|
Post by Charlie Honkmeister on May 29, 2017 12:02:02 GMT -5
If you wanted to go this way, one idea would be to use a plastic pickup cover, and adhesive copper tape, to try different slot configurations, instead of all that jeweler's saw/Dremel work.
|
|
|
Post by Charlie Honkmeister on May 29, 2017 10:20:55 GMT -5
I'm wondering about the string balance on your implementation. It's clear that the bulk response is transparent, but I'm worried about individual poles. The G string slot clearly has no conductive path, but from an individual pole point of view, the A string for example, does. It's possible for currents to travel in the side wall and circumvent the slot. I wonder if you could compare the G with the A using your small test coil to confirm or deny this? I have one more brass cover to butcher, and it will be a while before I can get more. I want to make the next experiment count. If there's any tonality or string amplitude difference because of the slot location, then there might be some interesting experimentation to be done with different number, location, and even depth ( less than full depth top-to-bottom slots ). Bill Lawrence was of the opinion that eddy currents are like seasoning in cooking, and aren't entirely bad. If you're on to something here where you can potentially not just eliminate, but selectively introduce, eddies up to maybe even a string-by-string basis, you just might have something really cool to further investigate. Might be worth experimenting a bit more to find out. -Charlie
|
|
|
Post by Charlie Honkmeister on May 28, 2017 13:56:23 GMT -5
Long time update (probably last I guess!) - I did finally add in the treble bleed for the Vol pot. All I can say is thank you to all who helped with this project - it has exceeded my hopes and continually amazes me with how incredibly versatile and intuitive it is. I have had complicated push/pull in/out of phase series/parallel switching setups before and always ended up feeling that there was too much going on for real-world live use, but here the very simple mods that retain the normal layout yet provide amazing enhancements to the normal Strat options is truly wonderful. The blend in positions 1 and 5 on the switch is just completely fabulous! Thanks everyone! All the best Thanks for the feedback on the blend pot. Having a no-load pot for the blend really seems to be the right way to go since you can disconnect it completely. Glad it worked out so well and you're happy with the tone versatility.
|
|
|
Post by Charlie Honkmeister on May 1, 2017 10:46:27 GMT -5
I don't think that increasing the maximum possible Q should be a sonic design goal in itself, but it can be a yardstick for measuring some other parameters (internal resistance of the coil, lower eddy losses, efficiency of the magnetic circuit) that are worth pursuing in a good pickup design. From a pure transducer engineering standpoint, designing a transducer that has a big resonant peak in the very frequency range you're trying to transduce, is just nuts. Nobody would do it that way these days; we're just faced with the consequences of what "they" did back then, for their own reasons, and are used to what "that" sounds like. If you are chasing a subjectively better tonality but one which is still familiar as an "electric guitar sound," you aren't going to need resonant peaks more than 6-8.5 dB, because when loaded by volume and tone controls, through a cable, into the amp input impedance, that's the practical Q you're probably going to get with a passive setup and a conventional design. And that's what we are all used to hearing. But maximizing the possible Q on the bench might be indicative that some other things in the pickup design are good and won't get in the way of you going in the direction of "vintage overall sound, but better." You'd have in general more flexibility to tailor the sound outside the pickup, as Mike (ms) pointed out. But if you aren't trying to "emulate but improve on", the tonalities we know and love from the instrument, and/or if you want the absolute most tonal flexibility possible, IMHO it doesn't make sense to have a resonant peak at all; you should go for "hi-fi flat" and then do any signal processing you want downstream. It is a goal because it allows you to use lower value vol pots, reducing the amount the tone changes with the volume control position. Or it allows you to run the tone control lower in normal operation so that you have some boost available for when you need it. Or some combination of the two. OK, higher possible Q allows you to move toward lower source impedance for the pickup, for about the same output level, so you can do those things. I'm on board with that. If you get a high efficiency magnetic circuit, you can lower turns and/or use lower AWG wire. The lower the source impedance, the more flexibility you have on usability of the volume and tone controls. Also, the cable capacitance will affect the pickup resonant frequency less. All good things.
|
|
|
Post by Charlie Honkmeister on May 1, 2017 9:27:35 GMT -5
I don't think that increasing the maximum possible Q should be a sonic design goal in itself, but it can be a yardstick for measuring some other parameters (internal resistance of the coil, lower eddy losses, efficiency of the magnetic circuit) that are worth pursuing in a good pickup design.
From a pure transducer engineering standpoint, designing a transducer that has a big resonant peak in the very frequency range you're trying to transduce, is just nuts. Nobody would do it that way these days; we're just faced with the consequences of what "they" did back then, for their own reasons, and are used to what "that" sounds like.
If you are chasing a subjectively better tonality but one which is still familiar as an "electric guitar sound," you aren't going to need resonant peaks more than 6-8.5 dB, because when loaded by volume and tone controls, through a cable, into the amp input impedance, that's the practical Q you're probably going to get with a passive setup and a conventional design. And that's what we are all used to hearing.
But maximizing the possible Q on the bench might be indicative that some other things in the pickup design are good and won't get in the way of you going in the direction of "vintage overall sound, but better." You'd have in general more flexibility to tailor the sound outside the pickup, as Mike (ms) pointed out.
But if you aren't trying to "emulate but improve on", the tonalities we know and love from the instrument, and/or if you want the absolute most tonal flexibility possible, IMHO it doesn't make sense to have a resonant peak at all; you should go for "hi-fi flat" and then do any signal processing you want downstream.
|
|
|
Post by Charlie Honkmeister on Apr 28, 2017 12:14:47 GMT -5
well explained Charlie! thnx! You were very accurate. I agree. We are talking about 3 technologies here. Still strikes me why no COSM in the SY-300, it would make more sense than putting COSM in the end of the chain (after the synthesis part) in the GR55. Maybe Roland has smth cooking for the future? There's two types of COSM modeling - one works on the individual string signals to model a particular instrument, allow alternate tunings/pitch shifts, etc., and the other place for COSM is in emulating the amp, speaker, microphone combination. The VG series and the GP-10 do both kinds of COSM, so they can model a particular guitar playing with a certain modelled pickup selected, through a particular amp with a certain mike, etc. The Roland all-in-one effects units (like the GT-100 )can do the second type of amp/speaker/mike COSM as well as a lot of other effects. But the tube amp/speaker/mike COSM modeling is mostly guitar-oriented. Not to say you couldn't use tube amp distortion/overdrive on a SY300 patch, but usually you would want to do something else. The SY-300 sounds are way, way different than what you want to sound like with a vintage guitar through a vintage amp, etc. Maybe the bottom line is that for now Roland wants you to buy one of their nice new multi-effects units to go with the nice new SY-300, GR-55, etc.
|
|
|
Post by Charlie Honkmeister on Apr 28, 2017 8:05:50 GMT -5
Greekdude,
Don't sell the GP-10 short; it doesn't take a back seat to any of Roland's GR series on pitch-to-MIDI performance. It's just miniaturized and put in a different price category (BOSS versus Roland VG series).
It doesn't have the onboard synth like the GR-55 does, because it isn't a guitar synthesizer - besides the pitch-to-MIDI, it has a COSM string remodeler which bases its sounds on the actual string vibration, not in converting to a MIDI or pitch signal which triggers a synthesizer.
The COSM modeling is capable of some synthesizer type sounds that have a high amount of tonal controllability by picking and damping techniques. But for certain instruments, keyboards in particular, the MIDI synth capability works better for having more variety of sounds available. So your choice of unit, say, a GR-55 versus a GP-10, would depend on where you want to go tonally, and what your needs are for let's say, playing live.
A single guitar with a GK interface can drive two Roland units at the same time, with a special cable, and several players do exactly that (GP-10 and GR-55 for example.)
Roland has always had somewhat of a problem with people understanding what the VG series (VG-8, VG-88, VG-99) did, versus the GR series, and the difference between COSM modeling -derived sounds versus synthesizer sounds.
Now there's a third category of Roland unit - the SY-300, which is a digital tracking synthesizer which doesn't require a GK 13-pin interface; it has enough computing horsepower to figure out the different string signals from a mono input, and re-synthesize tones from that, in real time.
A great resource for all of this stuff is the V-Guitar forums site - vguitarforums.com. This has forums on all the Roland guitar products, and other vendors' products as well, GK interfaces, guitars with built-in 13 pin capabilities, etc. There's a huge amount of information there .
|
|
|
Post by Charlie Honkmeister on Apr 25, 2017 7:18:18 GMT -5
Greekdude,
When you want to get into the whole world of guitar MIDI, etc. I don't think you can go wrong with the (Roland) Boss GP-10. With a GK-3 pickup it's about 400 dollars and has huge sound-making power on its own, and also includes guitar to MIDI that works very well. It also allows individual string outputs to a DAW so you can record and process with plugins, each string separately, as well as the combined output, all at once. That's only scratching the surface. I have one. It's amazing.
If you want probably the state of the art wireless MIDI to guitar interface, the Fishman Triple Play would be the unit to consider.
Just a couple of suggestions when you are ready to take the next step with some gear.
-Charlie
|
|
|
Post by Charlie Honkmeister on Apr 18, 2017 22:32:09 GMT -5
I had performed a test with a wire and some foil here guitarnuts2.proboards.com/post/78778/thread and it looks like you'd get about 10pF difference from proximity between bare leads and the shielded cavity. That might make a difference if you have an on board active tone shaping as you do, but in a standard passive set up, it's a drop in the bucket. The pickups I'm using are about in the 1.0 to 1.2 Henry range with a self-resonance of about 10 Khz at the end of the shielded cable. That's deliberate overkill for approximately an 8 Khz maximum resonant frequency when everything is connected, including a possible gain mismatch (not exactly 1) on the buffer so the capacitance isn't completely cancelled out at maximum pot sweep, so I'm allowing for some extra capacitance that way. Generally I see anywhere from 150 to 250 pF , all pickup coil wires shorted together and capacitance measured from that point to ground, from the commercial pickups I've tested. Bill Lawrence L-90's are the exception with about 90 pF (no shielded cable, just 5 separate insulated wires from the pickup.) So about 10-20 pF from the pickup switch and maybe a bit of additional wiring to the buffer board isn't much of a problem. If I were shooting for 20+ KHz resonant frequency, the pickup inductance would have to be about 550 to 650 milliHenries, stray capacitance would be a lot more critical, and the pickup output would be about 6 dB less, so we would be looking at about a 12 dB preamp to bring the signal back up. But 8-10 Khz maximum works very well for getting variable resonant frequencies from 1.2 to 5+ Khz to cover most classic tonalities we know and love.
|
|
|
Post by Charlie Honkmeister on Apr 16, 2017 20:17:52 GMT -5
Just as a reminder, capacitance is inversely proportional to the square (or cube, don't remember which) of the distance between the conductors. So shielded cables which have a relatively large distance between any center conductors and the shield foil or braid, inherently will have less capacitance. You can see that in RF coax cables, such as RG-59U, where high capacitance would mean unacceptable loss of signal at high (RF) freq uencies.
Because of the distance effect, if you wanted to gild the lily on low capacitance, all signal wiring should be run in the center space of a shielded control cavity and as much as possible not run next to the "walls" or back cover. Since the capacitance drops off very quickly starting at fairly small distances between conductors, you won't buy yourself much reduction in capacitance that way. But sometimes neat, centralized, bundled wiring with not much excess in the control cavity, can make a small difference, especially with certain styles of clean playing.
In all the cases relevant here (coax, twisted pair, two parallel wires, one wire over ground plane) the capacitance per unit length is not dependent directly on spacing, but rather on the inverse of the logarithm of a ratio. The ratio is spacing to wire radius, where spacing has the relevant interpretation for each case. It also depends directly on the dielectric constant of the insulator. There are two consequences: 1. You can not change by huge amounts because the log function varies slowly with change in the ratio. 2. If low capacitance is your goal, using small wire is just as important as large spacing. For rf coax cables, the goal is a characteristic impedance allowing efficient transmission. This characteristic impedance is the square root of inductance per unit length divided by capacitance per unit length. Very low capacitance is not required. Coax cable designed for low loss usually has an impedance of 50 ohms. Thanks, Mike, believe you are correct on all counts. But I did want to give a very general idea of what is involved here. Introducing characteristic impedance of a transmission line, dielectric constants, etc. wouldn't make a lot of sense to the non-EE or non-physics crowd. Generally guitar wiring (including the center conductor of most cables) is in the AWG 18-24 range and the absolute difference in radius of the conductors of wires in that range I wouldn't think would matter very much (not even close to an order of magnitude difference per unit length), especially since most wire is insulated and the insulation is on the same order of thickness as the wire itself, limiting the minimum distance possible between conductors. If you have a link to a formula reference for this effect, it would be great to share for those who want to chase this down in more detail using real data for insulated wire specs. -Charlie
|
|
|
Post by Charlie Honkmeister on Apr 16, 2017 9:56:02 GMT -5
I was wondering how this wouldn't make that situation similar to that terrible Monoprice cable in this photo, where you have lead wires enclosed inside some shielding, too. Maybe the fact that the 4 conductor wires of that pickup are so thin makes the shielding and the lenght irrilevant...? Thank you very much in advance!
The four conductor is higher capacitance, but it's only about a foot, which is not as bad as having it the whole length of a guitar cable. It's more interesting than that, though. Suppose you have the humbucker wired in series, you have two of those wires tied off at the other end of the cable. The signal is between the two coils, and subject to that capacitance, so you effectively have a capacitor to ground in between the two pickups, creating a tiny secondary resonance. It's so small, and of a high frequency, that you can't hear it, but you can see it in bode plots. Four conductor shielded pickup cables have quite a bit of capacitance to ground. It isn't normally a big issue since in most cases, the cable capacitance is the largest value and mostly determines how far the pickup resonant frequency shifts down. 20-60 pF won't matter much in resonant frequency when you are using a 15 to 20 foot cable with anywhere between 650 pF and 1200 pF capacitance. Just FYI, if one were to minimize pickup wiring capacitance, twisted pairs or just unshielded wires (as in some Bill Lawrence pickups) would work in a shielded control cavity. Just as a reminder, capacitance is inversely proportional to the square (or cube, don't remember which) of the distance between the conductors. So shielded cables which have a relatively large distance between any center conductors and the shield foil or braid, inherently will have less capacitance. You can see that in RF coax cables, such as RG-59U, where high capacitance would mean unacceptable loss of signal at high (RF) freq uencies.
Because of the distance effect, if you wanted to gild the lily on low capacitance, all signal wiring should be run in the center space of a shielded control cavity and as much as possible not run next to the "walls" or back cover. Since the capacitance drops off very quickly starting at fairly small distances between conductors, you won't buy yourself much reduction in capacitance that way. But sometimes neat, centralized, bundled wiring with not much excess in the control cavity, can make a small difference, especially with certain styles of clean playing.
If you lose high end definition before the pickup signal even gets to the amp, just turning up the treble doesn't always work to get the sound you like for clean playing.
I'm having to deal with this myself because I'm doing electrically variable capacitance for pickup voicing. If the inductance of the pickup is relatively high, I have to go to really low capacitance values to get a good resonant frequency range and be able to get the resonant frequency as high as I want for hi-fi or Fender/Gretsch tone from the pickup. Then stray capacitance in the wiring can be an issue, even with a buffer to eliminate the effect of the cable and amp capacitance.
|
|
|
Post by Charlie Honkmeister on Apr 14, 2017 10:25:43 GMT -5
Hey Greekdude,
As far as starting from scratch with MIDI on Windows, you need to start with some basic tools to see what is going on with the interface.
Check out: MIDI-OX - midiox.com . Also, MIDI Yoke is useful for distributing MIDI to/from multiple programs running at the same time. That might be later on though.
Next, a freeware/shareware DAW program to record and play back MIDI. Reaper is probably the best known and loved of this type of program but there are many out there.
Your Yamaha keyboard has a polyphonic MIDI synthesizer onboard with hundreds of sounds. As is, it can play back General MIDI (GM) files with multiple instrumental parts, drums, etc. All you need to do is to send a General MIDI file to it over the MIDI interface. The DAW can do this, or you can use a dedicated MIDI file player to do that. There are thousands of GM MIDI song files on the Internet if you want to play with those for a while.
The DAW program is the missing link to allow you to use MIDI and audio VST plugins. From there the sky's the limit on plugins, including soft synthesizers, including probably the Hammond organ sounds you crave.
Just a side note, there is a program, VSThost, which will allow you to use one or more plugins and allow you to play the plugin directly from your keyboard, without the overhead of a full fledged DAW.
Hope that gives a couple of useful directions.
Take care,
Charlie
|
|
|
Post by Charlie Honkmeister on Apr 10, 2017 12:39:17 GMT -5
Nice work on the coil, Antigua.
|
|
|
Post by Charlie Honkmeister on Apr 3, 2017 20:44:21 GMT -5
That makes sense, but won't the "R" function of the LCR meter do something similar? I'm not aware that there's a distinction between AC resistance and impedance.
It's the same thing for our purposes, I think. Now that you pointed that out, you can read AC resistance directly with your Extech, but only at the test frequency you selected (100/120 Hz, 1 Khz, 10 Khz, etc.) The impedance/AC resistance at 1 Khz is probably good enough for a quick sanity check of the coil impedance, but the formulas I gave will work for any frequency.
I'm not familiar with the Extech since I have a DER EE DE-5000 LCR meter, but they basically would do the same thing.
I would agree with MIke that #41 AWG is probably too skinny to let you get good current through the coil. You would have to use a lot higher voltage drive signal to get that current, and would actually run into resistive heating effects you didn't want.
|
|
|
Post by Charlie Honkmeister on Apr 2, 2017 21:15:31 GMT -5
LM386 Audio Amp BoardOn boards with an LM386, the restriction is that you have to have between 4 ohms and 32 ohms impedance load. It wouldn't be too much trouble to get that but it doesn't really match the coils you have already. The really low inductance of your two coils suggests you need a lot more turns for good magnetic field strength to couple to the string, which is a small target to intercept and react to flux. Wind a Strat coil with #30 to #32 AWG and end up with about 4 ohms DCR and you should end up with 250-450 turns and a lot more inductance. This would make a more suitable load either for the Pyle amp you are thinking about, or something smaller like a LM386. We could do this a bit more analytically with Salvarsan's coil calculator and some formulas, but you get the idea. The other idea is to put slugs, like humbucker slugs, into the plastic Strat bobbin that you wound the coil with, and both increase the inductance and focus the field. You could probably do that with the Strat bobbin exciter coil you have already, and it might work OK that way without having to wind anything new. I can probably make a little bobbin with a steel slug, and wind it up pretty good. I have an Extech 380193 LCR meter, could the resistance mode be used to determine if I've achieved 4 ohms impedance with a combination of winds and steel core? 120Hz or 1kHz test frequency? My understanding is these impedance ratings tend to be imprecise anyway. Since you have an Extech, finding the impedance of an inductor at a certain frequency is pretty easy. Here's a reference Web page. www.arca53.dsl.pipex.com/index_files/elecprop.htm1. Measure the DC resistance. 2. Measure the inductance at 100 Hz or 120 Hz test frequency. Find the inductive reactance with the formula X(l) = 2 * pi * F * L , where F is the frequency you wish to find the impedance at, and L is the measured value of the inductor. You can pick 100 Hz as a reasonable starting value for F. since the reactance will rise as the frequency is increased (inductive load.) 3. Use the formula Z= sqrt ((DC resistance) ^2 + (inductive reactance from step 2) ^2). That's the impedance at the frequency you picked. 4. Plug numbers in again, for a frequency of 1 Khz and see if the impedance at that frequency is "reasonable" for the power amp you're using. (usually 4 to 16 ohms.) Probably could be above that range even, but power transfer to the coil would be limited in theory if there's a big mismatch between amp output impedance and coil impedance. Impedance measurement can be complicated with an actual electromechanical transducer like a speaker, because there's complex impedance changes and back-EMF caused by energy storage and dissipation in the moving cone, surround, etc. But for a simple coil with no moving parts, things can be pretty precise.
|
|
|
Post by Charlie Honkmeister on Apr 2, 2017 14:42:21 GMT -5
Exciter coils shouldn't be a big issue. Just use the same one as you use for pickup testing. If you had a problem with 10 volts barely moving the string, the problem was current, not voltage. You have to get good current drive for the thang to work. The one I use is a plastic Strat bobbin without magnets/poles, and I had about 200 turns of #32 AWG on it. If you know the inductance and resistance of the exciter coil we could figure out the impedance. Parts Express and SparkFun have small speaker level power amp modules. Realistically, any small amp PCB capable of over about 100 milliwatts into 4-8 ohms should be fine. I have a coil I'd like to use with a DC resistance of 52.8 ohms, and the inductance is 0.67 mH, it must be about 150 turns of 41 AWG. That one is ideal because it's small and would be easy to position, but I also have a plastic Strat bobbin coil with 3.2 ohms resistance and 0.49mH inductance, which is much larger, about 70 turns of 26 AWG. I'm slightly worried about burning out the 41 AWG coil by exposing it to too much current. What do you think about a little amplifier like this www.amazon.com/Pyle-PCA1-30-Watt-Stereo-Amplifier/dp/B0012KZNP4/ref=sr_1_4?s=electronics&dd=lydIXCa_3W4e95ktHPGr5g%2C%2C&ie=UTF8&qid=1491065510&sr=1-4&keywords=small+audio+amplifier&refinements=p_90%3A8308920011 ? What I was thinking about is something like: LM386 Audio Amp BoardOn boards with an LM386, the restriction is that you have to have between 4 ohms and 32 ohms impedance load. It wouldn't be too much trouble to get that but it doesn't really match the coils you have already. The really low inductance of your two coils suggests you need a lot more turns for good magnetic field strength to couple to the string, which is a small target to intercept and react to flux. Wind a Strat coil with #30 to #32 AWG and end up with about 4 ohms DCR and you should end up with 250-450 turns and a lot more inductance. This would make a more suitable load either for the Pyle amp you are thinking about, or something smaller like a LM386. We could do this a bit more analytically with Salvarsan's coil calculator and some formulas, but you get the idea. The other idea is to put slugs, like humbucker slugs, into the plastic Strat bobbin that you wound the coil with, and both increase the inductance and focus the field. You could probably do that with the Strat bobbin exciter coil you have already, and it might work OK that way without having to wind anything new.
|
|
|
Post by Charlie Honkmeister on Apr 1, 2017 11:03:04 GMT -5
I'm having luck vibrating the string via sympathetic resonance using this audio exciter www.amazon.com/gp/product/B004AR5G9O/ref=oh_aui_search_detailpage?ie=UTF8&psc=1 What I have to do is set the guitar on "feet" so that the vibration is retained in the body and neck of the guitar, otherwise the table the guitar is placed on eats up too much of the energy. I rested the head stock directly on the exciter device, which delivers vibration energy to the guitar a lot more efficiently than attaching the device from the top, which is more how it's intended to be used. With the exciter tuned to the exact frequency of the string, it gets a pretty rigorous vibration, and the nice thing is that it's a very pure string movement, extremely still looking with no inter modulations of any sort. You can tell that it's favoring fundamental movement and not exciting much harmonic movement, but there is enough harmonic content there to be useful. I also tried an exciter coil, but at 10 volts it was just barely moving the string, so I'd have to either get a stronger amp, or make a better coil. One annoying thing is that when the exciter coil's core was magnetized, like a pickup, this causes a lot more string movement, but then the introduction of the magnetic field sort of spoils the experiment, though I could see string pull over dead center as be "neutral" in that it's pull is perfectly symmetrical. Exciter coils shouldn't be a big issue. Just use the same one as you use for pickup testing. If you had a problem with 10 volts barely moving the string, the problem was current, not voltage. You have to get good current drive for the thang to work. The one I use is a plastic Strat bobbin without magnets/poles, and I had about 200 turns of #32 AWG on it. If you know the inductance and resistance of the exciter coil we could figure out the impedance. Parts Express and SparkFun have small speaker level power amp modules. Realistically, any small amp PCB capable of over about 100 milliwatts into 4-8 ohms should be fine. If the audio exciter on the headstock works for you , that's great. Your own Sustainiac Model C, in a way.
|
|
|
Post by Charlie Honkmeister on Mar 31, 2017 12:53:18 GMT -5
Since the E-Bow itself has a pickup, what's being picked up, amplified, and re-injected into the string, depends on the E-Bow pickup position on the string. So you already have an initial signal having a spectrum with peaks and nulls, before you magnetically transduce a higher power version of that signal back into the string at a slightly different location. If you want to eliminate possible harmonic level inaccuracies because of the Ebow, it might make sense to use a piezo bridge saddle signal which has all harmonics present, and use it to drive a driver coil positioned at exactly 1/2 the speaking string length (first antinode.) If you take things to the point that you want to see what happens on string pull versus magnetic field strength on an actual picking stroke to the signal output of the string, things will get complex in a hurry. It would be fascinating to go there, but you would probably need an automatic picker, or a "pick-bot" to do that. The frequency of a G string is 196Hz, is there some way to skip the feedback loop altogether and exciter the string into vibration with a function generator and coil? I'm not sure how much power it would take, or what kind of driver could would be ideal for the task. An EBow seems to benefit from the fact that it's causing a feedback loop, so I imagine an arbitrary excitation would require more power, and the string would need to be tuned very precisely, so that it's natural frequency is complimentary of the induced magnetic frequency. The difference from the EBow is that the hopefully the function driven exciter would only reinforce the fundamental and some odd harmonics, where as the EBow reinforces everything it sees. It wouldn't be a problem to use a function generator signal to tune to the string frequency, at least for a very short time. Or, tune the string to the frequency generator. The problem is that you would have phase and frequency drift which would slide the exciting signal in and out of phase with the vibrating string. You would have to have a phase and frequency correction loop somewhere. That isn't necessarily a huge deal with some of the analog chips out there (e.g. CD4046 PLL) but maybe a bit much electronics hacking for the main thrust of what you're trying to do. Driving from a piezo string signal at the bridge, or a magnetic string signal picked up very close to the bridge, would be the most practical. Just as a possibility, and something I haven't done yet, is to take a pair of stereo earbuds, get one earbud, and inject Superglue into the center diaphragm part of the earbud. That fixes the coil in place, and you should end up with a single-string low impedance magnetic pickup. The output will be low mic level but a "normal" mic preamp should have enough gain to bring the signal up. That way you can use Blue-tac or something to fix it under the string of interest, and you have your signal to drive the exciter coil, with some power amplification, of course.
|
|
|
Post by Charlie Honkmeister on Mar 30, 2017 12:35:21 GMT -5
Since the E-Bow itself has a pickup, what's being picked up, amplified, and re-injected into the string, depends on the E-Bow pickup position on the string. So you already have an initial signal having a spectrum with peaks and nulls, before you magnetically transduce a higher power version of that signal back into the string at a slightly different location.
If you want to eliminate possible harmonic level inaccuracies because of the Ebow, it might make sense to use a piezo bridge saddle signal which has all harmonics present, and use it to drive a driver coil positioned at exactly 1/2 the speaking string length (first antinode.)
If you take things to the point that you want to see what happens on string pull versus magnetic field strength on an actual picking stroke to the signal output of the string, things will get complex in a hurry.
It would be fascinating to go there, but you would probably need an automatic picker, or a "pick-bot" to do that.
|
|
|
Post by Charlie Honkmeister on Mar 30, 2017 8:26:08 GMT -5
Great persistence and nice work, Antigua. You didn't mention when you did listening tests with an amp, the cable you used and maybe an idea of its capacitance. If you used a typical cable length, the cable (even if it's good quality) and amp input Z may have affected the lower Z output Strat pickup less than the higher output Z P-90, when actually plugged in. (Just a WAG for why you had to add additional RC loading to the Strat pickup to match tonality.) It probably does owe to the guitar cable or amp input. The fact that every rig is different would make it very difficult to associate any set of R and C values with a particular outcome. Antigua, you know that you're just talking yourself, bit by little bit, into an active buffered approach, right? No worries, I like passive, too.
|
|
|
Post by Charlie Honkmeister on Mar 30, 2017 0:21:05 GMT -5
Great persistence and nice work, Antigua. You didn't mention when you did listening tests with an amp, the cable you used and maybe an idea of its capacitance.
If you used a typical cable length, the cable (even if it's good quality) and amp input Z may have affected the lower Z output Strat pickup less than the higher output Z P-90, when actually plugged in. (Just a WAG for why you had to add additional RC loading to the Strat pickup to match tonality.)
|
|
|
Post by Charlie Honkmeister on Mar 28, 2017 22:11:18 GMT -5
The other approach that might work out for a single coil is possibly a tapped pickup where the full coil is a lower resonant frequency (P90) and the fractional coil is more traditional Strat. That's done on both coils in a humbucker, by the Joe Barden Two Tone. It does get a pretty nice single coil sound in "switched/tapped" mode, without much of a perceptible volume drop. Speaking of, I discovered (insofar as I've never seen it mentioned elsewhere) that tapping causes huge parasitic capacitance. The unused portion of coil couples with the used half. With the SD SSL-4 I measured something around 600pF additional capacitance caused by the unused coil. In turn, SD makes the tap point really low, so that you have a very low inductance combining with a very high capacitance, for an overall low output. Add to that the sharper 3rd order slope of an external capacitance, and it's not wonder the tapped tone is so unpopular. The solution there is to disconnect the tapped coil at both ends. This requires 4 conductor cable, while most tapped pickups have only three conductors. It would also require a DPDT instead of an SPST switch, which is not a real big deal. If this was done, you could have a tapped single coil with two (or more) convincing voices. If I owned and operated a pickup company, I would have done this yesterday. I remembered that you (I believe) had posted on the unused part of a tapped coil before. It makes sense that that that would be the case for the unused windings. The Joe Barden basically just shorts the two tap points, one on each humbucker coil, together. I have no idea why this wouldn't create response drop or high frequency loss, but it does work sonically. I have a HB Two/Tone bridge that I can send you if you like at some point, to test. FWIW, Gibson has a few patents on a bifilar pickup coil. Basically, wind the whole thing with 2 wires laid together instead of one. Besides creating a very exact 50% tap point if you wanted it, you could disconnect the second winding and because the second winding is not active, the effective spacing between and around the remaining winding that is working is increased, and that would reduce interwinding capacitance somewhat. Also, by passive loading of the second winding, either connected or not connected to the first winding, it might be possible to tailor the PU response in a more flexible way. Gibson Bifilar Coil Patent
|
|
|
Post by Charlie Honkmeister on Mar 28, 2017 15:53:33 GMT -5
The other approach that might work out for a single coil is possibly a tapped pickup where the full coil is a lower resonant frequency (P90) and the fractional coil is more traditional Strat.
That's done on both coils in a humbucker, by the Joe Barden Two Tone. It does get a pretty nice single coil sound in "switched/tapped" mode, without much of a perceptible volume drop.
|
|
|
Post by Charlie Honkmeister on Mar 28, 2017 15:37:40 GMT -5
Antigua,
Oops, forgot that it was you who did the excellent LTSpice tutorial , which I have turned multiple people on to. Mea culpa. I really appreciate your work on that.
I don't think it's nuts at all to want to get a P-90 sound in a Strat pickup, especially bridge position; several makers have tried to go this way to fatten up the midrange honk and make the PU better suited for crunch and distortion.
The ideal passive solution would be to switch components in and out to get the maximum versatility. But you're going to run into inconsistency in output levels with all passive when you switch voicing, and you still have the cable capacitance to deal with. That's what finally drove me to active buffered.
If you can tolerate lower output level in general, if you go for a lower inductance, lower DCR pickup, the job is easier because the cable capacitance will affect the pickup circuit less and allow you more tweaking range on setting resonant frequency and cutoff slope.
|
|
|
Post by Charlie Honkmeister on Mar 28, 2017 14:44:57 GMT -5
The series resistor affects the Q and the HF slope. The parallel resistor affects the Q, but not the HF slope in the high frequency limit. You can start outside the guitar with pots (instead of fixed resistors) for a given value of C. It will take some back and forth. I have mostly just done parallel resistors; it seemed good enough to me, but it did seem that there was more that could be done. That's a good idea, I'll give that a try. Even if it works out though, it sort of changes the game; we can't just say "instead of buying a new pickup, just put a cap across it", because it would turn out that you need a special recipe of capacitance and resistance to get to a particular goal. That's not entirely a bad thing, but it's much less simple and and straight forward, and would spawn a whole new pastime of figuring out pickup mod recipes. Antigua, with a EDA design program (LTSpice, Circuitlab, etc.) it's easy as pie to tweak a value and do a simulation run to look at the response. Since JohnH has confirmed that pickup models in general do work well to a reasonable level of accuracy, given that the design in question has "reasonably" low eddy losses, it's perfectly legitimate for experimentation to use a modeled pickup with parameters which are derived from real measurements of the pickup. I've been experimenting in this way for approximately two years, on the way to developing the electronically variable capacitor/buffered approach I have been talking about on posts. Specifically, what I've come up with for the input network is: where the pickup DCR is lumped into the simulation model of the inductor L1. So, ignoring C2, which is the cap for the variable capacitance feature, besides inductance, we are only looking at pickup DCR, stray capacitance (C_pickup), voicing capacitance (C1), series resistance (R1) and shunt/parallel resistance (R2) to provide voicing for the pickup, which covers resonant frequency and Q value before rolloff. I'm ignoring amp input resistance and capacitance in this model because I'm feeding a JFET buffer with 10 Megohms input impedance, and very low capacitance, with it. If this looks like JohnH's 4 and 6 component models, that's because we ended up in the same place. But we're interested in modeling the whole instrument, cable and all, alll the way to the amp input, to relate the component changes anywhere in the chain to how the instrument will sound. Just an extremely simple model like this will give lots of experimentation, and you already have some of the answers to the question, "What do I make it sound like?" since you have been doing some great work on testing and publishing pickup electrical parameters. From building, modding, and testing, I came up with some general observations: 1. Resonant peaks on the order of 5 to 8 dB provide familiar tonality and are closest to emulating the RC loading of a passive instrument connected to a cable with significant capacitance. This is true for the entire resonant frequency range of interest, approximately 1.4 KHz through about 4.8 KHz. You might need Q in the upper end of the 5-8 dB range in certain situations like, for example, emulating a particulary twangy Tele pickup. Parallel/shunt resistance (R2) is the most important controlling value for setting this Q value. 2. Series resistance (R1 in the pic) is needed to lower the Q into about the same 5-8 dB range, at low (less than 2 Khz or so) frequencies where the shunt capacitance is high. I'm sure that Mike's (ms's) observations are valid as well, and with a circuit simulator you can get very rapid feedback as to how a component change affects the response. Also, don't forget that with LTSpice you can run a simulated signal through the simulated circuit, and get output data files. I've heard this is pretty slow, though. I haven't particularly looked at the real-world slope of the rolloff above resonance but hope to do so when I set up a test bench similar to yours (Circuitgear CGM-101, driver coil, Stratotarts integrator board, etc., all of which I already have on hand. ) What you have here is very interesting and I'd like to do some experimentation like yours to see how the rolloff slope affects tonality. -Charlie
|
|
|
Post by Charlie Honkmeister on Mar 19, 2017 21:01:06 GMT -5
There are many simple ways of constructing a higher frequency pickup system. The simplest is to use the parallel configuration of a humbucker. Actually, you could use a high impedance preamp with a suitably chosen resistive load, and get an almost flat response up to the unloaded resonant peak of most existing pickups. That would get you up to 10k without breaking a sweat. A low impedance, low inductance coil can also get you there. But when you look at sonic preferences, most players dislike the tone of high fidelity (high frequency) pickups. For example, everyone raves about the Cavalier Lion. Well the frequency response is one of the lowest of any Tele bridge out there. Yep, absolutely agree with this and that is the approach I'm playing with now. A 4 Henry humbucker in series coil mode becomes a 1 Henry humbucker in parallel coil mode, with a self-resonance over 10K depending on stray and cable capacitance. Four-wire "vintage PAF" sets such as the SD Pearly Gates are right in the strike zone. So are quite a few single-coil sized rail humbuckers. Combine that with onboard variable load capacitance and reasonable resistive loading, into a high impedance buffer, and you get amazing tonal versatility with a twist of a pot.
|
|
|
Post by Charlie Honkmeister on Mar 19, 2017 19:13:49 GMT -5
I'll do what I always do, which is to share anecdotally but I can't give a 360° tour of a multi-year R&D project funded by companies who have earned the right to do so over decades of discovery and fair commerce. (I'm not suggesting that's what YOU are asking, just that it's a common theme) To answer the question directly, Fluence is the most obvious project, in that one thing I've mentioned publicly in the past (without revealing final iteration data) is that the original Fluence core had a resonant peak of 80kHz. I have the full spectrum at my fingertips during the voicing process. I can literally listen to the difference 1dB at 10kHz makes in realtime. Same is true of 16k, et al. Now, if you run a simple plot of a Fluence vintage Strat voice against a 54 Strat pickup they'll be pretty similar. Whether or not subtle differences are meaningful, or whether someone thinks they're not worth spending money is up to the consumer. But they're not imagined or conjured. As far as Zephyrs were concerned, we felt the subtle differences did manifest themselves north of the resonant peak, and in the time alignment/group delay. They will also not provide any chart or plot that verifies it and I do not work for them anymore. What do you mean when you say a core has a resonant peak of 80KHz? Mike, I understood what Frank meant. The Fluence is a multi-layered PCB coil with printed conductors, and is a low impedance pickup. Given the 80 KHz self-resonant frequency, the Fluence core pickup coil assembly (regardless of hum cancellation technique used) probably is roughly on the order of 20-25 mH and has a very low stray capacitance of about 60-90 pF. This means an overall system preamp gain of about 32 dB or so is needed for adequate output level. I've wound and experimented with conventional low Z pickups in very close to this range so have a ballpark idea of the numbers involved in the Fluence. Obviously Frank is under NDA , so I'm sure he couldn't comment. But it is certainly true that having a self-resonance like that, well above the audio range, allows a dead flat audio frequency (and well above) response, which can be voiced by EQ on the preamp board any way you please, given a good enough S/N ratio in the preamp. But AFAIK all the Fluence products only have two pre-selected voicings and you can't physically access a point which is just preamped and doesn't have the EQ applied. But during R&D they had the whole thing breadboarded instead of one epoxied unit, and could play with any EQ they wanted. There's actually a reason I asked Frank about the "over 5Khz" response. I'm still on the fence about whether anything over flat response to 8 Khz to 10 Khz is necessary for a magnetic guitar pickup. I'm intrigued by Frank's statements that it does matter and is audible. Currently I'm in the medium Z design space without preamping, and with about a 10-12 KHz self-resonance, at the expense of only about 6 dB of reduced output. It wouldn't be very hard to go lower on inductance, preamp about 12 dB or so, and get flat response to 22 or even 25KHz. So regardless of what Frank can or can't disclose, or what we can actually test in those products, it's still possible for someone to explore this question in detail.
|
|
|
Post by Charlie Honkmeister on Mar 18, 2017 18:57:52 GMT -5
Look, we all know where this is headed. I work for companies and I'm not posting graphs or screen shots, or going down too many rabbit holes with you guys. There are uninformed and downright false claims made in the name of science emanating from these kinds of homebrew experiments, and dismissive unbelief in other areas. You run these flawed experiments and then sprint to the nearest pickup forums to blast your newfound knowledge as if it's canon. It's killing me to watch some of you take this hobby of yours and turn it into a crusade. It's like I'm watching a kid first learn what salt and pepper taste like, and then they proceed to describe everything as "salty" or "peppery" to the entire world...until they learn about garlic. You pick and choose what you will listen to, such as an article by a dealer that says 5k+ doesn't matter in a guitar speaker, but you refuse to consider return path issues because the gauss meter zeros out at some point. Well, zero isn't zero on that meter. That meter is a bull in a china shop for what you're trying to determine. No, it's a bull in a dollhouse. It's Godzilla. If you think 5k is a cutoff, stick a 31-band EQ in the effects loop of your Mesa Triple Rectifier and boost 16kHz and see what freaking happens. Does nothing happen? Is the cabinet deaf to 16k? Map out the Q of the 16k fader and see how low it goes. The FACT is, I work with pickup technology where I can literally touch a button during the R&D process and toggle <1dB at 10kHz and everyone in the room hears the difference, and has an opinion about it. Of course It does different things with different amps, but for example I can pinpoint the exact moment on a Marshall where it goes from ear candy, to crashing the preamp. Yet some of you want to take what I'm saying as a lie, or as "unproven" because you're not on my R&D team and you didn't see it for yourself? I've done the string window studies. There are flaws in what you're going to conclude from this. You can either keep going and ignore what I'm trying to tell you, post up counterpoints to every one of my points, or consider them. Frank, string window and Gauss meter comments aside, do any of the commercial pickups/systems you have worked on, incorporate the stated and/or implied results of your R&D involving either direct audibillity of signals in the 5Khz-15KHz+ range, and/or interaction of those frequency components with tube guitar amplifier preamp, intermediate driver/phase inverter, or power amp/speaker stages? I'm only aware of your design input on the Seymour Duncan Zephyr and Fishman Fluence products. Would it be possible in theory to test one or both of these products in some form factor, and be able to verify what you are saying?
|
|
|
Post by Charlie Honkmeister on Mar 7, 2017 22:17:14 GMT -5
Antigua,
Some of the low end Gibson guitars (2017 Les Paul Custom Special, Firebird Zero, M2, SG Fusion) are coming with Gibson "double slug" DS-whatever pickups and from the specs on Gibson's site, they are in the vintage range. I'd be willing to bet they are Asian manufacture.
I don't see being able to come up with any short of buying them, but I'm going to keep my eyes open on Ebay since changing the PU's would be probably in the sights for some of the buyers of these guitars. I think from a Youtube video of the M2 control cavity that the PU's have the Gibson single Molex connector.
If any come my way I'll be sure to PM you.
|
|
|
Post by Charlie Honkmeister on Feb 16, 2017 20:28:32 GMT -5
I doubt that you get the direct eddy current affect right unless you use a coil causes a field very much like the field the string makes. It wouldn't be too hard to get close, or at least, consistent enough to make meaningful measurements. If you could get the right signal power level to the coil and use the right drive waveform, maybe an artificially tweaked one with emphasized high frequency content, I think you could make meaningful eddy measurements. It would be worth experimenting with.
|
|