limiting envelope and filter effects.

4. Modulation options are very limited.

5. The circuit is too complex for what it ultimately achieves.

There are many straightforward synth modules that could be pressed into service for guitar use based around 1v/oct and fully compatible. I've built many MFOS and Thomas Henry modules. The main thing is getting the triggering and P/V conversion (tricky). The rest is then pretty straightforward, as each module is no more complex than a stompbox. Still a lot of the same limitations above, but the payoff is proper fat synth sounds and way more possibilities to expand the system.

Suggest two audio oscillators (make one a Henry Thomas XR2206), a simple LFO (MFOS), a VCF (Parker-Steiner Synthacon very highly recommended) a VCA (Henry Thomas) and an EG will go a very long way and produce spectacular results way beyond the x911.

[Mark Hammer]

Thanks for that.  Very helpful.

[HarryBBD]

Thanks Mick. That kind of confirms my memory of trying (and not buying) the unit. The smaller Korg P/V unit that is pictured in this thread
has the same tracking problems. (as does the Gentle Electric unit, when applied to guitar !).

I still don't see any VCO in the Korg X-911 schematics.  Anyone want to help me out here   afaik there IS NO VCO on any of the schematics.
Maybe it was a separate Korg module or something like that

Mark is right on about the playing style vs synth technology. I need my guitar to be a guitar with a synth ~added~ so it must repond to whatever
technique I wish to play with, without a glitch. OK maybe a teeny tiny glitch would be OK. But no Octave hopping !!!

H^) harry

[Mark Hammer]

Yeah, I can't see anything that is obviously a VCO either.  There is the mystery module - the KLM-228 - which shows up in the schematic, but it doesn't seem to be situated at a point in the circuit diagram where you'd expect to see a VCO.  The service manual makes reference to some trimpots for setting up the VCO, so maybe that provides some clues.

[HarryBBD]

Ohhh I know (raises hand)

The KLM228 is just another board that ~is~ shown on the schematics, that circuit is at the bottom of another page. Its just a preamp
section, no magic there...  Someone said that the sustain is limited to the string vibration time, that could be due to the envelope follower... or they could be synthesizing sawtooth waves from the recovered square wave. To do that you need a fundamental square wave and a pitch CV. Frankly I don't see that in the schematics...  Its a real puzzle, as the mechanical print shows only the two PCBs.  There is a header that has the three sawtooth waves and a ground, this
might be an input, or an output...

H^)

[HarryBBD]

Mystery solved... the Korg X-911 table of contents shows TWO pages pf KLM229 schematic... only one is in the scan. Anyone
got the missing page

H^) harry

[wavley]

The big question is, do people think that guitar synths have hex pickups, or not. For me, I insist that my guitars be playable polyphonically while still outputting a synth waveform (ie HEX pickup).  That makes things a lot more complex.

I'm just a layman, and I only speak for myself; Guitar synth = hex pup.
Anything else is just a digital, or analog, circuit "faking" it. A synth to me is something like my GR-50. I play my guitar, and I hear synthesized strings, horns, drums, etc. An EH microsynth, Korg X911, or anything else in that vein, is just a circuit mangling the guitars original signal to sound like certain synthy sounds. Hell, a fuzz through a Blue Box sounds synthy! But, alas, in my book, a guitar synth it does not make  .

Before anyone jumps all over me, I realize that I'm treading in both fields here, but it's just that there is just a big difference in how the signal is processed (from what I understand). The X911 will never do what a real guitar synth can do, but it doesn't make it good or bad. It's just different. In my case, owning a "real" guitar synth, just makes it impossible for me to justify buying, or building this particular box.

Maybe that's why I built so many fuzz boxes, because there's no Germanium in my Digitech GNX3, or Boss GT Pro  . I love "real" wahs, because I've yet to find a digital processor that can give me "that" sound. I love analog as much as anyone, but everything has to have a purpose at some point, and some things are just better (subjectively) than others.

Sorry if I offended anyone, just my 2 bits.

Personally, I don't care to have realistic imitations of strings and horns.  I understand the allure of a "real" guitar synth, but I imagine this is just a difference in opinion as to just what a real synth is.  To me, my favorite keyboards are analog monosynths and early polysynths that don't do imitations of horns and other such things, they just sound like synths.  So personally I would love to have such a thing for a guitar even if it is a bit quirky, which I believe is the point of a project like this.  I own a few vintage keyboards, but I don't care to drag them out to shows in addition to my pedalboard not to mention that my job in the band is as a guitar player and singer.

I guess my point is, I would love to have a project where I can get some fat monosynth sounds out of a guitar.

[Mark Hammer]

Harry was over at my place for a synth/DIY "summit" about, hmmmm, I think it was 8 years ago now, and brought his guitar synth with him.  I tried it, and I also tried out the late Peter Snow's (3 years since his passing from brain cancer) Casio MIDI-equipped guitar (MG-510) with a Yamaha FB-01 module.  I ended up buying the guitar (which I had long covetted) from his widow this past year.  The MIDI/module-based system was certainly capable of a broader variety of sounds, but I have to say that Harry's system just struck me as being more responsive.  Maybe it was the difference between the rather lightweight and slender Ibanez he was using and the heavier and stiffer super-Strat style Casio, but Harry's was just livelier and had a greater sense of immediacy.

The MG-510 is slightly hampered by both the P-2-V conversion time and the time required to turn that into data and generate notes from the synth modules.  I imagine if I made a MIDI-to-CV interface that would add even more time (though clearly not as much as the original P2V).

If he's reading this, Harry can correct me if I'm wrong, but I believe his unit was in between a processor, and a true synth like the x911.  That is, it did not synthesize or generate notes via VCOs.  Rather, it used a divided G-Vox pickup, separately fuzzed each string output to generate a harmonically rich and instantaneous signal, and used envelope detection and more standard analog synth shaping techniques to produce the sonic variations, via transient generators, filters, VCAs, etc.  By NOT using P2V (again, I may be incorrect in my recollection), it was faster.  And of course, by not obliging the unit to "know" what note was being played, there were no pitch tracking issues.  It was an absolute  delight to play, and a real benchmark setup AFAIC.

[wavley]

Harry was over at my place for a synth/DIY "summit" about, hmmmm, I think it was 8 years ago now, and brought his guitar synth with him.  I tried it, and I also tried out the late Peter Snow's (3 years since his passing from brain cancer) Casio MIDI-equipped guitar (MG-510) with a Yamaha FB-01 module.  I ended up buying the guitar (which I had long covetted) from his widow this past year.  The MIDI/module-based system was certainly capable of a broader variety of sounds, but I have to say that Harry's system just struck me as being more responsive.  Maybe it was the difference between the rather lightweight and slender Ibanez he was using and the heavier and stiffer super-Strat style Casio, but Harry's was just livelier and had a greater sense of immediacy.

The MG-510 is slightly hampered by both the P-2-V conversion time and the time required to turn that into data and generate notes from the synth modules.  I imagine if I made a MIDI-to-CV interface that would add even more time (though clearly not as much as the original P2V).

If he's reading this, Harry can correct me if I'm wrong, but I believe his unit was in between a processor, and a true synth like the x911.  That is, it did not synthesize or generate notes via VCOs.  Rather, it used a divided G-Vox pickup, separately fuzzed each string output to generate a harmonically rich and instantaneous signal, and used envelope detection and more standard analog synth shaping techniques to produce the sonic variations, via transient generators, filters, VCAs, etc.  By NOT using P2V (again, I may be incorrect in my recollection), it was faster.  And of course, by not obliging the unit to "know" what note was being played, there were no pitch tracking issues.  It was an absolute  delight to play, and a real benchmark setup AFAIC.

Harry's synth sounds pretty danged interesting, I don't necessarily need to trigger vcos, just something that sounds synth-like, polyphony is obviously a plus but not a necessity for me.  I have a Casio MG-510 that I've recently re-capped and it still didn't fix the dead D string (corroded caps is a big problem with these things) so I break it out every so often and tinker with it but I have yet to use it live until I get it running reliably, I have messed around with using a Juno 106 as a sound module with it, but the guitar feels so different than the Jags, Mustangs, and SG-I I use that I doubt I'll ever make it a mainstay instrument, though I have seen JJ Cale playing one quite a bit.

edit: Sorry, I just realized, I have a PG-380, not an MG-510

[HarryBBD]

My guitar processor "Muffy" can be seen at
http://www.wiseguysynth.com/larry/guitar.htm
It is essentially a hex fuzz outputting quasi-sawtooth, square, pulse, and rectified pulse (gives a sort of octave up effect). It has six envelope followers
controlling six 12dB VCFs with resonance. It has six VCA's which can act as either gates of be controlled by the envelope, or bypassed. Obviously it cannot
sustain more than the string itself does.  There is an attack-delay circuit for each envelope follower.  There are no 'unobtanium' chips used.

If I were doing it again, I'd make a voltage controlled ADS generator (you really can't have 'release' but what I've designed is a unit that shares the decay and release times. Careful setting of a trigger level could give access to all sorts of percussive sounds. 

The actual P/V converter I added is about as complex as one of these boards, and that just gets you three strings monophonic. The best way to really get guitar synth imho would be a board per string, with P/V conversion, direct waveform synthesis from the guitar string, and an envelope follower / ADS generator, VCF and VCA. I can generate octave down square/sawtooth, fundamental square and sawtooth, (ready for this one ?) fifth up sawtooth, and octave up sawtooth from the guitar string directly.  I'd include a VCO with saw, triangle, and pulse waves that can be hard sync'd to the guitar waveform (this eliminates any pitch tracking delays, instead you get a waveform distortion which is not very noticible... or you can have the VCO and guitar wave ~not~ sync'd, that allows the direct guitar wave to cover up the inevitable delay in P/V tracking.  My P/V converter gets an output within one cycle of the fundamental so its maybe 2X faster than most commercial units... but it is complex. I also have a circuit that eliminates the droop when you lift your finger from a string. This makes the note go flat as a rule, as your finger does not damp the string in time to prevent converting that now slightly longer string length.  P/V conversion is kind of nasty about these things...

Downside, connecting six boards to a voltage controlled front panel and putting it in a box is not cheap. The overall complexity would probably be about the same as an all-analog polysynth, without the neat SSM or CEM chips to do the functions. Probably it would be like an olf Oberheim 8 voice polystnth (made from SEM modules) in complexity. NOT for the faint of heart, eh   You could buy a Roland V-guitar for the price. Or any vintage guitar synth, but at least this one would work...

Maybe I'll get the camera out a photo what I have now...

H^) harry

[wavley]

Honestly I've been thinking the best way for me to go is to just get a hex pickup and hex to midi converter, then take my parts Juno-106 (the keyboard is trashed) and chop it into a sound module leaving me a good number of keys in case I break my good Juno.  Of course, I've been planning to do this ever since I got my Casio... one of these days I can spend more time on my guitar and less time fixing my converted corn crib house.

[Mark Hammer]

edit: Sorry, I just realized, I have a PG-380, not an MG-510

I get them confused too.  The PG-380 has the voices on-board, where the MG-510 is simply a MIDI controller. I would imagine they're pretty much the same guitar in terms of playability, though, and I can't imagine the difference in their electronics has much of an impact on their respective weight.

I have a busted 106 too! 

[soffa]

I found the missing page from the X911 service manual:

http://img.photobucket.com/albums/v368/zero_thehero/schematics/X911-09.jpg

[pinkjimiphoton]

nice find! thanks mate!

[snufkin]

The X911 tracks a lil better than everyone here seems to remember however as stated it has serious limitations

I have one and also a penfold pitch to voltage converter board partially populated for over 3 years 

If you guys were going to create a forum based guitar synth project while P2V units are cool I think that a hexaphonic processor would be more interesting
you could spec it for a pickup from here http://www.ubertar.com/hexaphonic/

then a processing stage and envelope detector in to --> a fuzz/waveshaper --> in to a VCF and a VCA in to a mixer
If each were designed with tim escobedo style economy the parts count could be fairly low even though it would have to be done 6 times !

basically a super fast responsive analogue hex fuzz and filter would offer something that cheap pich to midi (Sonuss) and hex to midi (axon/roland) systems don't
so it would be very playable and worth building even though it would be a pricier than usual project

a good bench mark for processing would be the gr300 as it is held up as the fastest and most playable guitar synth in the west
  http://www.joness.com/gr300/patent.htm

the obvious downside of this proposal is that everyone who participates would need a cheap guitar to mod with a hex pu

[Jaicen_solo]

On my list of projects to get around to is indeed an analogue hex fuzz with VCA and VCF. Something similar to a GR100, but compatible with current 13-pin synths.
However, I think there is a market for an all analogue, P2V synth. Personally, i'd like a last note priority P2V, dual oscillators + sub osc, with hex fuzz and analogue VCA/VCF. If it was based on the GR300, it should be fast as well as accurate, but i'd take fast over accurate.

[pinkjimiphoton]

i agree, the 13 pin deal would be probably the easiest. i've got a couple of those, but the necessary skill to build something like that at this point elludes me completely.

i bought one of them cheap "guitar controllers" for xbox, so i could scavenge the hex pickup out of it if the time ever comes...power rock? power gig?

i figure after i harvest the hex pup, i can mount a humbucker in there and just use it as a lap steel.

[Mick Bailey]

One seriously limiting factor with the Korg is that it isn't what you expect a mono synth to be. The normal controls you would expect from a synth for true subtractive synthesis are missing and it relies heavily on an envelope follower to add some interest to a rather dreary set of presets. You quickly lose interest in the fake flute and brass sounds and limited palette. Even the most basic single oscillator mono synth (such as the MS-10 or Jen SX1000) has more dimension.

Given Korg's history with synths, it's disappointing that they didn't build something more flexible. It has more in common with the Micro Preset than (say) the MS-20.

I still think that there's a lot of mileage to be had out of fundamental extraction, even if it is a little glitchy. I've got 14 modules in my modular synth and they were all pretty straightforward to build. It would be really nice to use a guitar to control it - the note would need to be sampled and held to release the full drama of a slow filter sweep on a PWM waveform - a guitar's natural note length is way too short. S&H is simple once a sucessful F to V is achieved. Perhaps going down the PIC route may yield something reasonable?

A simple 'starter' synth can be built around the SN76477 - see the SN-Synth project and sound clip.