GRAPHTECH GHOST/HEXPANDER SYSTEM
|Simplified switching diagram approximation, done entirely from my failing memory. It seems to work this way though, more or less. Obvious stuff was left out, along with pin references to the circuitboards.|
Switch 2- Synth/Both/Guitar Quickswitch After playing for a while I realized that the Synth/Both/Guitar Quickswitch was actually a good idea, since there are times when you want to quickly switch between the pure sounds of each. A well-placed toggle switch would make this less cumbersome than the external GK-2A's switch, and therefore, more useable. I located this Quickswitch where the S1/S2 switch had been and drilled a second hole through the scratchplate below that to relocate the S1/S2 switch. This gives enough space so that neither switch blocks the fingers while operating the pots or switches; they're pretty easy to find and operate by feel. The piezo pickup extension cable (The one the cat got...) gave me the connectors I needed to install this, and I happened to have an old Radio Shack 3-position (center off) switch with two sets of poles.
The switch works the same way it does on the GK-2A pickup: The center position is connected to ground, and in that position, no contact is made with either of the other two poles. This gives you the blended sound of both synth and guitar. In the other two positions, the switch grounds the respective signals, cutting them from the outgoing signal. (This is vaguely similar to the way the S1/S2 switch works, except in that case they're control voltages and the contact between the grounded center position and the other two positions is momentary or continuous; the GR-33 circuitry deciphers the meaning.)
Switch 3- S1/S2 Quickswitch This was discussed previously: It's a momentary contact three-position switch. I attempted to "debounce" S1/S2-- the switch caused a fairly loud pop when when toggling in one of the two directions. The idea is to put a resistor and capacitor in parallel to the ground so that there's a slight time delay to even out the contact surge. I forget which values I settled on (I think I used the original tone cap + a 22K? resistor?) and while it doesn't kill the pop entirely, it does reduce it. It doesn't have any effect on the sound since the switch is handling a control voltage, not an audio signal.
Pot 1- Magnetic Volume No changes here; the standard 250K Strat volume pot to control the output from the pickup selector switch.
Pot 2 (Push-Pull)- Piezo Volume/Preamp Bypass Since I had the push-pull pot, I decided that I might as well install it. For reasons explained earlier, I decided not to use it for the "Mid/Dark" feature. However, it is useful as a Ghost preamp bypass switch, so that the guitar could be used totally unpowered-- just like a regular guitar. Unless you install a bypass switch, the unpowered guitar's magnetic signal at the 1/4" jack will be very weak and thin-- Frankly I'm amazed that there's a signal at all. My bottom line is that the guitar must be able to function as a standard electric guitar. The push-pull pot is a modal switch-- it's not one that you can easily toggle while playing. For me, this works perfectly since I consider the Piezo output to be a specialty mode.
As I suspected, the push-pull pot eats up a lot of valuable overhead space in the control cavity; Shoehorning the guts back into the guitar and sealing it up was challenging, but eventually everything fit without complaining electronically and without any badnews scratchplate bulges.
Pot 3 (Switched)- Hexpander Volume/Battery Discussed previously; controls the Hexpander volume by sending a control voltage out the 13-pin jack. The switch controls the battery connection and is easier than plugging/unplugging the 1/4" jack to do this.
For what it's worth, I later came to the conclusion that the Hexpander Volume pot wasn't absolutely necessary-- I prefer controlling the Synthesizer volume with a footpedal, especially for the organ patches. Running those patches through a Hughes & Kettner Rotosphere Leslie speaker simulator is really kewl. This would allow you to retain a traditional treble cut tone control... (is the job ever done?).
Unpowered: Plug into the 1/4" jack and get the magnetic pickups. The main controls are magnetic pickup volume and pickup selector. The synth/guitar selector is also active and turns off the magnetic pickups because I wired the 2nd set of poles to do that. Engaging the preamp circuit at the push-pull pot will attenuate the magnetic pickup output because there's no power going to the circuit.
Battery: Plug into the 1/4" jack and get the magnetic pickups, or switch on the battery power, and bring the preamp in circuit and get magnetic and piezo pickups. Active controls are magnetic pickup volume, piezo volume, piezo preamp in/out (bypass push-pull switch at piezo pot), battery in/out (at Hexpander volume), and synth/guitar selector (turns off magnetic pickup, or mag+piezo)
13-pin plug: Since the 13-pin system provides power, you might as well remove the battery so you don't unintentionally or needlessly drain it (I suppose that if you cut the 13-pin power by turning off the synthesizer, the battery will try to power both the Hexpander and Preamp circuit.) This gives all the above-mentioned possibilities, in addition to Hexpander and composite magnetic/piezo output at the 13-pin jack. All controls work as above, plus the S1/S2 switch, the Hexpander Volume and Synth/guitar selector (enables/cuts synth).
One thing to be aware of when reconfiguring plugs is that the concept of solderless connections has a downside; what goes on easily can often come off as easily. It gets worse when you remove redundant ground connections. Reducing a 12-contact plug down to 6 contacts removes 6 points which the connector was using to snuggle itself onto the terminal block. With 6 pins, it's not a problem, but if you reduce a 2 pin connector down to 1, or even with a 2 pin connector, the bend of a wire may be enough to coax a connector to disengage. Because of this, I used the extension cable's second 6x2 connector to consolidate the individual pickup connectors; They now plug as a single unit onto the circuitboard.
I was hoping to solve a problem I hadn't noticed in my initial testing: Even with the volume pots set at zero, there's still an audible signal coming out of 13-pin connector when the amp is highly cranked. This was noticible in both the synthesizer (faintly) with certain patches ("Tekno Taurus"), but also with the acoustic guitar sound (more obvious). The output thru the 1/4" jack doesn't have this problem: the magnetic and acoustic signals are completely cut by their volume pots.
My first thought was that it might have something to do with the guessed value of the Hexpander pot, so I substituted a 250K pot-- this made no difference. I then bypassed the pot altogether and shorted the hexpander volume connection to ground-- no difference either. That should have completely killed the sound. This ruled out the hexpander pot, which makes sense-- the piezo pot wouldn't have been affected by that anyway. I doublechecked to make sure that the circuitboards were getting their reference to ground. Interestingly, grounding the signals at the synth/both/guitar Quickswitch kills the composite acoustic/magnetic signal at the 13-pin jack, but still leaves a tiny bit of the synthesizer coming through. The 2nd optional guitar Quickswitch (magnetic/both/acoustic) doesn't completely kill the piezo-acoustic signal going to the 13-pin jack.
What a puzzle! To figure out why this is happening, one has to look at the symptoms and try to venture a guess as to how the circuit generally works. It's complicated by the two different outputs and what Graphtech calls an "intelligent" circuitry... whatever TF that is.
It appears that the piezo pickups outputs are processed separately by the preamp and hexpander circuitboards. The acoustic preamp circuit combines the six piezo pickup signals into one, processes it, and runs that signal out to the 250k pot (with an optional Mid/Dark output). The magnetic pickups signal comes into the preamp with the level set by the guitar's passive volume pot; the piezo/both/magnetic quickswitch allows selection between these two signals by grounding one or the other, or neither. The separate signals are available at the 1/4" jack depending on whether a stereo or mono plug is inserted: my guess is that if the ground and ring voltages match, the signal is blended and sent out to the jack's tip connection: If they don't match, the piezo signal is sent out to the jack's ring connection.
The Hexpander needs the six separate signals to do its filtering thing before passing the separate signals out the 13-pin jack (the GR-33 does the actual analog-to-digital conversion). The Hexpander volume pot sends the control voltage to the synth, where the actual synth volume control takes place. I assume that the Synth/Both/Guitar Quickswitch in 'Guitar' position simply overrides the pot and sends a zero volume voltage to the synth. The Hexpander circuit probabaly gets its piezo-acoustic output from its own circuitry-- without the preamp, the piezo-acoustic signal is passed directly to the 13-pin output. With the acoustic preamp installed, it probably taps the preamp circuit for the magnetic and acoustic signals since it sends a post-pot blended and post-piezo/both/magnetic Quickswitch signal out the 13-pin jack. The Synth/Both/Guitar Quickswitch in 'Synth' position probably grounds the output of that composite signal (which is why it can be cut off completely).
From the Japanese GK-2A website, the author has deduced these functions for the 13-pin jack:
This suggested some possibilities for further testing-- that is, if the info was accurate for the Graphtech system. My wild-assed guess was that the two circuits might operate with different ground reference points. I'd originally assumed that the ground was common to the whole shebang and wired the switches and pots accordingly. After I discovered this problem, I'd taken the guitar apart numerous times, redoing the ground connections, trying to determine whether there were two different sets of ground voltages. I scrutinized the circuitboard to see if all the ground pins were actually all tied together, or whether some might have a more complicated route to ground. Nothing seemed to solve the problem.
On the other hand, maybe there's something wacky going on in my synthesizer? That's always a possibility when you buy used gear. (I've already had to dig into the synthesizer to reflow solder on one of the intermittent output jacks-- and curse the fact that modern production efficiency demands that controls and jacks be soldered directly to circuitboards instead of being wired. Wired connections buffer the strain that usage puts on soldered connections since wire has some "give".)
As a matter of fact, I finally did what common sense should have told me to do at the beginning: Test the original GK-2A pickup. Even though that system doesn't have the piezo-acoustic sound, the synth output should tell me whether the problem was in the guitar or the synthesizer. Sure enough, setting the wart's Synth volume at zero didn't completely cut off the "Techno Taurus" patch at a highly cranked volume either. Duh. The problem was in the synthesizer. I'd wasted about two weeks of testing, going after the wrong culprit. It wasn't totally wasted time since it forced me to test and think about how the system might work. And it gave me some wiring ideas. That's the reason why I've gone into it with such mind-numbing detail.
Even if I don't fix the synthesizer, it's not a intractable problem. The sound level bleeding through is very low on the synthesizer side and it's dependent on the patch; the residual piezo-acoustic signal is more noticible, especially with the high-gain guitar toys later in that signal path. Fortunately, that signal (composite acoustic/magnetic) can be completely cut off with the Quickswitch (same as with the GK-2A). But there are times when you want to ramp down the guitar volume, not just cut it off. If you can't ramp down to zero, that's a pretty serious problem. One quick and easy solution would be to use a switching or blend pedal at the outputs of the synthesizer. Killing those signals before they got to the amps would be a surefire thing.
Before I realized that it was the synthesizer's problem, my inelegant solution was to use the 13-pin output only for the synthesizer; the guitar signals are plucked off the 1/4" output, where the volume controls can kill the signals completely. In this dual-output configuration, the Quickswitch has no effect on the 1/4" jack output, so it had to be adapted-- fortunately, it was a double-pole switch, so the other half was used to ground the magnetic pickup. This means that switching to the synth-only position kills the mixed guitar signal at the 13-pin output and also kills the magnetic signal going to the 1/4" jack. The piezo acoustic signal is unaffected, but could be turned completely off at the pot-- as I said, the piezo pot works in the 1/4" output. I figured that it was far more useful to completely switch between the magnetic pickups and the synth output: The piezo-acoustic sound is a specialty mode that doesn't work well with the usual high-gain gizmos in my guitar's signal path, so it can stay turned off at the pot most of the time.