|
|
Post by antigua on Jan 25, 2019 5:29:11 GMT -5
Here are a couple data files produced by the VElleman PCSGU200, the files are named for the pickups in question. These plots were made by driving the pickups with a voltage and measuring the difference across a 1 meg resistor, as opposed to using a driver coil. echoesofmars.com/misc/velleman_pcsu200_dimarzio_super_dist.txtechoesofmars.com/misc/velleman_pcsu200_fender_fidelitron_neck.txtEach file has two plot lines, the first is unloaded, the second is loaded with 470pF and 200k. I haven't had a chance to start on a javascript based data intake and plotter, but one thing I'm curious to see is whether or not the Fideli'tron shows any pre-resonances dips when integrated mathematically. I can also create plots with an exciter coil so that the results can be compared, but that will have to wait until later in the week because among other things that requires changing around my test rig. I have been fighting with these files for a while. Some rewrites of the way i handled the data internally were required to be able to load from a generic source. My initial data model made too many assumptions based on my own process. I digress, I'm having two problems with the data: The first is I don't think it's high enough resolution (enough data points across the frequency domain) for my statistical smoothing to work. No big deal, it's sparse enough to get peak readings without denoising. I'll see if I can determine the lower limit of datapoints for loess() to work without falling back on simpleLoess() and throwing up errors. The second is the data isn't behaving as expected. I'm expecting the response between 100 and 500 Hz to be linear (and thus flat when integrated) It's not. Is this a quirk of measuring with a resistor? It's kind of inconvenient as I'm using the height of the flat section to do some calculations.
 You're right. I had forgotten that these plots were made with a 1 meg resistor in parallel to the pickup instead of a driver coil. I have since figured out that the resistor method can't be used for this purpose because it does exactly what you see there. A lower value resistor flattens the slope, but kills the resonant Q, as you'd expect it would. A higher value resistor will lower the input until it hits the noise floor, and also seems to diminish the Q. It might be possible with a high value resistor and a much higher input voltage, but the need for higher input voltage complicates the setup beyond simply using a driver coil, not to mention it might risk pushing more amperage through the pickup than the coil wire can handle. The driver coil method seems to be the only easy way to go about plotting a realistic response curve, even though creating the driver coil itself can be a bit of a project. |
|
|
|
Post by straylight on Jan 31, 2019 21:18:08 GMT -5
Having a pickup winder really helps here. I will post 100 turn driver coils with a 100 ohm resistor on a humbucker bobbin, paraffin potted for GB £5 + postage cost to anybody who is interested. I will refund anyone who produces analysis and raw data I can use.
|
|
|
|
Post by antigua on Feb 1, 2019 0:42:43 GMT -5
Having a pickup winder really helps here. I will post 100 turn driver coils with a 100 ohm resistor on a humbucker bobbin, paraffin potted for GB £5 + postage cost to anybody who is interested. I will refund anyone who produces analysis and raw data I can use. I'll re-measure those pickups with a driver coil. The only real difference in the outcome, with a driver coil, is a much flatter inductive slope, and a slightly higher Q. I have a pickup winder but I had made by driver coils with a cordless drill before that. |
|
|
|
Post by antigua on Feb 2, 2019 2:57:07 GMT -5
|
|
|
|
Post by straylight on Feb 6, 2019 16:41:38 GMT -5
Well this is looking better. The velleman file layout is a bit weird, I ended up deleting the comment section at then end rahter than trying to not read it. R isn't really a language to do text parsing in, and I'm seriously considering writing a format converter in python. Whatever the difference between this file and the last one, my statistics work, there are some issues in getting a nice direct comparison, I CAN use anything that looks like this.
In your files I'm getting a data point every 100Hz, that's ok across 10kHz, a bit vague in the 1kHz range and not really helpful for looking at low end response. I think the 100Hz sample is picking up some mains noise? I'm generating ~300 or ~600 datapoints with my syscomp, but they're logarithmicly spaced, each frequency point being a multiplier of 1.01 or 1.02 is there a way to get data points distributed like that with your velleman software? I have a demo-mode PC Lab 2000 which pight be what your running and it looks youle you can squish the frequency step down to 1/48 octave.
 |
|
|
|
Post by straylight on Feb 6, 2019 17:27:11 GMT -5
Ok, I've got the Super Distortion into my DiMarzio Comparison. Data is compariable across the cutoff for shape. I think i'm going to try a smaller driver coil to see if I can reproduce the shape you're getting. Would be nice if my beautiful lines matched someone else's  |
|
|
|
Post by antigua on Feb 6, 2019 18:48:10 GMT -5
I'm not sure what causes that impedance rise below 200Hz, my understanding of circuit analysis is fuzzy, but I think it owes to the ~13k ohms of series resistance in the pickup itself. It was a lot worse with the earlier impedance plot because the overall series resistance using that method was well over 1 meg ohm.
|
|
|
|
Post by straylight on Feb 7, 2019 7:36:52 GMT -5
I would guess that it's something we're doing differently as I have never seen this. Higher resolution across 100-1000Hz will confirm wheter it's just noise. It's only the 1 data point that is way out. Right now, the interesting parts of that plot are below 4kHz and you have 40 points below that and 260 above that.
|
|
|
|
Post by stratotarts on Feb 7, 2019 8:00:32 GMT -5
I'm not saying that it is definitely the cause, but I have noticed that 50/60 Hz noise can sometimes produce uneven response below 100Hz. I've also seen irregular responses above 100Hz when the test loop headroom is exceeded and distortion starts to creep in. It's essential to have adequate S/N across the entire test frequency range.
|
|
|
|
Post by redefinerefine on Mar 15, 2019 5:57:57 GMT -5
Setting up the hardwareStep 1) Create a driver coil using the Strat pickup bobbin and the magnet wire. Wrap the magnet wire around the bobbin about two hundred times, or enough to fill the bobbin. This will give the coil an inductance of a several millihenries, and a resonant peak that is beyond the audible range, which makes it suitable for this purpose, since its own resonance will be very far away from the pickup's resonance. Either solder lead wires to the coil as seen in the picture, or use the ends of the magnet wire as the leads. You will have to sand off the insulation coat from the wire in order to get to bare copper for soldering, or for making electrical contact.  Firstly, thank you so much for creating such a fantastic resource! I would like to ask for a quick clarification. How are the pieces of flatwork held together for the exciter coil? I normally see strat flatwork as two separate pieces held together by the magnets. Are the cheaper strat pickups a molded assembly like a humbucker bobbin? Thank you again! |
|
|
|
Post by stratotarts on Mar 15, 2019 10:30:49 GMT -5
That is a one piece molded plastic bobbin. I wind my coils on the same type, which I bought brand new in bulk online from China. I think I'm only using 50 turns of wire instead of 200, but the exact number of turns does not make much of a difference.
|
|
|
|
Post by antigua on Mar 15, 2019 11:33:24 GMT -5
That is a one piece molded plastic bobbin. I wind my coils on the same type, which I bought brand new in bulk online from China. I think I'm only using 50 turns of wire instead of 200, but the exact number of turns does not make much of a difference. I've found that I need a lot of turns of wire to get a good S/N ratio, otherwise the induced flux field is a bit weak, even if the function generator I have is maxed at 8 Vpp. I think even more turns than 200 is desirable, because the resonance of the coil is still way above the subject range, even if you get a a couple thousand turns on there, so long as the exciter has an air core. I think I've settled at around 200 only because I didn't want the coils to become any larger than they were at that point. Another way to do it would be to take a plastic Strat pickup, remove the ceramic magnet and the steel poles, and then take off about 80% of the turns on the coil, reattach the wire to the eyelet, and you should be left with a decent exciter coil. Also I went through and updated a bunch of that post, because a lot of it was out of date. That's one reason I love this forum, no time limit on post edits. |
|
|
|
Post by pablogilberto on Jun 25, 2019 6:59:27 GMT -5
Hi antiguaThanks for this article. I have learned so much. I’m starting to setup my pickup measuring device using this as a guide. I have the following questions: 1. Exciter Coil I used about 200 turns with AWG 34 on a plastic bobbin and I am getting the following values: (Test Freq is 1Khz using DE-5000) Ls = 1,774uH Lp = 11.863mH Cs = 14.27uF Cp = 2.138uF Rs = 26.56 ohms Rp = 31.25 ohms Will this work well as an exciter coil? How do I calculate if yes or no? 2. Is it correct to note the RLC values being the Rs, Ls and Cp? (Not Rp, Lp and Cs) 3. I have Ken Willmot’s work and he is suggesting to add a series resistance of 100ohms for the exciter coil. Is this necessary? 4. When measuring R, L and C values, is it best practice to use the lowest and highest possible test frequency? i.e. L - 100Hz (lowest) C - 100kHz (highest) R - 100Hz (lowest) Thank you! |
|
|
|
Post by antigua on Jul 1, 2019 22:31:03 GMT -5
Hi antigua Thanks for this article. I have learned so much. I’m starting to setup my pickup measuring device using this as a guide. I have the following questions: 1. Exciter Coil I used about 200 turns with AWG 34 on a plastic bobbin and I am getting the following values: (Test Freq is 1Khz using DE-5000) Ls = 1,774uH Lp = 11.863mH Cs = 14.27uF Cp = 2.138uF Rs = 26.56 ohms Rp = 31.25 ohms Will this work well as an exciter coil? How do I calculate if yes or no? 2. Is it correct to note the RLC values being the Rs, Ls and Cp? (Not Rp, Lp and Cs) 3. I have Ken Willmot’s work and he is suggesting to add a series resistance of 100ohms for the exciter coil. Is this necessary? 4. When measuring R, L and C values, is it best practice to use the lowest and highest possible test frequency? i.e. L - 100Hz (lowest) C - 100kHz (highest) R - 100Hz (lowest) Thank you! Sorry for the late response, I was out of town. 1) those are good values for the exciter coil. The capacitance seems high, but that doesn't matter, given how low the inductance is. More turns of wire produces a stronger magnetic field, so the trick is to get a lot of turns while still maintaining a resonant peak that is about ten times greater than that of the pickup so that the exciter coil will provide a flat test signal. 200 turns is well below the limit, but you can probably go as high a seven or eight hundred turns and still have a resonant peak that is well beyond the pickups being tested, and the stronger field would produce a better S/N ratio, if needed. In my case, I tried to create a coil that was as tiny as possible in order to approximate the geometry of a magnetized guitar string, so I was only able to get about 100 turns of 44 AWG, but all the testing Ive done doesn't indicate that there was any real difference between a driver coil that is very small, or as large as the pickup, but the driver coil should be no larger that the pickup being tested, or else it will become inefficient. 2) You want to measure three things: Ls at 100Hz (which means an assumed series resistance at a low test frequency), Cp at 100kHz (which means an assumed parallel resistance at a high test frequency), note the distinction between Hz and kHz. And third, note the DC resistance (which is neither Rs, nor Rp). The DE-5000 has a DC resistance measurement more also. 3) I have tried it with and without the series resistance and I tend to get the same results either way (same Q factor, same resonant peak frequency), but Ken Willmott is an electrical engineer by trade, so I'd do it if he suggests it. You might try for yourself, with and without, and report if happen to find a difference. 4) touched on this with #2, you want low freq for inductance, high for capacitance. |
|
|
|
Post by stratotarts on Jul 2, 2019 7:03:26 GMT -5
3) I have tried it with and without the series resistance and I tend to get the same results either way (same Q factor, same resonant peak frequency), but Ken Willmott is an electrical engineer by trade, so I'd do it if he suggests it. You might try for yourself, with and without, and report if happen to find a difference. With an audio card, I recommend driving the test coil with a headphone output, which usually has a fairly high output impedance (example 100-300 ohms). The software oscilloscope built in signal generator usually has an output impedance of 50 ohms. So when the coil is directly connected, there is already an effective series resistance, just not as great as the combined series resistance of that and the 100 ohm resistor. The difference will only show up at a very high frequency, where the coil's inductive reactance becomes equal to, or greater than, the source resistance. The purpose of the 100 ohm resistor is to guarantee proper operation regardless of what kind of signal voltage source is used. |
|
|
|
Post by pablogilberto on Jul 2, 2019 10:10:56 GMT -5
3) I have tried it with and without the series resistance and I tend to get the same results either way (same Q factor, same resonant peak frequency), but Ken Willmott is an electrical engineer by trade, so I'd do it if he suggests it. You might try for yourself, with and without, and report if happen to find a difference. With an audio card, I recommend driving the test coil with a headphone output, which usually has a fairly high output impedance (example 100-300 ohms). The software oscilloscope built in signal generator usually has an output impedance of 50 ohms. So when the coil is directly connected, there is already an effective series resistance, just not as great as the combined series resistance of that and the 100 ohm resistor. The difference will only show up at a very high frequency, where the coil's inductive reactance becomes equal to, or greater than, the source resistance. The purpose of the 100 ohm resistor is to guarantee proper operation regardless of what kind of signal voltage source is used. I am using a 2i2 Focusrite scarlett and the Rightmark Audio software. Trying to do the same setup as Ken Willmott suggested. Why is the headphone output impedance considered as high? (100-300 ohms) as you say? I’m a bit confused. |
|
|
|
Post by reTrEaD on Jul 2, 2019 10:52:10 GMT -5
Why is the headphone output impedance considered as high? (100-300 ohms) as you say? I’m a bit confused. Traditionally, manufacturers have used internal resistors in series feeding output jacks for headphones. In part, to prevent the possibility of damage to the circuits if the headphone cable fails and shorts. |
|
|
|
Post by pablogilberto on Oct 22, 2019 19:46:24 GMT -5
Hello everyone!
Has anyone tried measuring mid-scoop pickups?
Like John Mayer's Big Dipper, Kinman Mayer set or anything similar.
I'd like to understand/see how the frequency response curve of mid-scoop pickups compare to "normal" pickups.
I'm planning to build a mid-scoop pickup.
Can you give suggestions on what should I consider to do this?
I mean, the wire gauge, number of turns (underwound or overwound), magnet types (alnico II or V).
Or anything relevant info such as resonant frequency or RLC values of the pickup.
Thank you so much!
|
|
|
|
Post by antigua on Oct 30, 2019 23:12:22 GMT -5
Hello everyone! Has anyone tried measuring mid-scoop pickups? Like John Mayer's Big Dipper, Kinman Mayer set or anything similar. I'd like to understand/see how the frequency response curve of mid-scoop pickups compare to "normal" pickups. I'm planning to build a mid-scoop pickup. Can you give suggestions on what should I consider to do this? I mean, the wire gauge, number of turns (underwound or overwound), magnet types (alnico II or V). Or anything relevant info such as resonant frequency or RLC values of the pickup. Thank you so much! Sorry I missed this post. There's no single coil with an inherent mid scoop, that's really just the end result of an in situ pickup with a high resonant peak. The pickup with the high resonant favors the treble, the moving guitar strings themselves favor the bass/fundemental, and so you combine those two things together, you have a lot of bass, a lot of treble, and not a particularly prominent mid range. According to this listing reverb.com/item/29201496-fender-john-mayer-big-dipper-pickups?gclid=Cj0KCQjw6eTtBRDdARIsANZWjYYwqUxdhUJleh-GzofyLVAKvPvaTc5C3PJHkGH6GuocWho-JY-m-iQaAkytEALw_wcB&merchant_id=143081433&pla=1&utm_campaign=7276534768&utm_medium=pla&utm_source=google the Big Dipper DC resistances are... Neck: 5.75 Ohms Middle: 5.88 Ohms Bridge: 6.23 Ohms The neck and the middle pickups are in between the CS '69 and the '67/'52 sets. The bridge pickup is similar to that of the Fat 50's set. They're all so close that it's debatable that anyone could really tell them apart in a direct A/B with a blind fold on. Unscientifically speaking, I feel I have an easier time working with the 57/62's, but I've never taken the time to A/B them in front of a guitar amp. |
|
|
|
Post by pablogilberto on Nov 15, 2019 3:58:12 GMT -5
Hi Antigua! I'm trying to do a circuit simulation in SPICE. I'd like to ask for your advice. I am using an AC voltage source in series with R L and C to represent the pickup. I'm getting the output from Capacitor. I understand that you chose 200k and 470pF to represent the potentiometers, capacitors and cable capacitance. What I want to do is find out how changing potentiometer and capacitor values affect the frequency response. To do this, I'm planning to include the following: 1. Potentiometer and Capacitor network 2. Cable capacitance For the guitar amp, do I need to include a parallel load of 1Meg to represent it? This is my schematic for a single pickup with 1 tone 1 volume. Thanks!  ![[url=/7j8Pd4N][img]https://i.ibb.co/Bcdb5NK/guitar-wiring-schem.png[/img][/url]]() |
|
|
|
Post by JohnH on Nov 15, 2019 14:45:55 GMT -5
Hi Pablo,
That's a good basic Spice model to experiment with, and you can learn a lot from it.
As drawn, the tone circuit is after the volume, which on an LP is '50's wiring'. But it is much more common to have the tone before the volume, and as you turn down volume, it makes a big difference. Actually, you could run it both ways and you will see for yourself.
1M is a good value to represent amp input impedance. The cable capacitance, for an average 3m cable can be 0.5nF = 500pF, which includes about 100pF for the amp input capacitance. A 6m cable will be about 900pF.
These models just pick up the basic electrical characteristics, so you will see a peak which in a real guitar, is muted by the effects of the string response etc. Also, this model is about the simplest useful model of a pickup, but doesn't represent a number of complex damping effects within the pickup, hence the peaks it gives will be sharper than reality. Still very useful though for comparing wiring ideas.
|
|
|
|
Post by antigua on Nov 16, 2019 13:47:25 GMT -5
What I want to do is find out how changing potentiometer and capacitor values affect the frequency response. To do this, I'm planning to include the following: 1. Potentiometer and Capacitor network 2. Cable capacitance For the guitar amp, do I need to include a parallel load of 1Meg to represent it? The instrument input impedance varies between amplifiers and effect pedals. It can be as low as 250k ohms, or beyond 1meg ohms. It just changes the Q factor of the circuit resonance, the lower values will cause a lower Q factor. 1 meg is a good representative value. That circuit assumes no eddy current losses, but any guitar pickup with a lot of conductive metal parts will have a virtual parallel load in the form of eddy currents, which is tricky to model as a physical relationship in spice, but someone who goes by Tele Tuscon has done it:  This is like the model you drafted, except the voltage source is made to be frequency dependent, and is inductively coupled with the guitar pickup, and resistor is also inductively coupled to the pickup, representing eddy current load. If you tell me the type of guitar pickup, I can't give you exact values that would correspond, but I can describe whether the eddy current losses would be minimal or maximal based on having tested a lot of different pickups. Fender style pickups retain a high Q due to the lack of steel or brass parts, PAF's and P-90's reduce the Q factor to the point of almost being flat up to the resonant roll off, while Filter'trons cause substantial roll off ahead of the resonant peak. What sort of decision are you trying to make with respect to the pots and caps? |
|
|
|
Post by pablogilberto on Nov 17, 2019 10:07:33 GMT -5
Thank you so much for the info!
|
|
|
|
Post by pablogilberto on Nov 17, 2019 20:26:53 GMT -5
What I want to do is find out how changing potentiometer and capacitor values affect the frequency response. To do this, I'm planning to include the following: 1. Potentiometer and Capacitor network 2. Cable capacitance For the guitar amp, do I need to include a parallel load of 1Meg to represent it? The instrument input impedance varies between amplifiers and effect pedals. It can be as low as 250k ohms, or beyond 1meg ohms. It just changes the Q factor of the circuit resonance, the lower values will cause a lower Q factor. 1 meg is a good representative value. That circuit assumes no eddy current losses, but any guitar pickup with a lot of conductive metal parts will have a virtual parallel load in the form of eddy currents, which is tricky to model as a physical relationship in spice, but someone who goes by Tele Tuscon has done it: This is like the model you drafted, except the voltage source is made to be frequency dependent, and is inductively coupled with the guitar pickup, and resistor is also inductively coupled to the pickup, representing eddy current load. If you tell me the type of guitar pickup, I can't give you exact values that would correspond, but I can describe whether the eddy current losses would be minimal or maximal based on having tested a lot of different pickups. Fender style pickups retain a high Q due to the lack of steel or brass parts, PAF's and P-90's reduce the Q factor to the point of almost being flat up to the resonant roll off, while Filter'trons cause substantial roll off ahead of the resonant peak. What sort of decision are you trying to make with respect to the pots and caps? Thank you for posting this! Do you have a detail explanation regarding this schematic model? I would like to understand it better. The why and how it was modelled. Many thanks!  Regarding pots and caps, I understand that changing values will affect the resonant frequency and Q factor of the pickup curve. What I'd like to study now are the following: 1. How much change in Volume Pot resistance is needed to hear any difference in sound/tone? i.e. If I have 2 pots rated at 250k, having the actual values of is 247k and 251k, will the sound differ? (assuming volume pot is at 10). Or do I need a drastic change, say from 250k to 300k, or 250k to 500k in order to hear any difference? 2. Similar with the pots question. How much change in Tone capacitor capacitance is needed to hear any difference in sound/tone? i.e. If I have 2 caps rated at 0.047uF, having the actual values of is 0.045uF and 0.049uF, will the sound differ? (assuming tone pot is at 10). If not, will changing the value to 0.033uF or 0.022uF give audible result? Or is the effect more felt when the tone pot is being rolled down? 3. I'd like to understand how different capacitor types (mylar, ceramic, orange drop and paper-in-oil caps) affect the overall sound. If I have different types of capacitor having the same values. Will they sound any different, and why? Is the difference because of tiny difference in actual capacitance values or is it because of different dielectric materials? Thank you! |
|
|
|
Post by stratotarts on Nov 20, 2019 9:33:25 GMT -5
What I'd like to study now are the following: 1. How much change in Volume Pot resistance is needed to hear any difference in sound/tone? i.e. If I have 2 pots rated at 250k, having the actual values of is 247k and 251k, will the sound differ? (assuming volume pot is at 10). Or do I need a drastic change, say from 250k to 300k, or 250k to 500k in order to hear any difference? 2. Similar with the pots question. How much change in Tone capacitor capacitance is needed to hear any difference in sound/tone? i.e. If I have 2 caps rated at 0.047uF, having the actual values of is 0.045uF and 0.049uF, will the sound differ? (assuming tone pot is at 10). If not, will changing the value to 0.033uF or 0.022uF give audible result? Or is the effect more felt when the tone pot is being rolled down? 3. I'd like to understand how different capacitor types (mylar, ceramic, orange drop and paper-in-oil caps) affect the overall sound. If I have different types of capacitor having the same values. Will they sound any different, and why? Is the difference because of tiny difference in actual capacitance values or is it because of different dielectric materials? Thank you! What you are asking, can not be definitively answered in purely technical terms, because it is a perceptual question, "what does it sound like?". A definitive technical answer can only be obtained with listening tests, properly implemented in such a way as to eliminate cognitive biases. You can form an opinion, based on applying the results of other listening tests to the subject phenomena. Such an opinion will often be right, but not always. Such an opinion will always be provisional and has to compete with other, often lesser informed, opinions because it has no direct experimental basis. This is why you will see 1000 different online opinions on your question #3, for example.
Generally, facts are regarded highly here, and the only way to establish your questions factually and definitively, is to perform the listening tests scientifically.
Having said that, the answers are:
1. Enough difference to create at least a 1dB difference in the loaded resonant peak 2. Almost the same as #1, except that it is a frequency shift rather than amplitude. Also the difference diminishes rapidly as you turn up the tone control. 3. Different cap types have no audible influence on the sound. It's because their characteristics don't vary enough for that.
|
|
|
|
Post by antigua on Nov 23, 2019 0:58:49 GMT -5
Thank you for posting this! Do you have a detail explanation regarding this schematic model? I would like to understand it better. The why and how it was modelled. Many thanks! Regarding pots and caps, I understand that changing values will affect the resonant frequency and Q factor of the pickup curve. What I'd like to study now are the following: 1. How much change in Volume Pot resistance is needed to hear any difference in sound/tone? i.e. If I have 2 pots rated at 250k, having the actual values of is 247k and 251k, will the sound differ? (assuming volume pot is at 10). Or do I need a drastic change, say from 250k to 300k, or 250k to 500k in order to hear any difference? 2. Similar with the pots question. How much change in Tone capacitor capacitance is needed to hear any difference in sound/tone? i.e. If I have 2 caps rated at 0.047uF, having the actual values of is 0.045uF and 0.049uF, will the sound differ? (assuming tone pot is at 10). If not, will changing the value to 0.033uF or 0.022uF give audible result? Or is the effect more felt when the tone pot is being rolled down? 3. I'd like to understand how different capacitor types (mylar, ceramic, orange drop and paper-in-oil caps) affect the overall sound. If I have different types of capacitor having the same values. Will they sound any different, and why? Is the difference because of tiny difference in actual capacitance values or is it because of different dielectric materials? Thank you! Regarded strictly as a signal filter, guitar pickups are RLC networks with both series and parallel resistances, so the basic model is basically nothing more than that, and it's rather accurate. But then the steel parts in and around a pickup cause eddy currents, the relationship between the eddy currents and the guitar pickup is the same as the relationship between a primary and secondary coil in a transformer, and so it can be modeled with LTSpice as a transformer. You don't really have to model all of it, just know that pickups with steep part, like screws and slugs will have a much lower Q factor than the simple pickup model indicates. With a PAF type humbucker, the steel causes a loss of Q factor comparable to putting a ~100k resistor in parallel with the pickup, so it rounds off the high end response of the pickup quite a bit. As said above, how you will know if you can hear a difference is if the amplitude changes by more than 1dB. 1dB to 2dB +/- not very audible, but a 3dB difference should be easily heard. If the resonant peak of the pickup is high, 5kHz and beyond, it will surpass the operation range of a guitar speaker and probably won't be audible. The lower the peak frequency, the more likely you are to hear the difference in the Q factor. If the tone pot is at 10, the cap value is effectively irrelevant. As you roll it closer to zero, the value matters increasingly. The differences between 1nF and 100nF is easily audible, but outside of that range, you won't hear much difference, as below 1nF will attenuate too high of a frequency, and beyond 100nF too low of a frequency. 22nF and 47nF are the standard values, but you can use 100nF to get a supper muddy tone, or 3nF to get a thick humbucker/P-90 like tone from single coils. The capacitor dielectric doesn't matter, so long as it's not electrolytic / polarized. The difference between film and ceramic doesn't matter at low voltage audio frequencies at indoor/outdoor temperatures, but in general, film caps are more reliable; they're usually more accurate, less microphone and have a more linear capacitance-by-voltage curve. I don't think the difference is audible, but I've never really tested it specifically. I don't know of anyone ever, in the history of the electric guitar, being able to guess the cap type just by listening to the guitar. |
|
|
|
Post by pablogilberto on Nov 24, 2019 5:09:41 GMT -5
What you are asking, can not be definitively answered in purely technical terms, because it is a perceptual question, "what does it sound like?". A definitive technical answer can only be obtained with listening tests, properly implemented in such a way as to eliminate cognitive biases. You can form an opinion, based on applying the results of other listening tests to the subject phenomena. Such an opinion will often be right, but not always. Such an opinion will always be provisional and has to compete with other, often lesser informed, opinions because it has no direct experimental basis. This is why you will see 1000 different online opinions on your question #3, for example.
Generally, facts are regarded highly here, and the only way to establish your questions factually and definitively, is to perform the listening tests scientifically.
Having said that, the answers are:
1. Enough difference to create at least a 1dB difference in the loaded resonant peak 2. Almost the same as #1, except that it is a frequency shift rather than amplitude. Also the difference diminishes rapidly as you turn up the tone control. 3. Different cap types have no audible influence on the sound. It's because their characteristics don't vary enough for that.
Thank you for your inputs! |
|
|
|
Post by pablogilberto on Nov 24, 2019 5:13:17 GMT -5
Thank you for posting this! Do you have a detail explanation regarding this schematic model? I would like to understand it better. The why and how it was modelled. Many thanks! Regarding pots and caps, I understand that changing values will affect the resonant frequency and Q factor of the pickup curve. What I'd like to study now are the following: 1. How much change in Volume Pot resistance is needed to hear any difference in sound/tone? i.e. If I have 2 pots rated at 250k, having the actual values of is 247k and 251k, will the sound differ? (assuming volume pot is at 10). Or do I need a drastic change, say from 250k to 300k, or 250k to 500k in order to hear any difference? 2. Similar with the pots question. How much change in Tone capacitor capacitance is needed to hear any difference in sound/tone? i.e. If I have 2 caps rated at 0.047uF, having the actual values of is 0.045uF and 0.049uF, will the sound differ? (assuming tone pot is at 10). If not, will changing the value to 0.033uF or 0.022uF give audible result? Or is the effect more felt when the tone pot is being rolled down? 3. I'd like to understand how different capacitor types (mylar, ceramic, orange drop and paper-in-oil caps) affect the overall sound. If I have different types of capacitor having the same values. Will they sound any different, and why? Is the difference because of tiny difference in actual capacitance values or is it because of different dielectric materials? Thank you! Regarded strictly as a signal filter, guitar pickups are RLC networks with both series and parallel resistances, so the basic model is basically nothing more than that, and it's rather accurate. But then the steel parts in and around a pickup cause eddy currents, the relationship between the eddy currents and the guitar pickup is the same as the relationship between a primary and secondary coil in a transformer, and so it can be modeled with LTSpice as a transformer. You don't really have to model all of it, just know that pickups with steep part, like screws and slugs will have a much lower Q factor than the simple pickup model indicates. With a PAF type humbucker, the steel causes a loss of Q factor comparable to putting a ~100k resistor in parallel with the pickup, so it rounds off the high end response of the pickup quite a bit. As said above, how you will know if you can hear a difference is if the amplitude changes by more than 1dB. 1dB to 2dB +/- not very audible, but a 3dB difference should be easily heard. If the resonant peak of the pickup is high, 5kHz and beyond, it will surpass the operation range of a guitar speaker and probably won't be audible. The lower the peak frequency, the more likely you are to hear the difference in the Q factor. If the tone pot is at 10, the cap value is effectively irrelevant. As you roll it closer to zero, the value matters increasingly. The differences between 1nF and 100nF is easily audible, but outside of that range, you won't hear much difference, as below 1nF will attenuate too high of a frequency, and beyond 100nF too low of a frequency. 22nF and 47nF are the standard values, but you can use 100nF to get a supper muddy tone, or 3nF to get a thick humbucker/P-90 like tone from single coils. The capacitor dielectric doesn't matter, so long as it's not electrolytic / polarized. The difference between film and ceramic doesn't matter at low voltage audio frequencies at indoor/outdoor temperatures, but in general, film caps are more reliable; they're usually more accurate, less microphone and have a more linear capacitance-by-voltage curve. I don't think the difference is audible, but I've never really tested it specifically. I don't know of anyone ever, in the history of the electric guitar, being able to guess the cap type just by listening to the guitar. I am really learning a lot. Thank you so much for this. |
|
|
|
Post by aquin43 on Nov 24, 2019 7:46:46 GMT -5
What i am trying to say.. we need to get rid of this 6dB connection to faradays law. In my view it just isn't existing. I found this rising slope NOWHERE in any literature. Help me out here. Did you try to put in 6db and induction into google? first thing that should pop out is: Faradys law produces a 6dB rising slope. But it doesn't. 6dB is connected to the high pass the driver coil is producing. (so it seems, somewhere some guitarist or coil producer tried to explain this rising slope and just made a missinterpretation that many other people just took for granted). Let me try to get to my point. First, lets do some fundamental mathematics here E String 82,4Hz e String 329,6Hz FORUMS interpretation on how the law of induction works is, the output increases with 6dB/octave. So let me have a look at that. 3dB is a doubling the mean value. meaning... 6dB is 4x the power. so when i am plugging the E string you are trying to tell me that the guitar pickup will transmit the e string, which is the 4th harmonic of E, with 18dB difference, meaning like more than 10x louder than the E string, even though i plug them with the same magnitude. Show me a guitar that works like that please. Never heard this effect. If we would apply this interpretation of yours to harmonics to guitar strings as well, this would mean the harmonics, that are always swinging as well, would be MUCH louder than the fundamental. Put in the standard model of a PU into LTspice and it will show you how a PU really works. There is no 6dB slope. So... a PU works like a lowpass with a consistent transmission in the low end, until the resonant peak increase kicks in. The bandpass you measure is because of the NEW SYSTEM you created. A driver coil (high pass) and a PU (low pass) connected together (by induction) will produce a bandpass form. The rising slope of 6dB is connected to the order of the filter. Change the values of the driver coil in LTspice and you get a steeper slope. I'll be back on you later. BR A little humility might be advised when attempting to tear up known physical law. Can you really not see the glaring error in your E - e thought experiment above? Arthur |
|
|
|
Post by perfboardpatcher on Nov 24, 2019 8:15:43 GMT -5
|
|
