|
Post by Charlie Honkmeister on Jan 4, 2017 16:33:57 GMT -5
Ken, We have to have some ferromagnetic material in play to guide the field and increase the inductance since an air coil pickup would be too large to fit any of the standard packages and it would be a bear to get the magnetic field guided to the strings. Ferromagnetic material does not have to be conductive. There are ferrites that can do the job with any permeability that you want. If you want to get permanent field to the strings in an air core pickup, just use Neo magnets inside the coil under the strings. Neo's permeability is about 2% higher; so it is effectively the same. If you want such an air core pickup for some reason, use very fine wire. That's true. Ceramic magnets aren't conductive AFAIK. I didn't remember that about Neo magnets as well. The basic idea is to reduce inductance increase caused by the magnet and magnetic structure, so we can get more turns on the coil and get more output. But like anything else, there's a tradeoff with winding capacitance of course. In parallel coil mode, several "vintage neck" pickups are self-resonant at about 10 Khz with fairly decent Q - 12+ dB peaks. The tradeoff is 6 dB less output than series coil mode, which I don't think is a big deal at all. I'm still testing a couple, but I may not need to go very far to get close to the goal.
|
|
|
Post by Charlie Honkmeister on Jan 4, 2017 16:28:18 GMT -5
I'm not sure what the necessity is behind the double rail design, but if you absolutely needed rails, laminations would take care of the eddy currents. As far as a coil tapped humbucker retaining the same output, improving upon series/parallel; I bet that's not strictly true, as it's lowering the inductance in "tapped mode", but it's probably not nearly as low of an inductance as parallel mode. If the capacitance is somehow kept lower through the tapping, it will also yield a higher resonant peak for a higher inductance, or overall voltage output. The real evil of parallel mode is that you have a low inductance, but still a very high capacitance. Normally, low peak means high voltage, but a parallel humbucker essentially robs you of both. If you want a high Q factor, that's not hard to do, you just have to do as Fender does and stick to plastic, fiber and magnet in the construction. If you use ceramic or neodymium, the only damping losses that will exists are those that owe to the wire resistance. In general, such a high Q factor is unusable, Fender uses 250k pots just to further dampen the Q factor of their single coils. A pickup with a very high resonant peak is essentially flat, so it would be a good platform for tone shaping, which they claim is the basis of the Fishman Fluence's tone shaping, again prompting the question, why only two voicings? I was never clear on what technical benefits the Fluence enjoyed due to the PCB layered coil design. Since we're on the subject, I discovered something I believe is not common knowledge, when you tap a coil and the other side of the coil is still connected at one end, it does not conduct current, but it does still cause a very high capacitance to occur. On the Seymour Duncan SSL-4, the tapped resonant peak suggests that the unused portion of coil added about 450pF of capacitance, which is equivalent to an other 12ft or so of guitar cable. Yes, it's that bad. I always wondered by Seymour Duncan's tap point would cut the coil in half, rather than say, cut it down by 3/4, and that's the reason. You have to cut the inductance down that far in order to offset the very high capacitance, and output voltage suffers greatly as a result. It's a design flaw, IMO. If you do a coil tap, and you want to retain high output, you should use a four conductor cable so that the outer portion of coil can be disconnected at both ends. You can then set the tap point later in the wind, say the 3/4 or 4/5ths mark, knowing that the extra guitar cable's worth of capacitance won't be raining on your inductance. Antigua, Thanks for your comments! I don't necessarily have to have a rail design, just used the Barden Two/Tone as an example, and also it's easier to play with laminations or thinner material with a filler on a rail design, than it would be to do the same thing with a polepiece design. Also, with a rail design, there's no worry about polepiece spacing and having the polepieces line up for different string spreads, or dropout on big string bends. Really, reducing eddies and inductance would be a way to get a high self-resonant frequency with high Q as you said, which could be damped down with loading and capacitance to whatever was desired. Reducing the iron content of the cores/rails would probably allow more turns for greater output for the same inductance. The main idea of reducing eddies is probably secondary to reducing inductance, but the idea is to have the most output and clarity possible, and do all voicing external to the pickup. That's interesting about what you discovered with a connected section of the coil on a coil tap. -Charlie
|
|
|
Post by Charlie Honkmeister on Jan 4, 2017 10:22:45 GMT -5
Ken, Thanks - what I was getting at was "minimal" eddy losses. We have to have some ferromagnetic material in play to guide the field and increase the inductance since an air coil pickup would be too large to fit any of the standard packages and it would be a bear to get the magnetic field guided to the strings. Halbach arrays, anyone? So, steel rails but not too large, like a Joe Barden HB or a Bill Lawrence L-500 or L-90. The only other design I have seen which might work fine would be a humbucker with 12 individual magnets in the pole piece positions, like two traditional Strat pickups next to each other, and a non-metallic base. A slightly underwound Firebird design is tempting. The mini-humbucker size (aperture width) is really underrated IMHO. Paul Reed Smith is doing some PU's in that approximate size for some of their models.
|
|
|
Post by Charlie Honkmeister on Jan 4, 2017 0:39:37 GMT -5
Stratotarts drafted up a circuit design for the V5 integrator and had some PCss' produced, presumably in the East somewhere. I don't know how you go about doing that, but I'd think you could do the same with your schematic, and have them pre-attached to a suitable pot that mounts into most electrics, and boom, you have a product. I've been interested in something like a pickup modeler for a long time, and I was hoping the Fluence would be just that. Even though your device requires batteries, I would buy it, for at least one or two of my guitars. The cool thing is that it's really just a buffer, albeit a really good buffer. It can be used for a piezo buffer, mag buffer without variable res tone, or mag buffer with variable res tone. I have it set up so that there's a plug in 8 pin DIP socket which takes a DIP header where you can solder in any RC combination on the input stage as well as the VR cap, so it can work with a variety of PU's , or just jumpered to be a piezo buffer with 10 Meg input impedance. I have SMT PC boards in hand for three variants of the buffer and have built four so far (one clinker with blown traces, three working). I'm getting better at hand soldering the SMT devices and, as long as it's not smaller than an 0805, it's really just a different technique and doesn't require anything more than watching a couple of videos on YouTube to get some hints/tips, and practicing. Of course in production, we are talking solder paste, reflow ovens, and stuff that's more normal for that kind of board. If you have the PCB standardized design files (I use DipTrace), you can get ready-to-stuff boards for about $5.00 for qty. 3 1 inch by 1 inch boards from OSHpark in about a week. So far, the LSK389A (dual monolithic JFET ) board absolutely wins on noise. There's also an extremely subtle tone quality difference that I like, although it may be selection bias on my part. I'm going to be trying another op amp on the op amp SMT board when I get the parts. Yes, a production board which is integrated with a tone pot, would be a good product. Sincerely, thanks for the encouragement. -Charlie
|
|
|
Post by Charlie Honkmeister on Jan 3, 2017 23:59:11 GMT -5
So here's the quest(ion) - if we had a Gibson humbucker size pickup with as close to zero eddy losses as possible, then wouldn't we have more design freedom to engineer the response electrically to be closer to anything we wanted, including getting pretty close to emulating a slew of other humbuckers?
Some of this thread was sparked long ago with a set of Joe Barden Two/Tone HB's that I purchased back in 1998 or so. I was trying to figure out just how the Two/Tone could get so close to the single coil type tonality, without dropping volume like a conventional coil tapped approach.
Fortunately or unfortunately, I blew up the neck pickup beyond repair, trying to get a new longer cable soldered to it (what kind of decision was it to use flea clips instead of grommets to terminate the coil wires !?!) , and stripped it to see what was going on. I still have the bridge Two/Tone and the skeleton of the neck PU available for pics if anyone is interested .
So, it looks like several things were done, and it looks like I can get close with off the shelf Mojotone parts.
Here's some of the "features" of the Two/Tone. Of course, it's a double rail pickup.
1. No metal parts in the whole darned thing except for thin steel rails inside the coils. Unless you want to count the two thin strips of copper tape which ground the rails and the two ceramic magnets. The baseplate and bobbin tops are plastic/phenolic.
2. Rather conventional rail bobbins. One small ceramic magnet per rail, positioned outboard of the rail relative to the center of the pickup.
3. The "two/tone" feature is done by tapping both coils at about the 60% point and bringing out two wires for the taps along with the normal two wires per coil for coil start and coil end.
Just as a FYI, with the Mojotone 53 mm fiber baseplate, and their rails, magnets and rail bobbins, it might be possible to make a "faux Barden" , albeit the Mojotone rails are somewhat thicker than the JB rails. So... maybe laminated rails? How far do we want to pursue reducing eddies?
I could care less about the two/tone feature because of my variable resonant frequency / variable cap control buffer. If I can get the inductance in the 1 Henry range and the self-resonance about 10K, we should be able to get just about any combination of resonant frequency and Q we want within reason.
Comments? Looking for some encouragement to actually pursue doing this. I have the Mojotone parts in hand already and wouldn't take much to take these to Sonny Walton (who has a much better winder than I do) and wind some up.
Thanks,
Charlie
|
|
|
Post by Charlie Honkmeister on Dec 30, 2016 10:28:23 GMT -5
Antigua,
I do understand where you are coming from by trying to stay on the passive side when working on enhancing the instrument's tone possibilities. A lot of players feel the same as you do, and I for one, would be appalled at the idea of having to hack up a quality vintage instrument to put anything irreversible in it. I saw too many examples of that in the '70's and '80's.
Having a battery and going active is a big river to cross for a lot of players. But let's use the example of bass players. The majority of basses are active with a battery these days, because performance wise, when it counts, an active bass is capable of a lot more tonality than a passive bass. For those players, that advantage trumped everything else. The same players might have a vintage Precision or Jazz passive in their arsenal, use and enjoy them, but not use them for touring or gigs.
Using active electronics has been done over and over again for electric guitar since it's a mature instrument/market with a lot of people trying to innovate over several decades. I've never been happy with anything active I've built into a guitar until now. For some reason, the EQ wasn't right, or I had too much adjustment and had to dink with too many controls, or it just didn't sound natural, have the mojo whatever that is, etc.
So, the buffer/variable cap trick is absolutely minimalist in terms of how much electronics is added, and uses the pickup itself to form the resonant circuit. That's a key difference from just taking the tone as is, and then doing all kinds of multiple stage filtering to get different sounds out.
I can't overstress how "natural" the variable cap/buffer tone control sounds. But, I did my best to engineer it that way, by using Circuitlab simulation/modeling to select components (loading resistance on the pickup, capacitor values, a few tricks like that) to get realistic resonant peaks that are typical of a passive instrument with pickups loaded by tone and volume pots, and a reasonable cable length to a typical amp input. The surprise was, the modelling and simulation worked and the goal of " dial up another pickup with one knob" was IMHO pretty well achieved.
BTW all of the wiring mods/switching arrangements we have all tried over the years should still be valid with this approach but I prefer to use volume controls after the buffer and, I have no (or vary little) use for passive tone controls as a pot, any more after doing this and hearing the results.
-Charlie
|
|
|
Post by Charlie Honkmeister on Dec 28, 2016 1:49:16 GMT -5
Here's some typical response plots of the general circuit above in resonant (movable peak) and lowpass (tone control is just lowpass or treble control) modes, controlled by the switch S1. I've improved the circuit since then to be able to get more even height peaks in the middle pot positions in resonant mode, but this should convey the idea of what the circuit above can do. The curves are at 10% points in the pot value. Please note that at about the 90% point of the tone pot in lowpass mode , the pickup resonance is nulled out and the response is flat within a dB or so to 8 KHz.
|
|
|
Post by Charlie Honkmeister on Dec 28, 2016 1:18:00 GMT -5
JohnH, Thanks and I do appreciate the links to the buffer design and mixer threads. I do like and have used simple single-JFET buffers before, but what I have been chasing is low-inductance pickups which without any tone control, have a self-resonance of about 8-10 KHz. This is to be able to get really acoustic-like tonality and/or flat response to over 8 KHz when desired. Using pickups with inductances in the 1 to 1.2 Henry range and then attempting to do the variable capacitor thing with a buffer, requires the buffer to have a relatively low output impedance, for consistent resonant peak amplitude as you sweep the tone (resonant frequency) control. If you decrease the drain resistor to lower the output Z of the buffer, the single JFET buffer starts to become a bit of a current hog for the signal swing you want to get out of the buffer to drive both the tone control pot and the volume pot, on out through the cable. There's two main ways around this that I'm pursuing. One is to use an opamp buffer with a low current opamp (example, TI OPA188, and others), and the other is to use a 2-JFET configuration where the bottom JFET is a voltage-controlled current source in push-pull. This is called a Borbely follower, after Erno Borbely, who came up with it as a JFET version of the tube amp world's White cathode follower. It is only a bit more complex than the simple JFET buffer. Here's my version of the complete system, with resonant/lowpass switching on a push-pull pot switch. (BTW through hole TO-92 J201's will work with this, SMT version has different specs and doesn't work for current draw): I have the SMD board design for this version, but am actually going with a board which uses a Linear Systems LSK389A ultra-low-noise JFET pair. The opamp version is extremely straightforward as well; just substitute a non-inverting opamp buffer for the JFET follower. I have built the opamp version and have just gotten boards from OSHpark to build the LSK389A version. The version I have in the demo guitar uses a commercial very high quality buffer module, the Creation Audio Labs Redeemer. However the current consumption of the Redeemer is about 1.8 mA and I need to have a lower cost lower current buffer overall for general use as a commercial product. The current plan after testing is to open source the board designs on OSHPark, and maybe do a through-hole version of the board if there's some interest in such a thing. -Charlie
|
|
|
Post by Charlie Honkmeister on Dec 26, 2016 15:28:13 GMT -5
Thanks!
Yes, the variable capacitor with a buffer idea is exactly what I implemented for the variable resonant frequency control. It looks like several people picked up on this in roughly the same time frame.
The nice thing is that the buffer can both provide the benefits of a buffer in eliminating tone dependency on the length and quality of the cable, and also drive the variable capacitance tone control.
What I'm pursuing is high quality buffers which have very small current drain. The scenario is that there may be 3 buffers in an instrument, one for each pickup as in a Les Paul 2H configuration, and one buffer for a piezo bridge.
So the target is about 500 uA drain per buffer.
|
|
|
Post by Charlie Honkmeister on Dec 25, 2016 0:39:45 GMT -5
I also agree with you on the Fishman Fluence. They spent a whiz-bang amount of money developing that, but ended up with just two voicing selections. I've heard they sound good but there's no convincing advantage that the unique coil construction translates into better sound in any way. It's a well engineered product though, just underwhelming in its value proposition.
The way I look at it is that those two voicing selections are just two points on my VR tone control's pot, and I have a lot more choices on the same pot, for a lot less money.
|
|
|
Post by Charlie Honkmeister on Dec 25, 2016 0:32:51 GMT -5
Antigua, Appreciate your comments. Even with all the external "revoicing" filters/boxes/thingies out there, some of which I've bought before (!), not to mention all the other gee-tar gizmos for 45+ years which have promised to take a player's tone to the pearly gates, there still might be a market for a small, cheap, simple onboard product to really increase the versatility of the instrument for most genres and most amps/FX chains that most of the players will plug into. After all, just in replacement pickups alone, pro and semi-pro players are pretty well used to the idea of dropping $300-500+ for a set of pickups, a wiring scheme, and maybe the tech's installation time, in chasing a "better" sound. In fact, probably 60 to 80% of the current electric guitar/amp/effect/accessory market is based on the business model that a guitarist will chase "the tone" for 10-20 years and never be satisfied, leading to more gear purchases over the long run. Far be it from me to change a proven business model in a $2 billion dollar industry, but I think there's a niche or two available. Maybe it's just working with small builders to custom-voice their instruments, or offering kits on Ebay to allow someone to improve the axe they have without spending the traditional amount of money, or something like that. At least one of my posts a while back was to point out the original articles/Web pages on the variable capacitance tone control, and that it is considered "prior art" and public domain, so one of the big player MI companies couldn't ring-fence the idea with patents. This is a pretty wide-ranging thread and I don't want to divert the stream too much, so I'd like to suggest that we can move the variable resonance tone control discussion to the original thread I posted, guitarnuts2.proboards.com/thread/7823-Charlie
|
|
|
Post by Charlie Honkmeister on Dec 24, 2016 17:32:29 GMT -5
Antigua, you said: "Is a wah-wah pedal anything like pot controlled capacitance? It sure sounds a lot like what I would expect such a thing to sound like." You are 100% on the nose with that supposition. The pot-controlled capacitance trick was first seen in the original Hammond/Vox Cry Baby wah pedal in the '60's. Here's a post I made on the subject which links to R.G. Keen's analysis of the wah pedal. music-electronics-forum.com/t37793-post362519/#post362519In terms of "does it sound like a wah pedal?" the answer is, It depends. If you make it peaky, with large Q values, it can indeed sound like a wah. But in this case we are carefully controlling the Q value and resonance range to get a desired result. Also, there's only a single resonant peak. You aren't taking a pickup signal which already has one resonant peak, and processing it further with a wah with another resonant peak. The variable resonance tone control really is completely changing the pickup tonality in a musically useful natural-sounding way. For what I was targeting, the desired result is: be able to go from Gibson humbucker tonality (1.8 - 2.5 Khz) to Tele/Strat tonality (3-5 Khz) with one knob on the instrument. If you set the Q value to produce a consistent 8-9.5 dB peak over the whole range, you end up with a very good emulation of what the desired emulated pickup sounds like , loaded by open volume and tone controls, and loaded by some reasonable cable capacitance. The guitar sounds good like that, and I also have the push-pull switch to convert the resonant peak control to just a simple lowpass, to be able to have a "flat" or simply rolled off response from the pickup if desired, for going into a DAW and using plugins. So I haven't seen the need to vary the pickup loading that much, as Roly Roper did with his variable R control. Of course pickup magnetic structure , slant, and position under the strings will play a part in how closely you can emulate another pickup type that's physically different. But so far I have been quite pleased with the results I have been getting with my cheap Chinese single coil sized Strat rail humbuckers in my Strat. The VR buffered tone control really is useful musically and it's darn close to a "dial a pickup" knob. It's great to be able to share this with some technically astute folks such as yourselves. -Charlie
|
|
|
Post by Charlie Honkmeister on Dec 23, 2016 16:21:20 GMT -5
Antigua, Many thanks for taking the time to give a step by step tutorial on using LTSpice to sim guitar circuits. Also, ms, thanks for the Python example. I'd like to throw in a couple of side comments along the lines of Antigua's post, having messed with this stuff for a while now. (I use Circuitlab but that's a paid service; free is good too.) 1. A good LCR meter is a godsend if you are really delving into this pickup voicing stuff to any large degree. The DER EE DE-5000 has been tested to be as good as the Extech (the handheld gold standard) by the technically knowledgeable folks in the Pickup Makers forum. The absolute necessity here is to do your pickup testing at a 100 Hz or 120 Hz test frequency. The reason is that eddy losses in a pickup are proportional to frequency and low test frequencies minimize this error. The DE-5000 is about 110bucks with both sets of test probes on Ebay, less than two big name brand pickups. 2. I completely agree with newey on "marketing." A lot of the mumbo jumbo and mystique in the pickup business serves those who manufacture the pickups. A lot of times, musicians are paying lots of money to swap PU's when they could have some other alternatives (pot values, loading resistors, amp settings, pedals, different tone caps) for creating a different expressive tone with their instrument and rig, as you pointed out. The same "mystique marketing" thing applies to guitars, tonewood, amps, pedals, you name it. 3. I'm a firm believer in active buffering and have (re) discovered that it's really easy if you already are going to use a buffer, to use the buffer to creat a pot-controlled variable capacitor so you can tune and tweak the resonant frequency to your heart's content. This IMHO is vastly better as a tone control than the conventional passive tone control. With some good measurements of the pickup with the LCR meter, I can plug those numbers into Circuitlab to set the resonant frequency and Q of the buffered pickup circuit, without having to worry about what cable/amp/pedal the player is going to use. So that's really the next step of modelling the signal chain - engineering the signal chain. For anyone who's interested, here's a link to a demo I did, and some technical information, on the variable resonance tone control technique: music-electronics-forum.com/t43081/
|
|
|
Post by Charlie Honkmeister on Dec 6, 2016 23:23:37 GMT -5
Just a quick note on buffers: The buffer which drives both the cable and the variable rez tone control has to be able to drive about a 2.5K load to ground and have fairly low output impedance. So simple single-JFET buffers can work, but the source resistor has to be so low that the buffer is drawing quite a bit of current. There's three buffer approaches which seem to work and give acceptably low (well under 1 mA, and ideally close to 500 uA) current drain: 1. JFET Borbely follower. This is just a JFET version of a White cathode follower. 2. Hybrid MOSFET/JFET modified Borbely follower (JFET on the bottom as current source, MOSFET on the top.) 3. Low current low noise op amp. My first proto will be with the OPA188 which has under 500 uA current draw typical. I have PCB designs and SMT boards for (1) and (3) above and will be doing those proto boards in the next couple of months. The PCB's can be used for piezo buffers as well. Please let me know if you are interested in my posting the schematics for these. Just as a teaser, here's the PCB top side for the opamp version. Take care, C.H.
|
|
|
Post by Charlie Honkmeister on Dec 6, 2016 11:56:46 GMT -5
|
|
|
Post by Charlie Honkmeister on Dec 5, 2016 13:01:53 GMT -5
Thanks very much for the heads up; this looks to be a huge resource. I have shared this with my guitar-oriented FB groups.
Appreciate it!
|
|
|
Post by Charlie Honkmeister on Dec 5, 2016 10:13:22 GMT -5
Newey,
Thanks for your welcome to the forum. I really like the simplicity of this idea, and also that the buffer is fully functional as a buffer, as well.
I am prototyping two onboard buffers, one with an opamp and one with discrete JFETs, to offer a lower cost option than the CAL Redeemer buffer I used in the demo guitar. If there is some interest, I will update the forum with progress, and availability of boards.
Take care,
Charlie
|
|
|
Post by Charlie Honkmeister on Dec 4, 2016 23:19:15 GMT -5
Two to add to the mix: 1. Joe Barden Two Tone - I got a set of these in about 2003 or so. Very nice single coil sound from a humbucking package. Wiring the DPDT switch to the pickup is a bit of a pain. 2. (near future) - my variable resonance control on a humbucker. Might be worth the wait, I have the solution for Strat sized single coils but am doing humbuckers in the next month or two. The key spec is that the pickup you use has to be "vintage" wind (7-8K DCR or thereabouts) and about 4 Henries inductance in normal series coil mode. music-electronics-forum.com/t43081/-Charlie
|
|
|
Post by Charlie Honkmeister on Dec 4, 2016 22:54:15 GMT -5
EDIT: moved to Tone Control Discussions by sumgai, Dec 5 '16. Explanation below.
Everyone, Here's a link to my post on successfully demo'ing a one-knob variable resonant frequency control for electric guitar. It is extremely simple and the version I demo'ed used completely off the shelf parts including a commercial buffer. This provides an onboard tone control which is IMHO vastly better than a conventional passive tone control. The version I demo'ed in a Strat is able to vary a single pickup resonance from about 1.4 KHz to above 5 KHz. This provides an extreme amount of versatility while retaining clarity even at low resonant frequency settings. linkPlease follow the link including the two other links to the "theory" posts, and I would be happy to clarify or share more information on this technique. Take care, Charlie
|
|