Need advice/help for my school project!

[dhkstyle]

Hello, everyone! I am a high school student working on a school requirement, which is writing a physics lab on a topic of my interest. Since I enjoy playing the guitar, I decided to look more into guitar pickups. The research question that I decided on is "How does the frequency response of a guitar pickup change with the number of turns of wire?"
To carry out this experiment, I used a circuit model from the blog, Electric Guitar Pickup Measurements | kenwillmott.com

This is the circuit diagram that I tried to imitate:


As a student who is broke and has only few equipment, I used alligator clips, 3.5mm jacks, USB sound card with input and output, 100ohm, 47k ohm, and 470k ohm resistors, a small magnet that I borrowed from school, enameled copper wire, and cardboard.

1) Pickup: I watched this video (DIY Guitar Pickup | HACK A WEEK) to make this simple pickup. I basically attached the magnet shown to two pieces of cardboard, and then wrapped it with copper wire about 200 turns.


2) This is a picture of the test coil. I attached a plastic tube to two pieces of cardboard. Then I wrapped it with the same copper wire about 100 turns.



3) The whole setup of my experiment is unlike the circuit diagram that I posted. I didn't connect any resistors because the software, Right Mark Audio Analyzer, had errors


Using this setup, I ran the software RMAA and got some results:



My question is whether this result is valid. I still have a very basic understanding of the electronic concepts that are happening. From my understanding, the peak is the resonant frequency. What I don't understand is why the graph has many bumps (spikes?) unlike the results shown in the blog.

In addition, I ran another software, audiotester v3.0, with the same setup but got a significantly different graph.

Why are the two graphs significantly different? I have a feeling that the results are very inaccurate, but I just can't verify it myself.

To sum up, I have very basic understanding of this topic. I am clueless to what I should do next and whether I'm on the right track. Please, if there is anyone who can guide me through this project, I would really appreciate the generosity.

Thank you, everyone, in advance...

[dhkstyle]

Maybe I was asking too much... Can anyone give me some advice on measuring the resonant frequency without using an oscilloscope and function generator? Unfortuantely, my school can't afford these equipment, so I need a cheap way to measure frequency response (resonant frequency)...

[dstrat]

this is complicated..

you would need to be able to measure the L (inductance) C (Capacitance) R (Resistance) then can be found with math.

lot of guys around here can explain this stuff,. I'm not one of them;

best read this

https://en.wikipedia.org/wiki/LC_circuit

https://en.wikipedia.org/wiki/Electrical_impedance

[epizootics]

Hi Dhk, kudos for tackling something like this as a high school project with minimal budget and gear! You didn't pick the easiest subject here, and I'd reckon your teachers will appreciate that no matter what results you get.

Because of the design of your pickup, the output is going to be very low. From the pictures, I'd say the magnet wire you used is quite a bit thicker than what is usually used for guitar pickups. This should result in a lower resistance (with the same number of turns, the thinner the wire, the higher the resistance will be). Also participating in the very low output is how few turns of wire you have here. A typical Strat pickup is wound with about 7500 turns of AWG 42 wire. A pickup like that will have a DC resistance of about 6000 kOhms. If you have access to a digital ohm meter at school, I'd suggest you start by measuring the DC resistance of your pickup. The resistance of your coil is correlated to its output (even though it's not a perfect prediction, as it also depends on the type of magnet you used, the shape of the coil, etc.) - the lower the resistance, the lower the output. The problem a very low output poses when it comes to measuring frequency response is that the level of the signal you will get from your pickup (measured in dB) will be about as loud as the parasite noise any coil picks up. That includes humming from electrical devices around you, including your computer. This will make the results hard to analyse, as you might have a lot more going on around the same level as your signal.

This said, on the first graph, you DO get a response at 1kH. The many smaller spikes after the first one look like octave harmonics of the initial tone. How "clean" is the signal you send through your test coil? Ie. is it pure 1kH or does it inclue upper harmonics?

I am no expert in all of those fields, but I am sure there are a lot of guys here who can point you in the right direction. Don't give up, this is a cool project!

[nickb]

The signal level when using RMAA is too high. The red warning indicating clipping are showing. Also, you are exciting it with 1Khz and yet you get a series of odd harmonics. That is a sure sign of clipping.

The resistors serve the purpose to isolate the test coil from the audio card. They, or some other means of isolation is required, for example a simple op amp based buffer will do it with no loss (attenuation) of the signal. The problem with the resistors is they attenuate the signal to 47/(470+47)= 0.09. This big loss might be why RMAA was not happy. You might try the microphone input in that case or adjust the audio settings to boost the signal again.

The two things you need to accomplish is a clean drive to the exciter with no clipping and similarly a big enough signal into the audio card, but not so big that you overload it.

Good luck

[dhkstyle]

Thank you for your generous advice! One question is where or how can I get a simple op amp based buffer? Is it possible to make one with a low budget? Also, would the buffer be enough to get rid of clipping?

Thank you~

[nickb]

You can build a buffer with some gain for less than $2. It will not help clipping. You may not need it anyway. Let's keep this simple, at least for now.

First try connecting the test circuit output to the line input of the audio card. Using you audio card's management tools ot the windows one, make sure the line input is enable and the gain is all the way up. Next use RMAA and attempt do a test. You'll get the test screen up where it sends a test signal to your circuit and listens to the output. Look at the VU meters and adjust the audio card output level to try to get the VU meters in to the -1dB to -3dB range. If you can't get the VU meters high enough try connecting to the microphone input instead of the line input. Play with the input and output levels to get the VU meters in good range.

[Gnobuddy]

One question is where or how can I get a simple op amp based buffer? Is it possible to make one with a low budget?

Is your school providing you with any resources at all? Are there any school labs, that might have some equipment you could use?

A buffer (in this context) is an amplifier circuit, which you need to place between your guitar pickup, and the input to your sound card. The reason is that the characteristics of the pickup coil might otherwise be affected by the low input impedance of your sound card, so you will end up measuring the wrong thing. As was previously mentioned here, using the 470k and 47k resistors provides the pickup coil with the high impedance it wants, but has the side-effect of also cutting down the (already weak) signal.

So, my first question is, is there any sort of amplifier at your school - or home - that you might re-purpose for this experiment? You mentioned that you play guitar - what sort of equipment do you play it through? You might have a guitar effects pedal that might work as the buffer you need.

Another thought: your school probably has some sound-reinforcement (loosely, P.A.) audio gear, used for assemblies, or theatre, or other school functions. Is it possible for you to get temporary access to any of that? If your school has an audio mixer with at lest one high impedance input, you may be in business; you could put your audio software on a laptop, and take that with you to where the mixer lives.

It certainly is possible to make your own buffer, but that becomes another project of its own. It also requires some fairly expensive equipment - like a soldering iron, solder, wire cutters, et cetera - that you may not have. That's why I suggest looking for some existing equipment that can serve as your buffer.

Also, would the buffer be enough to get rid of clipping?

You have to find out where the clipping is occurring now, before you can answer that question. My bet is that the clipping is at the other end of the chain - not at the pickup coil, but rather, at the part of the sound-card that's driving your test coil. Did you put that 100 ohm resistor (R1) in series with your exciter coil? That resistor needs to be there. Without it, your exciter coil will behave like a short-circuit, and cause overloading of the headphone output of your sound card / audio interface.

One other question: do you, by any chance, have one guitar with single-coil pickups, and one guitar with humbuckers? Or a guitar with tapped humbuckers? If so, you already have two pickups with different numbers of turns in the coils, and you could conduct your experiment using your guitar, rather than your home-brewed pretend pickup coil.

(The pickup coil you wound has far fewer turns of wire than the real thing, and that will cause the signal from your fake pickup to be very, very small.)

-Gnobuddy

Edit: Nick posted while I was typing up this post.

[dhkstyle]

Thank you so much for your answer!!
So I found out that my school has a really old p.a. system and an audio mixer (would need to verify this though).


Would it be possible to use this equipment? On the back of the machine, there are four mic inputs. I'm not sure how I can connect this machine to my circuit though.

I also figured out that the clipping occurred because I did not put the 100-ohm resistor in series with the exciter coil. Thank you very much!
I guess now the only thing that I need to resolve is the weak signal.

I also have a guitar with two humbucker pickups, but I'm not confident enough to remove them. I'm worried that I will damage the guitar as a whole.

[Gnobuddy]

On the back of the machine, there are four mic inputs. I'm not sure how I can connect this machine to my circuit though.

I don't recognize that device, so I can't answer your question in detail. I also don't see any of the usual mixer controls in that photo - only what appears to be a cassette deck. Can you post a photo of the back of the supposed mixer? There should be gain, volume, tone and other controls associated with each of those mic inputs. Can you find a brand and model, so that we can find out what is actually in that black box?

At any rate, you do need to verify if the device shown in the photo can actually be used as a mixer. If it can, your fake pickup coil goes to one of the mic inputs, and the line level output from the mixer goes to your sound-card input. (VERY IMPORTANT: if that mixer only has speaker-level outputs, DO NOT connect them to your sound card. Speaker-level signals are too large, and might fry your sound-card.)

If you were asking how to make the physical connection to the mixer, mic inputs (except for cheap karaoke equipment) usually use a rather large connector with three pins (holes) in a triangle, called an XLR connector. You would need to connect your fake guitar pickup coil to the appropriate XLR (male or female), as required to match up with the one on the back of your mixer.

I'm concerned that the device in your photo might actually be cheap karaoke equipment, and not useful for your experiment. But I can't tell unless you can provide some of the other information I asked for (brand, model, more photos including the rear panel).

I guess now the only thing that I need to resolve is the weak signal.

As you probably know, since it is the topic of your experiment, the signal voltage from the pickup coil is proportional to the number of turns of wire in the coil. Real guitar pickup coils have many thousands of turns of wire in them - about eight thousand turns in a stock Stratocaster single coil pickup. Your fake pickup coil has only two hundred turns. So, all else being the same, your fake pickup will produce a signal that is forty times weaker.

As you can see, there is no mystery as to why your coil is producing such a small voltage. I hope you did some research into real guitar pickups before winding your two-hundred-turn coil.

On a similar note, how are you preparing for your project? What material have you read so far? You will need to understand how all the components in your experimental setup work, and what they do. At a bare minimum, that should include the sound card, the exciter coil, the guitar pickup coil, and what is going on in the analysis software to produce those graphs you posted.

I also have a guitar with two humbucker pickups, but I'm not confident enough to remove them. I'm worried that I will damage the guitar as a whole.

You don't need to remove them! The beauty of the exciter-coil measurement method is that you do not need direct access to the pickup wires. Just place the guitar face-up on a suitable soft surface, plug a guitar cable into the guitar, place your exciter coil on top of one of the guitar pickups, use the guitar pickup selector switch to select that pickup, set guitar tone and volume to maximum, and measure the guitar pickup signal coming out of the other end of the guitar cable (i.e., that end gets connected to your sound card input, via 470k/47k resistors).

You must use those resistors (470k/47k), or a high input impedance buffer; do not attempt to connect the guitar cable directly to the input of your sound card. The low input impedance of the sound card will drastically alter the behaviour of your humbucker pickups, and what you measure will not tell you how the pickups actually behave when used properly.

-Gnobuddy

[dhkstyle]

Thank you again for your detailed answer!!

I figured out the p.a. system I posted earlier is too old for usage. I had found an audio mixer in the music room that I managed to make it work.


https://www.amazon.com/Peavey-XR8600.../dp/B003O8P7CG

Using this mixer, I did some test with my fake pickup (now it has 890 turns, still significantly lower than the turns of an actual pickup). A problem that I encountered is that the peak of the frequency response is larger than 22,000 Hz (it is beyond the graph). Then, I did the same experiment using an actual guitar pickup, but got the same result in which the peak of the frequency response is not shown on the graph. Could it be that the test coil is hampering with the results? I read that the test coil must have a resonant frequency way beyond that of the pickup. The test coil i'm using now has 65 turns. I'm planning to wind more turns later.

I have read some resources to prepare for this project. Primarily, I browsed through this forum and the old archive to get information on measurement techniques. For instance, http://www.syscompdesign.com/assets/...ar-pickups.pdf & BuildYourGuitar.com :: The Secrets of Electric Guitar Pickups are two main resources that kind of got me started with this experiment. I looked at Measuring the Electrical Properties of Guitar Pickups | GuitarNutz 2 , which also helped me understand. In addition, I have done some research on the technical background by studying RLC circuits. However, I am still very lacking in this area as my advisor (physics teacher) is also clueless.

To briefly summarize what I understand of this experimental setup is that the sound card sends out a sine wave of frequencies ranging from 20Hz to 22,000 Hz to the exciter coil. The exciter coil causes a disturbance in the magnetic field of the guitar pickup, thus inducing a voltage in the pickup according to Faraday's law of induction. The voltage from the pickup then goes into the mic input of my soundcard, and Rightmark software analyzes it. What I don't understand is how Rightmark performs its analysis. I also don't know why other softwares produce different results. I have tried using Audiotester, ARTA, and HOLM Impulse, but they all produce varying results.

I apologize if I'm asking too much, but could you recommend me any resources that could help me with the background knowledge? As my physics teacher can't help me, I am solely relying on the internet for help.

Another question is how can I make a better pickup? I'm currently living in Argentina, and I find it very difficult to get materials such as a bobbin, and alnico magnets. I am using materials available at school, and I'm wondering if I can still improve the design of my fake pickup.

Thank you so much for your generous advice!

[Gnobuddy]

I had found an audio mixer in the music room that I managed to make it work.

Progress, excellent!

Using this mixer, I did some test with my fake pickup (now it has 890 turns, still significantly lower than the turns of an actual pickup). A problem that I encountered is that the peak of the frequency response is larger than 22,000 Hz (it is beyond the graph). Then, I did the same experiment using an actual guitar pickup, but got the same result in which the peak of the frequency response is not shown on the graph.

Your actual guitar pickup certainly has a resonance much lower than 22 kHz. I have measured a few pickups, and they usually have a peak somewhere below 5 kHz. If you're not seeing that peak, there is a problem with the measurement. The reference you originally cited (Ken Willmott) - did you look at his figure 5, and read the section titled "Equalized Response Curve Graphs"? You need to be using an integrator to see the actual frequency response of your pickup coil.

Based on what I've heard so far, my guess is that building that integrator circuit is also going to be a major problem for you. But without it, you can't really measure the frequency response of your pickups. So we may have run into a dead-end here.

There is a possible way out, though - the integration can also be done in software. The question is, does any of the software you have tried include this capability? I haven't used any of them, and can't give you an answer. But you can research it and find out.

Speaking of software, you posted three very different pictures in your first post. Two were frequency response plots, while one was a distortion spectrum. These are entirely different measurements - that's why I said in my earlier post that you need to understand what the software is actually doing, before you can understand the graphs it generates.

What I don't understand is how Rightmark performs its analysis. I also don't know why other softwares produce different results. I have tried using Audiotester, ARTA, and HOLM Impulse, but they all produce varying results.

If your signal was too small for the sound-card, you would get nonsense results from any software. But after you began to use the Peavey mixer to amplify your fake pickup signal - are you still getting different results from each different piece of software?

If so, you have to find a solution. I suggest you pick one piece of software (use the one Ken Willmott did), research it, understand exactly how to use it, and stick with it. (You also need to find out if it has the integration capability mentioned earlier.)

I apologize if I'm asking too much, but could you recommend me any resources that could help me with the background knowledge? As my physics teacher can't help me, I am solely relying on the internet for help.

If you tell me which area you want to learn about, we can try to find relevant resources. There is a huge amount of background material that would be useful, and I don't know where you need to start. (For example: when you omitted that 100 ohm resistor in series with the excitation coil, what was your reasoning for omitting it? Related question, what is a short circuit? Related question, what happens if you short-circuit the output of an amplifier? (headphone amplifier inside your sound card, in this case.)

Another question is how can I make a better pickup? I'm currently living in Argentina, and I find it very difficult to get materials such as a bobbin, and alnico magnets. I am using materials available at school, and I'm wondering if I can still improve the design of my fake pickup.

I have also made pickups when I had zero budget and nearly zero resources. I used a piece of ceramic magnet from a broken pocket-radio loudspeaker. I got wire from a junked relay coil I found. I made one bobbin by gluing together pieces of plastic from a small transparent plastic box (probably acrylic). I made a second bobbin using scraps of very thin plywood for the flanges, and a solid piece of wood for the central former; I drilled holes through the wood and inserted screws to act as pole-pieces, with the ceramic magnet on the bottom.

The pickups I made didn't look like a Fender single coil pickup, but they worked. (By the way, Fender actually seems to have stolen his pickup design from DeArmond https://www.premierguitar.com/articl...ar_Microphones )

I did this long before the Internet existed, so I didn't know the right number of turns to wind, and only wound 4000 turns. The pickups were low-output, and the guitar sounded more like an acoustic guitar than an electric guitar. The pickups worked, but not the way I had wanted them to work.

You can also improvise your own pickup design, but you need to understand how the pickup works, first. So: do you know the meaning of "magnetic circuit"? How about "magnetic flux"? How about "reluctance"?

Guitar pickups work because the vibrating guitar string alters the reluctance of a magnetic circuit that also includes a permanent magnet, and passes through a coil of wire. Those are the requirements, but other details are up to you imagination.

There is one more question I would like to ask you. If your plans do not work out, and you are unable to solve all the problems and measure the pickup frequency responses you intended to measure, what is your backup plan? (It's always good to have a back-up plan...)

-Gnobuddy

[dhkstyle]

I'm embarrassed to say that I had to look up most of the words you told me

So, a short circuit is a circuit that allows current to flow with no resistance. I have no clue what it would do to the output of an amplifier, but I guess it will do harm to it? My reasoning behind omitting the 100 ohm resistor was to simply increase the gain so Rightmark could perform its analysis.

I kind of know how the pickup works: the vibration of the strings causes a disturbance in the magnetic field, thus changing magnetic flux and inducing a voltage in the coil. However, magnetic circuits are still very new to me.

For now, I don't have a backup plan. I think it's too late for me to change a research question. My teacher keeps telling me to just stick with bad data and write why my experiment failed and what I can do to improve it.

Thank you as always for helping me! I would appreciate a lot if you could recommend any resources that could get me more informed on this topic.

[eschertron]

I have had more than my share of 'failure analysis' papers to write, no shame it it

The silver lining is that if the failures pushed you to understand the scientific principles behind the devices and the experiment, your research paper can provide the justification for a proposal for a new, more likely to succeed, experiment. This is what the experimental method is all about. Just ask Tom Edison!

[Gnobuddy]

I'm embarrassed to say that I had to look up most of the words you told me

Please don't be, learning always starts from ignorance. Much better to look answers up now, than to never look them up at all, yes?

I asked you those particular questions, because I guessed that you did not know the answers, and so you would teach yourself a lot in the process of trying to find them. Trying to find your own answers to questions is a very effective way of learning; things will sink in and make connections in your brain, in a way that just doesn't happen when someone simply gives you the answer.

Another thing to think about: a big part of any research project is to first find out what you don't know. Once you know that, you can go looking for information to fill in those gaps. If you're lucky, you will find somebody else has already figured it out for you. If not, you then try to figure it out for yourself.

I like eschertrons comments too - we're kinda both saying some of the same things. The important thing is to learn as much as you can from the project, and the way to do that is to ask questions, find answers, perform experiments, and do it all over and over again, until it finally works the way you wanted. Often it takes many tries to get there, and sometimes you never get all the way there, but you can still learn a lot in the process.

I have no clue what it would do to the output of an amplifier, but I guess it will do harm to it?

Good guess! Ohm's law tells you that a short circuit will draw an infinitely large current if you apply a voltage to it. So the amp will try to supply too much current, and that can burn out the amp.

These days, amps often have protection circuits built into them. These circuits sense when there is too much output current, and try to limit the current, so that the amp doesn't burn out.

That is why you were experiencing clipping problems earlier. Because you had a short-circuit on the output of your headphone amp, and it was trying to stop itself from delivering damaging amounts of current into the short circuit.

My reasoning behind omitting the 100 ohm resistor was to simply increase the gain so Rightmark could perform its analysis.

Makes perfect sense, if you don't know what a short circuit does to an amp! But now you do, so that is a mistake you will never intentionally make again!

However, magnetic circuits are still very new to me.

I think magnetic circuits are an old (and now unfashionable) way to look at some electromagnetic phenomena. But I think it is actually a very good way to understand guitar pickups, without getting lost in Maxwell's equations and vector fields and other fairly complex math.

For example, you mentioned that vibrating strings will cause changes in the magnetic field. But which way should the field point? In which direction should the strings run? Where should the coil be positioned? Should the strings vibrate up and down, or side to side, or does it matter?

If you go back about ninety or a hundred years, to around 1915 - 1930 or so, there were people tinkering with guitars, trying to invent some way to pick up the guitar string vibration and amplify it electrically. Many of those early attempts failed, because the would-be inventors didn't understand or think about the answers to some of the questions I just posed. Many of those tinkerers probably had very little education too, so they didn't have much background to go on. And there was no Internet, of course, so they couldn't just look up answers with a few mouse-clicks.

Ninety years later, you have the benefit of all sorts of knowledge that those people didn't have access to. You are so lucky!

For now, I don't have a backup plan. I think it's too late for me to change a research question. My teacher keeps telling me to just stick with bad data and write why my experiment failed and what I can do to improve it.

Can you maybe modify the research question a little bit? For example, can you research some early attempts at making guitar pickups, and write about them? Even better, explain why they didn't work? Maybe figure out a way to fix the problems, so they would actually work?

I would appreciate a lot if you could recommend any resources that could get me more informed on this topic.

You can help me by narrowing down the words "this topic" a little, and it will be easier to find good resources.

But I suggest looking for very old books on electricity and magnetism. Find out how to make an electromagnet, how a buzzer or electric bell works. What is a pole-piece, and why do we even need such a word? How do electric motors and generators work? What are eddy currents? Why are the iron cores inside motors and generators made out of thin laminations, and not one big chunk of iron?

Believe it or not, every one of these things will teach you something about how electromagnetism works, and how guitar pickups work. (Pickups have a lot in common with both electric bells and generators.)

-Gnobuddy