curtisa

Hello all, A while back I bought a pre-slotted, pre-radiused birdseye maple fretboard fitted with black dot markers, with the intention of building a 7 string. As these things always happen, time and focus got diverted until recently when I started getting interested in guitar building again. So I went through my stocks of timber and pulled out this fretboard to have a closer look at it (it was still in packaging since I bought it). When I took off the packaging I got the vague impression that the slots looked a bit off-square, but I wasn't sure if it was the grain of the wood, or the edges of the timber playing tricks on my eyes, so I spent a bit of time doing some measurements on it, and lo and behold the slots are indeed out of square. If I draw a centreline along the length of the board up through the middle of the position dots and lay a protractor on the board the slots appear to have slightly less than 1 degree skew towards the bass side of the nut. As far as I can tell I have a few options, but I wouldn't mind finding out if anyone has a particular preference as to where to from here: 1. Build the neck as-is with the <1deg skewed slots. Fit the nut with a matching tilt and intonate the difference at the bridge. Visually I might be able to get away with it. It may be the way the raw fretboard has been cut is exacerbating the skew of the slots, making it look more noticable at the moment than it actually will be once the board is tapered. 2. Deliberately re-taper the fretboard to correct the skewed slots. There is room on the board to do this for a 7 string neck. Only drawback I can see is that the dot markers will obviously appear to drift towards the bass side the closer they get to the nut. By my measurements the "D" string will pass through the lower 3/4 of the 3rd fret dot and be perfectly aligned at the 19th fret. Perhaps the other alternative here is to re-taper and shift the centrline upwards so that the visual offset of the dots is spread out across the full length of the board (-1mm @ 3rd fret, + 1mm @ 19th)? 3. If (2) is an option, re-route all the position markers into something different that masks the drift of the dots (eg ovals, Les Paul block inlays, Ibby Universe Pyramids etc). A lot of work though... 4. Suck it up and bin the board. Seems a shame to waste what is a quite nice birdseye board, but maybe I just need to bite the bullet and admit that it's not worth the effort? Thoughts?

Hi Chaps, Been out of the guitar DIY loop for some time now, but recently got my interest rekindled when I stumbled on PVX Guitars. I realise that it may be viewed as a bit of a cop-out, but as someone who isn't entirely confident with the more precise woodworking skills required to turn out a quality instrument a kit guitar seems like a pretty good place to start out. With all the "serious" woodwork done already I'm pretty sure I could handle the assembly, bridge location/installation, finishing etc Anyone have any experience with PVX Guitars, or know of anyone who's built one of their body/neck kits?

Looking closer at the plots of the SM57 and SM58, the reason why the 58 works better on vocals than the 57 does is due to the earlier rise in the upper-mid response. The 58 has a circa 3dB rise at 3K, which is exactly where most people perceive "presence" in a spoken voice. The dip at 7K will help tame sibilance on close-miked vocals without ruining overall clarity in the vocal sound. That said, a person who sings with a particularly strident voice may not need the extra upper-mid help that the SM58 provides, so the SM57 might work better in that application. The same could be said for any miking situation - the combination of the source and the mike (and the position, and the room, and the preamp, and the...) will determine how the result sounds. I'd suggest borrowing some mikes to experiment with, both condensor and dynamic - it's really down to your own personal preference and what you think works best for your source material.

I prefer the Audix i5 on electric guitars to the SM57. You will need to know the basic differences between dynamic and condensor to a degree. As a very general rule of thumb dynamic mikes will handle higher sound pressure levels without destruction, but have a more limited frequency response. Condensor mikes are more fragile, won't take punishment as easily, and require phantom power from the mike preamp, but have more sensitivity to lower sound pressure levels and a broader frequency response. Some condensor mikes will run on batteries, so that can be used to your advantage if your mike preamp doesn't provide phantom power. Because of their robustness dynamics are ideally suited to close-miked drums and cranked guitar cabs. Condensors, with their extended high frequency response are ideal for acoustic instruments, cymbals and vocals. If you can stretch your budget a little more, theres also the Shure SM-7. It's a dynamic mike, but a lot of people favour it on vocals aswell. Devin Townsend uses it a lot on his albums, and James Hetfield can be seen using one on the Metallica documentary that came out a couple of years back.

Yep, that'll do nicely. Be prepared for it not to sound too special though for the afore mentioned reasons.

Using the pre out may work OK. However, there are a few caveats: It won't sound like a guitar amp - you'll lose the natural high frequency roll-off that the speaker introduces, so it will sound very bright and harsh in the headphones, especially with distortion. The preamp out may not have enough drive capability to push a set of headphones, which may introduce problems of it's own in terms of further tonal ugliness, distortion or (worst case) destruction of the preamp out stage through continued overloading of the poor little chip feeding the headphones. If you can feed the preamp output into something thats designed to drive a set of headphones (eg stereo, mixer) it'd be better. Using a switch to disconnect the speaker in a solid state amp should be fine, just make sure that anything you put in there doesn't have a chance to short onto any surrounding metallic parts or circuitry. Although you may find that plugging something into the preamp output automatically interrupts the signal on the way to the power amp until you complete the external loop into the "pwr in" socket. Try plugging an un-terminated lead into the "pre out" or "pwr in" sockets and see if the speaker cuts off.

The idea with avoiding ground loops is to avoid creating more than one path between two earthed parts back to ground. Star grounding is the surest method to avoiding loops. The shortest, most direct path for each component is also good practice. In the case of shielded wire it can mean only connecting the shield at one end, but it depends on the application. For example if you had a pickup with a shielded cable, the shield will already be connected at the pickup end. Earthing the shield on the cables at the pickup-selector-switch end will be fine unless the shielded components of the pickup itself come in contact with another point connected to earth (eg conductive paint, earthed pickup covers etc). In this case you've created a loop between ground lug on output jack/pickup cavity shielded paint/pickup/pickup cable shield/earth connection on switch/ground lug on output jack, and the loop needs to be broken somehow. A little bit of exploration of your guitar will reveal what's earthed and what's not, and using that info will help you build a "map" as to the best way to run your earths in the guitar. For example, you may find that the shield on the pickup cables does not connect to anything on the pickup that will come in contact with an earth inside the pickup cavity when installed (eg plastic encapsulated single coils), in which case you can safely earth the shield of the cable to a common point inside the control cavity without creating a loop within that section of wiring. The ground lug on the volume pot can also be connected to this common point safely aswell, as can the ground lug on the output jack. A shielded cable between the volume pot and the output jack cannot however, as you will then be creating a loop between the cable shield/ground lug of output jack/ground wire of output jack/volume pot ground lug/cable shield. In this case the loop can be broken by either ditching the earth wire between the output jack and common point (my preference), or lifting the shield of the cable at one end only. Errr, dunno if that's helped things or confused them further!

A week is possibly a little bit extreme. The tube amps I've seen only have a few hundred uFarads of capacitance per voltage supply, which will dissipate fairly quickly after turn-off. Probably all you really need to do is unplug the amp, leave the main power switch turned off, and turn the standby switch on. This will connect the main filter caps to the amp circuitry to speed up the discharging process. I would imagine they would be discharged to a safe level within half an hour, but always check with a multimeter to be sure. Making up a little "cap discharger" tool is a good idea - just solder some crocodile clips to a 10K 1W resistor and cover the exposed solder joints/wire with some heatshrink or electrical tape. Clip the resistor across the caps with the amp turned off and the amp is safe to work on straight away - $1 insurance.

Not unless you have the space for a lot of them. Just had a quick look at the Whammy manual on the Digitech website, and it says that it requires at least 9.7V @ 820mA to operate. Ignoring the fact that the required input voltage is higher than a normal 9V battery, you'd probably need 3 or 4 in parallel to last more than an hour or so of continuous use.

If the Whammy's control pedal is built anything similar to my old Digitech RP6 control pedal (which I suspect it is), it doesn't use a pot, it uses an LED and photo-transistor combination - you're not going to be able to adapt a pot to this circuit for internal guitar use without some serious circuit surgery. As a bare minimum you're going to need a schematic for the Whammy to find out what's what, something I'm not sure you're going to easily track down. Probably more trouble than it's worth, and more than likely extra routing required to fit what must be a fairly substantial PCB.

Anything that looks like it's about to fall off - wiggle the wires and see if they stay put (looks can be deceiving) - or a solder joint that has a fractured look, or dry crispy look around its edges.

First thing I'd suggest doing is opening the pedal up and looking for any wires that have broken free - maybe the battery clip, the input/output sockets, the footswitch, or the pots. I've had Behringer mixers that had really bad hum that was tracked down to a broken wire on the power supply connector.

Yeah, weather was crazy for a bit there, wasn't it? Hobart and Melbourne airports were shut down because of high winds, and a truck crossing the Tasman bridge had a shipping container blown off the back of it onto the road! Was chatting to one of the electricity distribution operators today (I work in the power industry) who said they were flat out servicing customers with blackouts for a period of about 36 hours. Progress has slowed a little now that I'm back at work from my time off. However I have drawn up a PCB layout ready for etching of a class-D driver. It's all surface mount components, but not the really tiny stuff. Everything should be workable with a regular soldering iron and a bit of patience. Board size is 80mm x 55mm, a bit bigger than a credit card. I can get it a little bit smaller still but at the expense of height (I'm using right-angle plug-in headers for the external connections to the circuit. Using vertical headers instead will make the board a bit smaller, but the connectors will stick straight up requiring a deeper cavity inside the guitar.

Update from me: All it needs now is a pickup cover and a pair of core "endcaps". The black wire at the top of the photo is the connection to the copper shield surrounding the two cores. I ended up rewinding the two cores using 0.2mm wire, which enabled me to fit 256 turns on each bobbin (over 100 more than the last attempt), plus plenty of spare room for the self-amalgamating tape and copper foil. The cores are connected in parallel-antiphase, the total resistance being 12 ohms, a little higher than I had hoped but still within shouting distance of usable (my first sustainer came in at 13.6 ohms and works fine with the old LM386 circuit, so I don't think an extra 4 ohms is going to cause too much trouble). The baseplate is made from two layers of 1.5mm thick black plastic I had lying around, superglued together, and cut to shape. Two rectangular windows (25mm x 11mm) are cut into the baseplate for the two magnets to sit into, and a U-section of plastic is fitted underneath the magnets to stop them falling out the bottom of the driver. The U-section also enables me to slide a piece of steel underneath the magnets to experiment with focusing the magnetic field more upwards, a la Col's FEMM plot above. There are actually two layers of copper foil - a narrower layer (6mm) that wraps around the two bobbins that fits within the upper and lower flanges of the bobbins, and a second full-height layer (10mm - the one that's visible in the photo) that is soldered to the inner layer so that it doesn't slide off. I'm really pleased with the way this has turned out this time. Given how much longer it's taken me to build it, the damn thing had better work or I'll be pissed off!

I wouldn't have thought an extra 1mm of plastic was going to make much difference? Anyway, the idea is that I will contstruct a pair of steel "endplates" that sit on top of the laminations of each coil, flush (or even slightly proud) with the surface of the pickup cover. I'd imagine the proximity of the steel to the strings would have more of an effect than the poximity of the windings. I couldn't find anything locally at all! I could find bags of craft shops with magnets, but they were all the wrong size, shape, north-south orientation...In the end I bit the bullet and went with Aussie Magnets. A bit more expensive than I was hoping, but I was getting desperate, and in the end I didn't give a hoot where the magnets came from, just as long as I could get the right type. @Col: What happens to the pattern/intensity of the field if the steel bar at the bottom of the sustainer is removed? I'm just putting the finishing touches on the magnet bracket/baseplate for my sustainer and I've installed a channel underneath the magnets where I can insert a similar steel bar to see what difference it makes.

Update from me: Haven't been able to do much more on the sustainer over the last week-and-a-half as I've been busting my gut building natural stone walls in the garden. However I've finally found a pair of Ferrite/Ceramic magnets that fit perfectly under a single coil pickup. Having finally got my hands on the magnets I got all re-inspired and started working on the driver again - I've built up a baseplate that holds the magnets in position and supports the two coils, and will eventually provide the connection points for the cable exiting the driver assembly, in much the same way a normal single coil pickup construction does. I've built it out of laminated plastic (cutting the holes for the two rectangular magnets removes a lot of plastic, so it needs extra strength), and I'm just waiting for the glue to dry. I still need to get a black single coil pickup cover to hide the coils, but that shouldn't be too hard to source. It's looking like I'll have to rewind the coils using thinner wire as I can't fit the bobbins under a single coil cover (too wide), but with the apparent preferred arrangement of parallel-connected coils, I'd probably have to rewind my slightly-too-low-impedance coils anyway. Been having a closer look at the sustainiac amp and thinking that it might be a pretty good design to fiddle with - the driver actually sits within the negative feedback path of the amp which none of the current solutions do (LM386, linear BTL, class D etc). In our "traditional" approaches the amp doesn't really "know" what the driver is doing because the load is connected outside the negative feedback loop. The sustainiac patent puts the driver within the negative feedback path, which may help further iron out any non-linearities in the amp/load combination. More later...

Still no real progress on the driver itself yet. Still waiting to find some magnets. Any thoughts on the small rare earth disc magnets that can be bought from Jaycar? I was thinking perhaps I could install one or two under each coil, and the circular shape of the magnets makes construction of retaining brackets much easier. This afternoon I breadboarded up the driver circuitry from the sustainiac patent, just to see how it works. I don't have the appropriate MOSFET output devices, so I only made up the part of the amp shown on page 18 based around U3b (the HF oscillator), U3C/U3d (the MOSFET switch drivers), U5a (gate inverter) and U1b (driver input stage). The PWM action of this amp does work as is, although I've found that you need to take care that the input signal doesn't become too large otherwise you end up over-modulating the PWM signal, causing the outputs to latch up. Probably another reason to incorporate an AGC element in the design - A class D amp doesn't clip in the same way a linear LM386 does when overdriven. U5a, C13 and R16 perform the anti-crossover functions that prevent the two output devices from being switched on at the same time, safeguarding against the destruction of the output MOSFET's by shorting out the supply. All it does is create an extra delay in the switching waveform for the upper MOSFET element so that is cut off before the switching of the lower MOSFET element commences. Measured current draw with no input signal when running (but obviously with no load connected) was 4mA @ 9V. Switching frequency was actually closer to 28KHz with the components given, but near enough is close enough. The switching frequency can be increased by reducing the value of C12, although I don't really know if there's any benefits/drawbacks from doing so with this particular design.

I knew there had been some discussion about a TDA7-something-or-another some time back, but I couldn't remember what the part number was exactly. Both the TDA7052A and B both have the DC volume port, it's the TDA7052 (no suffix) that doesn't have it. The TDA7052A is available from Farnell.

The coils have been potted in PVA. I was using a small paint brush to apply a layer of glue to the coil every 20 turns or so. The photo you have there is the last one I took before I applied the final coat of PVA and selaed it up using self-amalgamating silicone tape. If you look closely at the photo you can see the white of the previous layer of PVA just showing through between the copper wire. With the combination of thicker wire and shorter winding length, the whole coil feels more "solid" anyway, so I'm expecting the driver to be less prone to vibration and microphony. The windings on the full-length bobbin on my previous driver felt very loose as it was difficult to get the copper wire tight on the longest run of the bobbin (from bass to treble). True, but with the class-D amp, I'm already running at full voltage (switching hard positive to hard negative), so the pulsewidth of the switching action will then define the amplitude of the average driver current. Time and trials will tell, but I reckon the class D chip will be quite happy driving a 2x2 ohm series inductive load. I have two TPA2000D1 2W class D chips on their way to me for trialling in the next week or so, so I'll be able to have a play with the new system soon. The chip is still an SMD device, but it has a slightly wider pin spacing than most others I've been able to find, so it should be just easy enough to manually solder. Failing that I can get the dual MOSFET driver chip used in the sustainiac patent for around 50 cents a pop, and try building the amp that they use in the patent. @Pete: I had a look at the datasheet for the NJM2073 you suggested. The mute scheme they show seems a bit crude, forcing the output stages into cutoff - could be prone to pops when engaged? I found a possibly better solution in the TDA7052B, a 1W BTL chip with DC volume control in an 8-pin DIL package. The DC volume port is handy because you can implement the mute function, AGC (if necessary) and drive level all in the one pin. Could be quite handy if you're looking to save space while maximising flexibility and expandability of the system.

Thanks for the kind words on the driver, everyone. Still searching for appropriate magnets. Couldn't find anything locally. Been looking at the Aussie Magnets website. Having a bit of trouble working out how they determine the magnetic direction of each magnet. I was considering getting a couple of different types for comparison - maybe a pair of ferrite magnets and a pair of Alnico ones. I need one magnet for each coil, and the magnetic direction needs to be North on the biggest flat face of the magnet to South on the opposite side. Yep, I was actually intending to remove the 5-way switch in the HSH S470 and go with a 3-way "John Pettrucci" switching scheme - with a mid driver fitted the guitar will be reduced to a HH configuration anyway. I modified my RG7620 from the fancy pants 5-way HH configuration down to the JP 3-way scheme years back, and I much prefer it to the way it was originally fitted out. Thanks Pete, I'll have a closer look at it tonight. 28AWG (0.32mm). Thicker wire than the original 0.25mm specced by Pete, but apparently the same guage as used by Sustainiac in their stealth system. Could only fit on 140 turns on each coil, and have yet to measure the DC resistance, but if it turns out to be a bit low I was thinking I could connect them in anti-phase series (like a regular humbucker) and still get the same effect. If it turns out to be no good with 140 turns I can easily strip it all out and rewind it with more turns of thinner stuff. Those two coils only took me about 15 minutes each to wind and tape.

Quote (Pete): Hmmm...yes, not normally a problem, but when my circuit is on, the bridge pickup may well be buffered by the preamp and so effectively be active. Is it possible that this could add to noise. This is why I was wondering whether a buffer on the whole guitar would help in some way...or if noise was created by the switching off of this buffer, even if no gain is apparent when it is in operation...any thoughts? No, I don't think in your application it's necessary. It's required in the sustainiac patent diagrams because they're buffering both neck and bridge PU's, using active FET switching, and when the sustainer is running you have the option of choosing the bridge and neck pickups together - there has to be a way of introducing some degree of isolation between the two active PU systems, and the 22K resistors do this nicely. In your case the idea is to only use the bridge PU as the active pickup while the sustainer is running, and disconnect any others from the system. With only 1 pickup active during sustainer operation there's no need to prevent cross feed, and hence no need to do resistive mixing. Quote (Pete):Ok...so are positions 2 and 4 on a fiveway (neck/middle and middle/bridge) now N+B as well? I am guessing so. These combination positions have been causing me some problems and limiting what I can do with 5TDP selector passively. I don't think the sustainer patent actually uses a 5-way switch. It's only shown as a 3-way switch on the diagrams. In normal pickup mode you don't get the option to do the usual bridge-middle middle-neck positions so it is a little limited in that respect. However the logic and FET switching circuitry will become quite a bit larger if you were to make the jump to 5-way switching, and you're left with two choices - with the sustainer active you either have to use the pickup selector with 3 "dead spots" (bridge-middle, middle, and middle-neck wouldn't be available as the driver has to be used); or provide extra multiple coiltap options to allow things like bridge humbucker, bridge coil tap+neck coil tap, neck parallel etc...The FET switching required would double and the logic would rapidly get out of control. Quote (Pete): Aha...now I think the really pertinent part is being revealed and may be of use. I get very little to no on switch noise, the sequential cascade needs to be different in different directions and this little detail seems to provide some control over this. More than likely, yes. I too get very little switch-on noise, it's the switch-off that's the problem for me. And I suspect that it's for the reasons I gave earlier. It's not a problem at switch-on because the LM386 takes a handful of milliseconds to become active, and the mechanical switch has taken care of the necessary bridge pickup selection/neck pickup bypass etc by the time the chip has become stable enough to pass signal to the driver. On switch-off however, the amp is the last thing to "die", not the first. And herein lieth the problem Quote (Pete): I think what I need to have happen is that the circuit power goes off but the bridge pickup remains connected for a little longer before coming out of bypass. You don't ever want the circuit on while the other pickups are in circuit. Also, a small delay could help in discharging the coils/s perhaps and stop the pop. So, for a little while after coming out of bypass, you want it to behave as a single bridge pickup guitar with nothing else connected. On turn on, you want an instant disengagement of the other pickups and connection to the bridge and perhaps a delay in the application of power to the circuit. I think what you really need is something that turns off the amp before it releases control of the driver, and then returns the guitar to normal configuration. Quote (Pete): Ok...if you look again at the 2073 data sheet, there is a muting circuit drawn there...not an actual pin like modern d-class chips, but a simple add on all the same... I think we're looking at different versions of the datasheet - I don't see anything like that on mine. Which one are you looking at? I think you will hear some switch-on/switch-off noise if you are running a perfect LM386 system, but probably only a tiny click. But the most significant cause of the switch-off thump at the moment is (I believe) the un-synchronised hard-switching of the amp and neck pickup. De-thumping circuits in big amps work by disconnecting the speakers via a fast-acting relay. The thump still happens inside the amp, you just don't hear it. Not really practical in our sustainer system. I suspect that the back EMF in the driver isn't the problem, it's the pickup winding that's the biggest offender. There may be a back EMF in the driver on switch-off, but it's probably quite small (low turns, low inductance, thus a small EMF). However, the coupling effect into the pickup winding will cause a tiny EMF spike into a massive EMF spike by way of the winding ratio between the driver and pickup windings - quite possibly hundreds of volts! And while this voltage may not be directly applied to the guitar circuitry during switch-off (the voltage will probably be slugged somewhat by the additional load imposed on it by the rest of the guitar system when everything is reconnected at switch-off), it will quite likely be induced and/or electrostatically coupled into surrounding circuitry (eg, nearby pickups, unshielded signal wires etc). I remember when I first built my sustainer, connecting my voltmeter across the pickup winding while the driver was active and I was seeing peaks of 100+ VAC as the guitar signal was being applied. I really hate to say this, but I think the stacked pickup/driver will always be troublesome because of this. Probably a bit hit-and-miss. How do you ensure that the "anti-EMF" will be the exact opposite of the EMF generated at the point of reconnection for example? Yes, definitely the way forward. I really think the synchronised switching (and/or cross-fading) of the system is the biggest clue in making the scheme pop-free. Indeed, let's hope so! I hope I haven't come across as too negative about the existing DIY sustainer system in recent posts, I'm only throwing my hat in the ring and trying to help iron out the bugs. And finally, to raise the tone a little out of geekspeak, some DIY porn: The Current Sustainer Build (Part 1) 1. The original humbucker pickup 2. Humbucker in pieces 3. Core removed from first bobbin 4. Old pickup winding wire removed 5. Bobbin end separated 6. Bobbin marked for shortening 7. Shortened bobbin ready for glueing 8. Bobbins glued and drying 9. Donor-transformer laminations being removed 10. Transformer laminations marked for cutting 11. Laminations cut from "E" section 12. Laminations cut for one bobbin 13. Laminations being filed to same length 14. Test fitting laminations in bobbins 15. Laminations re-varnished to maintain conductive isolation 16. Copper foil removed from cable offcut 17. Winding the bobbin coil 18. Almost done winding 19. Winding finished 20. Bobbin tape applied 21. Laminations fitted 22. All done! To be continued...

Edit: D'oh! Too many quotes! Quote (Pete): I take it hten that the FET switches are used to produce a sequential start up and shut down procedure. What is the sequence then? Looking at sheets 17 and 20, I don't think there is any sequence. The "gate filters" are just there to provide a quick fade up/fade down function to the gates of the FET's. The logic driving the gates is controlled by the selector switch, which if unfiltered, would cause pops in the signals as the FET's were turned off and on. The bridge and neck pickup gate filters have the same time constants, whereas the middle pickup/driver gate filter has a slighlty different time constant (long on fade up, short on fade down). There is also an additional filter shown at the bottom-right corner of sheet 20 that provides an additional cross-fade of the middle pickup output when the sustainer is turned on. Without analysing it closer it looks like when the sustainer is turned off and on everything starts simultaneously, but the delays from the time constants involved means that the driver supply, middle pickup output etc all finish their switching operations at differen't times. Quote (Pete): Again, I am not sure of the functions of the next blocks...the current source amplifier is the d-class switching amplifier for the driver I assume, the drive current limiter, some kind of AGC? Does the drive knob VR1 then control the threshold of this limiter? Probably not a part of the switching solution in any case. Yep, all correct. Quote (Pete): Moving on, the "low noise preamp" on sheet 19 is entered by MPU2...I am a little lost, let's see. This preamp is perhaps to turn the middle driver into an active pickup. I appears the signal is going from right to left ending in MPU2. Yup again. The driver is the middle pickup when used "backwards", but because of the lower number of turns the output of the driver is much lower than a normal guitar pickup would be, hence the requirement for a special low noise preamp on sheet 19 to boost it back up to more reasonable levels. Quote (Pete): Below this is a momentary switch on off flip flop, the adjoining comparator block I assume is producing control voltages V3 and V4...am I on the right track? Voltage V3 is always present - it's the battery supply that's turned on when you plug a lead into the guitar's output jack. V4 is the switched voltage going to the driver amp to start it up and shut it down when the flipflop is toggled. It's just a common point for all the FET-switched ouputs of each pickup to be combined before they're sent to the volume and tone controls, and finally the output jack. No different to the "common" connection on a normal 5-way selector switch. The 22K resistors (R30, R31) are added to prevent the ouput of one pickup from trying to back-feed into any other pickup connected to the common point, not normally a problem with a passive pickup scheme, but can be with an active system like this one. Mostly. The pickups have one FET each, plus the bridge and middle have an additional FET each to turn off the signal feeding the "EQ" sections - the EQ sections are only feeding the driver amp, so you need to switch them aswell when you've got the sustainer running and you're changing pickups at the same time. You mean U6d (sheet 20)? It's turning the output of the middle pickup on and off when 1. used as a normal pickup and controlled by the pickup switch, and 2. when the sustainer is turned on the pickup output needs to be turned off. It's just a filtered version of V3. V3A is only used by the logic gates. Filtering this supply helps keep unwanted switching noise out of the V3 supply feeding the analog circuitry. Very common technique. V4 is only used on sheet 18 to turn the drive circuit on and off. Note that the MOSFET's stay energised all the time (V2), but without any drive (V4 switched off) they'll stay cut off, not drawing any power. Probably helps control turn on/turn off noise. V4 is a bit like the "mute" function on the class-D chips. With the pickup selector in the middle position and the sustainer "off", the outputs of U5a and U5b are "high", cutting off the signal from the Neck and Bridge pickups, and allowing the signal from the middle pickup through. With the pickup selector in the mid position and the sustainer turned "on", the voltage at pin 13 of U6d starts to rise. When it gets high enough the output of U6d goes high, fading down the middle pickup output; and the the outputs of U5a and U5b go low, fading up the bridge and neck pickups. So with sustainer off your pickup selector operates as bridge-middle-neck, and with sustainer on you have (mid sustainer+bridge)-(mid sustainer+bridge+neck)-(mid sustainer+neck). The diode allows the fade time constant on that gate to be quick in one direction, and slow in the other. I think you mean bridge and neck. The current source amp only hears the pickups that are selected via FET's Q2 and Q3 (sheet 17). And yes, the EQ is used to condition the pickup signals for the driver. The bridge pickup looks like it's being band-passed at about 720Hz, (which seems odd, maybe it's more of an all-pass phase compensation circuit?...or just a typo!), and the neck pickup has a band-pass filter on it with some bass cut at 480Hz and a treble limit of 3.3KHz. I think there's two things going for this system. One is the cross-faded signals and voltages created by the various filtered-gate FET's and transistors, and the other (unfortunately for our DIY sustainers!) is that the driver does not create a step-up transforming effect with any stacked pickup combos because the driver is the pickup.

Fresh Fizz's analogy with the cinema projector was quite good. The MOSFET's push and pull current through the load (the driver) in a squarewave at very high constant frequency (has to be much higher than what is audible). Audible sound in the load is reproduced by varying the duty cycle of the square wave in direct proportion to the input signal. The inductive nature of the load provides a filtering effect, smoothing out the variations in the duty cycle and we percieve the effect as sound. If the duty cycle remains constant we hear nothing. DC motor speed controllers work on the same principle - by varying the duty cycle of a fixed frequency squarewave you can vary the amount of "average" voltage applied to the motor, thus varying the motor rotational speed. Because the MOSFET driving elements are either fully on or fully off (creating the squarewave), ohms law says that the power dissipated in the amp will be extremely low (MOSFET fully on, R=very low, V across MOSFET=very low, P=V^2 / R is also very low; MOSFET fully off, no current flow in MOSFET, no power to dissipate). Getting back to FF's projector analogy, each frame of the footage is only shown for a fraction of a second. If the frames are shown fast enough our eyes "filter" out the flickering and we see the result as fluid movement on screen. Easy enough to try though. It could also just be that the LM386 becomes unstable driving the load if the power supply voltage suddenly disappears. Just thinking about this now, this could be another drawback of the transforming effect of the driver/pickup stack. The pop (or "fart" I'm getting) could also be caused by the disconnection of the driver and simultaneous re-connection of the pickup winding while the driver is still being driven. The transforming effect of the combo system produces lots of volts (at very low current) in the pickup winding as we've already proven. If the pickup is reconnected to the system before the induced voltage on the pickup winding has had a chance to settle down to zero you'd get a pop. If you're only running with one pickup you probably wouldn't have this problem as you're not attempting to re-connect the combo pickup back to the system, and so the voltage across the pickup winding has a chance to decay naturally Why it seems worse on pickup selections other than the bridge I'm not sure. Perhaps it's the physical distance between the driver and the selected pickup which exacerbates the problem? You could also be looking at some kind of EMI pulse from the driver when it's disconnected, maybe even the back EMF in the pickup winding, that could be picked up by the "closer" pickup? One way I can see to get around this is perhaps utilising some kind of staged switching sequence. When the sustainer is turned on: 1. the combo pickup is disconnected (and possibly shorted out), 2. input signal to driver amp is set to zero, 3. power is applied to driver amp (possibly even with the slow ramp circuit), and 4. input signal to driver amp is gradually raised up to nominal (over maybe 10mSec or so). The power-down sequence would be the reverse order. Obviously you're not going to be able to do this with mechanical switching, so you're looking at a FET for the light duty switching operations (driver input signal fade in/fade out), a transistor for the power supply switching, maybe a relay or some other high-voltage transistor switch for the pickup leads (FET's are probably too fragile to cope with the big voltages induced by the driver combo), and something to perform the sequenced switching (maybe an 8-pin PIC?). There may be a way to do it by some kind of zero-crossing detector aswell, so that the driver signal is turned on and off when the input signal passes through 0V. Maybe triggering via TRIAC's? I dunno. Probably getting a bit too complex for what you're wanting I suspect? Only because it enables you to do fancy switching operations like I described above. Plus sustainiac describe using the driver as a pickup in itself which requires a special preamp to boost the signal up to normal guitar levels. Shoudn't damage the amp, no. 4mA constant drain on a 9V battery might be a little steep, although the datasheet doesn't mention if that 4mA is with or without a load connected. You might find that the quiescent current drops if the chip is feeding no load. And anyway, if I'm right about my theory about the transforming effect of the combo setup, the sudden disconnection of the driver while the amp is still running would still cause the sudden back EMF in the pickup winding, resulting in the "pop" or "thump". You'd still need to do some sequenced switching and signal ramping to get around it. Yes, class-D in only SMD outlines is a pain I agree, however there are means and ways of manually soldering SMD chips out there that I'm hoping I can perform myself. The LM386 is probably so lasting because it's so common and easy for anyone to use. It also finds use in ohter areas like servo drivers and oscillators, which class D chips can't really do without a lot of other support components. Dunno about the 2073, it looks like quite an old chip. Can you still get it? There may be other ones out there that do the same thing but better. I think Col found some TDA7xx things that might be better suited. I can't see anything on the datasheet that suggests this chip has a muting function, although it does say that it has "zero turn-on noise". The muting function is handy as it'll eliminate another fancy sequenced switching element. With the chip in mute mode you can probably switch the thing on and off without having to worry about pops and thumps (no signal output, no voltage across the sustainer, no voltage stepped up via the pickup, no pop...maybe...). If you can find a BTL linear amp that has the muting function I'd grab one for experimenting with. It's normally only a single pin that is either tied to ground or the supply via a switch.

Yup, I know the design principles behind class D amps, I was more thinking out loud. And the trade off's with distortion decrease as the switching frequency increases (the MAXIM chips for example operate with a switching frequency of 1+ MHz). More importantly, we're not concernened with super high fidelity operation as the amp is feeding a "noiseless" driver. The patent shows the guitar as a fully active preamp system for the pickups which runs 100% of the time, and the sustianer circuitry is only switched in via a slow ramp up/ramp down transistor on the supply line feeding the class D amp section (see fig 11 of the sustainiac patent - CR4, R49, C28 and Q8 - it's switching the voltage V4 which supplies the driver amp). The slow ramp up/down is meant to minimise pops in the driver circuitry. Not sure how well it'll work with the old LM386 though. The noise I get when I switch my current sustainer in and out sounds more like a "thhbbpbpbp", which suggests that it's perhaps the chip "motorboating" (low frequency oscillation) as the power supply voltage rises and falls - the LM386 may be unstable if the power supply varies at a (relatively) slow rate. The class D amp may be a different ketlle 'o fish all together when the supply voltage is ramped up and down. Maybe worth a try though? It's only an extra couple of components. Interestingly, many of the dedicated class D chips I've been looking at have a low-power standby option which "poplessly" turns the output on and off, and when enabled (usually by switching a voltage to a single pin on the chip - SPST switching as a possibility?) puts the chip to "sleep" drawing around 0.1uA! In my current sustainer re-exploration, I've pulled apart the humbucker and started re-building it into the bilateral driver. Things are going really nicely at this stage (touch wood) - I've managed to cut the bobbins and re-join them into a pair of half-bobbins, and all the laminations are cut and test-fitted into the new bobbins. I've given the laminations a quick coat of varnish over the exposed edges where I cut them from the donor transformer (to maintain conductive isolation between the lams when they're packed into the bobbin) , and I'm just waiting overnight for them to dry. Looks really neat. I've been taking lots of photos along the way, so hopefully I'll be able to do some kind of photo-essay in the next few days.

Yes, but I'm more interested in trying out the class D switching amp option this time round rather than the classic linear LM386 circuits we've been using thus far The patent images show a discrete class-D output stage made from a comparator and a pair of N- and P-channel MOSFET drivers. At a switching frequency of many 10's of KHz the amp won't "see" the DC resistance of <3ohms (unlike the linear LM386), as the impedance at that frequency will be much higher - perhaps 10's of ohms? Depends on the inductance of the coils given the number of turns and the laminated core material. Of course, that's not to say I couldn't use a dedicated class D chip amp. It's possible that when the patent was written class D chips weren't as common, and it was cheaper for Sustainiac to roll their own. I just came back from town with a throw-away humbucker that the guys at a music store gave me. The pole pieces on each bobbin are actually one-piece steel running the full length of the pickup, which I think may work to my advantage better than the traditinal six-slug single coil bobbin. I can cut the middle of the bobbins out and re-glue the two ends together to make the shorter bobbins required for the bilateral driver, and they already have perfectly rectangular core windows already moulded into the bobbin. Yup, that's pretty much how I made my first sustainer with a solid steel "rail". The donor pickup was a crappy Yamaha single coil with six slugs that I removed and then cut the middle out to make a long rectagular slot to install the new core into. I'm not entirely sure that they'd purposely put in inaccurate information in their patents. It might be out-of-date or superseeded. It's in the company's interest to describe the product fully and accurately so that if they need to demonstrate the system or make a claim against an patent infringement, they can do so in the knowledge that the information is "all there". Otherwise any legal defence will be compromised by innacurate information in the patent.