Author |
Message |
Inventor
Stream Operator
Joined: Oct 13, 2007 Posts: 6221 Location: near Austin, Tx, USA
Audio files: 267
|
Posted: Sat Dec 27, 2008 3:54 am Post subject:
Alternate Guitar Pickup Concepts Subject description: Getting weird here... |
|
|
I got to thinking this morning, which is always dangerous, and I started wondering about alternate ways of doing guitar pickup based on using the strings themselves.
The string is a conductor, about the length of an FM antenna. When it vibrates, it's length changes somewhat due to the string stretching, or so I would think. Therefore one might imagine that if the string were part of an FM-ish frequency oscillator, the oscillator's frequency would change as the string vibrates. This change would be subtle but possibly enough to detect. In fact such a tuned circuit might produce an FM radio signal that is transmitted from the string itself.
I did a quick google search and sure enough someone has a patent on transmitting from guitar strings, but what I am thinking of may be somewhat different than simply using the string as an antenna. For example, the pickups have small magnets in them, so there must be a tiny voltage induced across the length of the string.
Let's say we took an opamp and put a large resistor in the feedback loop, then connected its inputs to two adjacent strings just above the nut. When either string vibrated, it would produce a voltage that would be amplified by the opamp. With enough gain, such a voltage might be detected. This would eliminate the need for coils, replacing them with opamp circuits.
I don't know where I am going with this train of thought, but we can definitely observe that the train has left the station, haha! Thoughts? Ideas? Impossible? Impractical? Been done before? You tell me. _________________ "Let's make noise for peace." - Kijjaz |
|
Back to top
|
|
|
DrJustice
Joined: Sep 13, 2004 Posts: 2114 Location: Morokulien
Audio files: 4
|
Posted: Sat Dec 27, 2008 1:41 pm Post subject:
|
|
|
Hm... Indeed, why aren't we using a fixed magnet and just picking up the induced currents in the strings? I'm sure there are good reasons. One might be that bare fingers touching the strings would disturb the minute currents in the strings. Today we can get coated strings (Elixir), but capacitance can't be helped by that in case it matters. Perhaps there's simply less noise, trouble and overall better performance using coil pickups? Or it could be that the olden days, when the eletric guitar was invented, the electronics weren't up to it in compactness and/or performance, so nobody ever tried it? There's also the factor of the general world of guitar avoiding modernity and innovation like the plague; everything must be exactly as it was on the first day the guitar was invented Well, at least is has to be Vintage, '57, '59, PAF etc..
You mention picking up the difference between two strings. Could the strings just be connected at one end for a common reference, and then the current/voltage changes picked up for each string at the other end? I wonder how strong the signals are. I want to know now - someone please rig a test circuit for this
Another possibility is to use an optical pickup...
DJ
-- |
|
Back to top
|
|
|
Inventor
Stream Operator
Joined: Oct 13, 2007 Posts: 6221 Location: near Austin, Tx, USA
Audio files: 267
|
Posted: Sat Dec 27, 2008 2:05 pm Post subject:
|
|
|
DJ, the reason I mention the potential between two strings is that there is no ground wire coming up from the bridge unless you run one up. So for an unmodified guitar you get a complete circuit from one string to another. Of course, they are all connected to each other at the bridge, so the bridge should perhaps serve as ground. I dunno, whatever.
Just since you asked, I put my scope on the 5mV setting (the minimum one) and connected the probe's ground to the low E string and pressed the probe tip to the A string, both just above the nut. My scope shows a wide noise band at this setting, but when I plucked the string a slight variation in the noise band occurred. I'm guessing somewhere in the ballpark of 100 uV signal. That's enough to boost with an opamp and send it down the cable to the amp.
It could also be vibrations in the way I was holding the probe, so let's not get excited yet. I don't have a probe with a clamp tip (no budget for one), so again I dunno. But it certainly is possible that this works and we now have a new way to get signals from the guitar. Wouldn't that be kewl? I really enjoy little moments of discovery and excitement like this, even though I know and recall that most of the time something practical comes up to spoil the fun.
Would anyone else care to try this out? Maybe with a scope or maybe just running signal into an amp? I'll try soldering some alligator clips onto a 1/4" panel jack and testing the amp theory if I get motivated later. But first I've got some beer to enjoy! Cheers and bottoms up, lol! _________________ "Let's make noise for peace." - Kijjaz |
|
Back to top
|
|
|
Inventor
Stream Operator
Joined: Oct 13, 2007 Posts: 6221 Location: near Austin, Tx, USA
Audio files: 267
|
|
Back to top
|
|
|
DrJustice
Joined: Sep 13, 2004 Posts: 2114 Location: Morokulien
Audio files: 4
|
Posted: Sat Dec 27, 2008 4:32 pm Post subject:
|
|
|
Great!
I just tested with a scope, and I get almost 5mV peak to peak. That's using the bridge as ground and measuring the strings by the nut. There's almost as much noise on the scope, but the long dangling measurement leads are probably the main reason for that. As your sound clip reveals, there's actually a good amount of signal above the noise.
For more power, maybe the winding on the strings should be isolated...
DJ
-- |
|
Back to top
|
|
|
Inventor
Stream Operator
Joined: Oct 13, 2007 Posts: 6221 Location: near Austin, Tx, USA
Audio files: 267
|
Posted: Sat Dec 27, 2008 4:51 pm Post subject:
|
|
|
Hehe! Thanks for testing it, DJ. Some random thoughts:
1. I checked and the E string is no louder than the D/G string, so shorting the strings together does not lessen the effect. This means that we can just short all the strings at the neck and run a ground wire down to the bridge.
2. My guitar produces something around 40 mV peak to peak when played, so your measurements indicate that with your connection method (better than the one i tried) we are only a factor of 10 lower in signal.
3. Bigger magnets would probably boost the signal by a factor of 10.
4. Insulated strings might help with the plinky sound.
5. The plinky sound may be due in part to the fact that we do not have a coil's resonance lowering the upper harmonics. So we are hearing the true signal of the strings and not the pickup-distorted one.
6. Filtering would restore the sound quality possibly.
7. The plinky sound may also be due to the finger resistance. Maybe I'll try a test with gloves or plastic covering the strings.
I'm not sure what to do next, I'm open to suggestions. Wow, this is interesting! _________________ "Let's make noise for peace." - Kijjaz |
|
Back to top
|
|
|
Inventor
Stream Operator
Joined: Oct 13, 2007 Posts: 6221 Location: near Austin, Tx, USA
Audio files: 267
|
Posted: Sat Dec 27, 2008 5:59 pm Post subject:
|
|
|
Just a couple of other comments:
1. When I wear insulating gloves, the plinking does not go away.
2. I think the plinking may have something to do with the fact that the frets are conductors.
3. I have the guitar in DJ's configuration with wires shorting the tuning posts together and a ground wire running from the headstock to the base of the guitar where the cable is. It works better I think, only need to set the amp on 3 with max distortion gain.
4. The amp is picking up a local rock and roll radio staion, haha! Must be all that string antenna action plus the rectification from the high gain distortion. It just gets one station.
5. Maybe the frets need to be insulated or maybe they need to all be grounded. I'll experiment later. _________________ "Let's make noise for peace." - Kijjaz |
|
Back to top
|
|
|
DrJustice
Joined: Sep 13, 2004 Posts: 2114 Location: Morokulien
Audio files: 4
|
Posted: Sat Dec 27, 2008 6:11 pm Post subject:
|
|
|
Inventor wrote: | 3. I have the guitar in DJ's configuration with wires shorting the tuning posts together and a ground wire running from the headstock to the base of the guitar where the cable is. It works better I think, only need to set the amp on 3 with max distortion gain.
|
I only had the strings connected together at the bridge, which happens anyway since it's all metal. The measurements were done at the nut for each isolated string - thinking hexaphonic output.
DJ
-- |
|
Back to top
|
|
|
Inventor
Stream Operator
Joined: Oct 13, 2007 Posts: 6221 Location: near Austin, Tx, USA
Audio files: 267
|
Posted: Sat Dec 27, 2008 6:49 pm Post subject:
|
|
|
I was thinking the same thing, and I meant that but modified with the short. I was unclear in my brevity. I cast all blame upon the Budweiser corporation. _________________ "Let's make noise for peace." - Kijjaz |
|
Back to top
|
|
|
Inventor
Stream Operator
Joined: Oct 13, 2007 Posts: 6221 Location: near Austin, Tx, USA
Audio files: 267
|
Posted: Sun Dec 28, 2008 1:54 am Post subject:
|
|
|
I've given this some more thought. The plinking issue I believe is due to the frets shorting the strings together. If we view each section of the string over the magnet as a current source, which seems I think to be the correct model, then shorting two strings together at a fret gives the current a return path so that not much of the signal gets out to the amp.
Also, my method of shorting all the strings together has the same effect. This explains why DJ got 5mV signal while I am having to crank my amp up all the way on distortion setting to get an audio signal. The right way to do it is to put a current to voltage converter on each string. Then it is simply a matter of summing them together if that is desired, or running six conductors down the neck if hexaphonic output is desired (plus ground and power).
Also a simpler passive approach would be to just use six high-valued resistors, one per string, and tie them together at the neck. This would lose some signal due to return paths, but if the resistors were comparable to the input resistance of the amp then the signal loss would not be excessive.
I am putting Scotch tape over the frets to solve the shorting issue, though there remains the issue of finger resistance shorting things somewhat, plus if the player is grounded that's another return path or noise source. To resolve that we would use coated strings with the ends sanded to expose the conductive strings.
I am going to experiment with these ideas soon. _________________ "Let's make noise for peace." - Kijjaz |
|
Back to top
|
|
|
DrJustice
Joined: Sep 13, 2004 Posts: 2114 Location: Morokulien
Audio files: 4
|
Posted: Sun Dec 28, 2008 7:15 am Post subject:
|
|
|
Ah, yes. Those pesky metal frets are ruining things a bit. Changing to non-metallic frets (lute players use gut, nylon or carbon I read) is not a real option, at least for quick experiments. That's where Elixir strings would help. It's the only make of polymer coated strings I know - are there others?
As you mention, the coils in the pickups works as low pass filters. I suppose some EQ'ing is needed to get that traditional sound. Perhaps even using a small transformer in the output stage?
DJ
-- |
|
Back to top
|
|
|
Inventor
Stream Operator
Joined: Oct 13, 2007 Posts: 6221 Location: near Austin, Tx, USA
Audio files: 267
|
Posted: Sun Dec 28, 2008 9:22 am Post subject:
|
|
|
Thanks for the pointer to Elixir Strings, DJ. I haven't changed the strings on my guitar yet, so when I do, I know what to get. Should probably sand the ends off, yes?
The coil is actually a band-pass filter not a low-pass, DJ, and that's important because it affects tone greatly. There is the inductance and resistance of the coil, and also an inter-winding capacitance that makes it a tuned circuit. A "hot" coil has fewer windings and therefore a higher center frequency. All coils attenuate the fundamental so that it is lower in amplitude than the harmonics.
I looked at the waveforms on my scope and it was difficult to see due to 60 Hz noise, but I did not see the characteristic guitar shape to the waveforom. Probably this is due to the enhanced fundamental and upper harmonic frequencies. In other words, and according to theory, we are getting the true signal from the string, not a filtered one.
I suppose that guitarists who want a traditional "coil" sound will want to band-pass filter their output signal, which leads us to the possibility of a tone knob that allows you to set the "hotness" of the coil sound. Now you can emulate your favorite coil by setting a dial.
That's all I have for now, cheers! _________________ "Let's make noise for peace." - Kijjaz |
|
Back to top
|
|
|
Inventor
Stream Operator
Joined: Oct 13, 2007 Posts: 6221 Location: near Austin, Tx, USA
Audio files: 267
|
Posted: Mon Dec 29, 2008 10:01 am Post subject:
|
|
|
I called Fender to discuss this concept and find out if it is new or not and they told me that they don't accept unsolicited design proposals. They suggested that I speak to a patent attorney. What a let-down! I thought Fender was a progressive, forward-thinking company that embraced new ideas. So much for that concept. Bah, it's probably patented or impractical anyway. _________________ "Let's make noise for peace." - Kijjaz |
|
Back to top
|
|
|
DrJustice
Joined: Sep 13, 2004 Posts: 2114 Location: Morokulien
Audio files: 4
|
Posted: Mon Dec 29, 2008 12:32 pm Post subject:
|
|
|
Inventor wrote: | ...So much for that concept. Bah, it's probably patented or impractical anyway. |
Don't let Fender or any other big company get to you! Remember that traditional guitar designers are generally firmly stuck in retro land. I've found no patents or other references to this method of picking up the signals. If it's possible to extract a good sounding signal with a smallish circuit, I't would be practical. Maybe it was impractical before coated strings and compact electronics existed? It certainly bears thinking about and experimenting with. In the mean time you should apply for a patent if possible...
And re. the coil pickups, you're right of course, it's a bandpass response, but I'd expect the high slope to be quite a bit steeper than the low one(?). I guess an analysis of mag pickup responses vs direct string pickup reponses is needed next, so we can derive the required EQ'ing.
Heh.. I was thinking that this whole thing fits in nicely with my childhood conception of electric guitars: surely there's electricity in them strings and playing should give you a good jolt...
DJ
-- |
|
Back to top
|
|
|
Inventor
Stream Operator
Joined: Oct 13, 2007 Posts: 6221 Location: near Austin, Tx, USA
Audio files: 267
|
Posted: Mon Dec 29, 2008 1:11 pm Post subject:
|
|
|
Well, DJ, I also posted this to harmony central and justinguitar. Someone at harmony central found a patent that is a direct attempt to do just this very thing. It is patent number 7105731. Here is the google patents page on it:
http://www.google.com/patents?id=iQN7AAAAEBAJ&dq=7105731
The primary distinction here is that this patent involves an isolated bridge and sensing of each string independently with a return wire, while our approach has a grounded bridge. I don't know if that's enough to warrant a new patent or not, but you know my story. I'm so jaded by the business world and patents and all that that I'd rather avoid the whole thing. I've had more trouble and more problems over patent disputes than almost anything else in business.
I can, however, DIY myself such a guitar and share the information on how to do it with everyone, and I might do just that. After all, something doesn't have to be novel to make it a good DIY project.
I think you're right about the high slope being steeper than the low one, possibly due to the distributed nature of the coil. That would definitely complicate modeling.
Wow, another one bites the dust. Why do I get my hopes up for these things? I guess hope springs eternal as they say. Thanks for your encuragement, it helps a lot! _________________ "Let's make noise for peace." - Kijjaz |
|
Back to top
|
|
|
DrJustice
Joined: Sep 13, 2004 Posts: 2114 Location: Morokulien
Audio files: 4
|
Posted: Mon Dec 29, 2008 2:14 pm Post subject:
|
|
|
Ah, a recent patent even. My mention of patenting should have had a after it. In todays patenting climate, I'd say that US patent 7105731 is at a safe distance, prior art wise, from a common reference in the bridge and a hexaphonic output. But anyway, don't take my patent suggestion too seriously.
The method suggested in this thread is more practical than 7105731 since no modification to isolate the strings at the bridge is necessary. I also wonder about the AD/DSP/DA stage - there seems to be AD/processing/DA delays at every plug point soon - I don't know if that's really desirable. Or perhaps the needed EQ'ing is too much for a small analogue board. Anyway, if you choose to pursue it, I'm certainly interested in hearing about the results.
DJ
Edit: typo
-- Last edited by DrJustice on Mon Dec 29, 2008 4:23 pm; edited 1 time in total |
|
Back to top
|
|
|
Inventor
Stream Operator
Joined: Oct 13, 2007 Posts: 6221 Location: near Austin, Tx, USA
Audio files: 267
|
Posted: Mon Dec 29, 2008 3:14 pm Post subject:
|
|
|
Well, DJ, although I did *think* of hexaphonic output, you're the one who tried it and mentioned it in the thread. Also you were the first to actually *try* the grounded bridge and take measurements. To me this would seem to make you one of the inventors of the patent, a dubious honor indeed!
If you still feel that our work here is novel enough to warrant a patent application, and you'd like to be a half or partial owner, then please contact me via pm. You may, of course wish to decline and I'll understand.
I don't know if I'll pursue patenting, but if I do I think I will do it myself this time instead of paying attorneys. I just can't afford the attorney fees, but the filing fees - in full or in part - I could perhaps make budget for those though I'm not sure how. _________________ "Let's make noise for peace." - Kijjaz |
|
Back to top
|
|
|
seraph
Editor
Joined: Jun 21, 2003 Posts: 12398 Location: Firenze, Italy
Audio files: 33
G2 patch files: 2
|
|
Back to top
|
|
|
Oskar
Joined: Jul 29, 2004 Posts: 1751 Location: Norway
|
Posted: Mon Dec 29, 2008 5:24 pm Post subject:
|
|
|
I know that other manufacturers do various coated string solutions, whether they're suitable for Inventor's use, as regards insular properties, I wouldn't know.
Martin strings have special coated ones.
As do Cleartone, Black Diamond, DR, Ernie Ball, and probably quite a few others. _________________ Where there are too many policemen, there is no liberty. Where there are too many soldiers, there is no peace. Where there are too many lawyers, there is no justice.
Lin Yutang (1895-1976) |
|
Back to top
|
|
|
Inventor
Stream Operator
Joined: Oct 13, 2007 Posts: 6221 Location: near Austin, Tx, USA
Audio files: 267
|
Posted: Mon Dec 29, 2008 5:55 pm Post subject:
|
|
|
Thanks guys, that's good to know. So I should have a choice of coated strings at Guitar Center. I'll continue working with what I have for now due to cost, though. And thanks for chiming in on my little thread I got going here, it's keeping my interest for the time being.
Today I built the simplest of circuits, an inverting amp with a gain of 10. It boosted the signal enough that it played on the clean channel and the noise was lowered significantly. I didn't get to play with it much, though, before I did something stoopid and I think I blew up the chip. I was going fast and sloppy, long leads everywhere and not thinking. Gotta slow down and take my time and do it right next time. But for a few minutes it worked so I'm happy with the results. I figure a gain of 100 should get the job done. More on that later, now it's time for a nap. sleepy time... _________________ "Let's make noise for peace." - Kijjaz |
|
Back to top
|
|
|
urbanscallywag
Joined: Nov 30, 2007 Posts: 317 Location: sometimes
|
Posted: Tue Dec 30, 2008 8:32 pm Post subject:
|
|
|
Hi Inventor,
I think I have more questions than answers, but your project looks interesting. I am more of a DSP guy.
From what I can find, guitar pickups act as LPF not BPF as mentioned here. It makes sense, bass guitars go pretty low. A guitar pickup maker told me the frequency response of an electric guitar is basically dictated by the cabinet.
My first suggestion was going to be to use a differential amplifier to cancel the 60 Hz hum, but it was in the patent.
I still can't figure out exactly what's going on here. What is the transducer? The guitar string actually has voltage on it? It must be induced by the magnetic pickup, would this work without one? |
|
Back to top
|
|
|
Inventor
Stream Operator
Joined: Oct 13, 2007 Posts: 6221 Location: near Austin, Tx, USA
Audio files: 267
|
Posted: Tue Dec 30, 2008 9:11 pm Post subject:
|
|
|
urbanscallywag wrote: | Hi Inventor,
I think I have more questions than answers, but your project looks interesting. I am more of a DSP guy.
From what I can find, guitar pickups act as LPF not BPF as mentioned here. It makes sense, bass guitars go pretty low. A guitar pickup maker told me the frequency response of an electric guitar is basically dictated by the cabinet.
My first suggestion was going to be to use a differential amplifier to cancel the 60 Hz hum, but it was in the patent.
I still can't figure out exactly what's going on here. What is the transducer? The guitar string actually has voltage on it? It must be induced by the magnetic pickup, would this work without one? |
Hi urbanscallywag and thanks for reviewing the project. My understanding of it is that the frequency response of an electric guitar is due to the resistance, inductance, and capacitance of the windings in the pickup. I don't know what this guitar pickup maker is thinking, it is totally different from my understanding! Strange.
It is true that the body of the guitar is also a resonant structure and it resonates with the string vibration. Also we have the sound pressure of the amplifier that contacts the strings and can create sustain or feedback, but those factors are essentially secondary influences compared to the bandpass filtering effect of the pickups.
Guitar pickups have thousands of winds and a large area so their inductance is high, the wires have resistance since they are so small so there is resistance, and the insulation is very thin so there is capacitance. These parasitics form a distributed resonant circuit.
You are correct that the magnets are inducing the currents, yes. The concept would not work without the magnets. It works the same as a generator in your car. There is motion between the wires and the magnetic field which creates a current. This current then flows through the source and load impedances, creating an output voltage.
I hope that all makes sense. Cheers! _________________ "Let's make noise for peace." - Kijjaz |
|
Back to top
|
|
|
urbanscallywag
Joined: Nov 30, 2007 Posts: 317 Location: sometimes
|
Posted: Tue Dec 30, 2008 9:31 pm Post subject:
|
|
|
Just because there's an RLC doesn't mean its bandpass, though it could have high Q. The capacitor is modeled as a shunt to ground, meaning it rolls off high frequency, not blocks low frequency.
http://buildyourguitar.com/resources/lemme/
Sorry by cabinet I meant speaker cabinet. He told me that out of everything from string to pickup to amp to speaker, the speaker is what defines the frequency response of what a guitarist hears the most. I trust his knowledge, he worked at a well known guitar amp company.
What does using the string rather than the coil gain (though I do admit there's a coolness factor)? |
|
Back to top
|
|
|
Inventor
Stream Operator
Joined: Oct 13, 2007 Posts: 6221 Location: near Austin, Tx, USA
Audio files: 267
|
Posted: Tue Dec 30, 2008 10:14 pm Post subject:
|
|
|
urbanscallywag wrote: | Just because there's an RLC doesn't mean its bandpass, though it could have high Q. The capacitor is modeled as a shunt to ground, meaning it rolls off high frequency, not blocks low frequency.
http://buildyourguitar.com/resources/lemme/
Sorry by cabinet I meant speaker cabinet. He told me that out of everything from string to pickup to amp to speaker, the speaker is what defines the frequency response of what a guitarist hears the most. I trust his knowledge, he worked at a well known guitar amp company.
What does using the string rather than the coil gain (though I do admit there's a coolness factor)? |
Oh, how revealing! I was indeed thinking that the RLC divider was a bandpass, but now I see it is a second order low pass filter! I was viewing the peak at the resonant frequency as a band pass filter, but that is wrong. It is a second order low-pass filter as you say, thanks for the explanation.
That is a really cool article you found, I really enjoyed it, well read half of it so far. I had no idea the cable capacitance was more important to the frequency response of the guitar than the winding capacitance. Great to know that, and more stuff in the article too.
Yeah, I've heard that the speaker cabinet defines the sound the most, and this seems to be the case. I thought that by cabinet you meant the body of the guitar.
Actually I don't quite know yet what using string rather than coil gains, haha! It could be that it's cheaper because you don't have to wind coils. It might have a flat frequency response so that we can have more control over the frequency response that we do want by adding different kinds of filters. Or maybe it's just different and not better or worse, I don't know.
Thanks for your post, you clarified a critical misconception on my part and provided me with a nice guitar theory resource. I welcome any additional thoughts you may have, urbanscallywag! _________________ "Let's make noise for peace." - Kijjaz |
|
Back to top
|
|
|
DrJustice
Joined: Sep 13, 2004 Posts: 2114 Location: Morokulien
Audio files: 4
|
Posted: Wed Dec 31, 2008 5:12 am Post subject:
|
|
|
urbanscallywag wrote: | From what I can find, guitar pickups act as LPF not BPF as mentioned here. It makes sense, bass guitars go pretty low. A guitar pickup maker told me the frequency response of an electric guitar is basically dictated by the cabinet. |
The coil itself is an LPF, it passes DC after all. But the pickup function seems to exhibit a band pass response. The output of the guitar is not AC coupled, and if I move a string very slowly by hand, even with a huge deflection, I get no detectable signal.
DJ
-- |
|
Back to top
|
|
|
|