It would be interesting to get some good frequency response plots for guitar pickups, and to see the effect of adding a volume knob with a certain impedance, even if it is up at full.
One of my homework assignments in my "Guitar Amplification and Effects" class has the students make some plots showing how things change with different choice of the volume control (with the volume up at full), to show how the volume control loads everything down.
Great video. About the frequency response measurement, have you considered an op-amp integrator after the TL072 buffer? That should cancel the differentiating effect of the exciter/pickup coils and restore a normalized frequency function.
I have a test rig to approximate the frequency response and output impedance of the whole guitar wiring, which is essentially an HSS strat pickguard, but instead of connecting the start of the coils to ground, they are connected to an input jack. This allows me to send a sweep through the test rig before it goes into other devices (mainly pedals I am designing) in order to check the interaction between the pedal's input stage and the guitar wiring.
To negate the effects of the derivator zero, can't we just use an OTA as a Variable Gain Integrator between the oscilloscope and the excitor coil (with another buffer) and get an integrator effect to cancel the zero with unity gain?
@@LantertronicsI recently found out that the integrator idea has been done and produces great results, a youtube channel called "Guitar MD" has a comprehensive guide on how he measures pickups that's the same idea just executed with a pc instead of an oscilloscope.
Without the buffer I would suspect your output wave from would mirror the input wave form but have a different amplitude. An audio listen to source and output would confirm or discredit the visual on the scope. Exciting the pickup with an Elbow through the scope would also be revealing.
Can you select the gen as a trigger source in the scope? If not, use second channel to trigger off gen source ( you can turn the trace off and still use trigger). This will create a rock solid trigger. An FFT will help you find the interference frequencies, causing that distortion.
@@Lantertronics I spent a couple of decades showing people how to get rock solid triggers, having worked for HP Test & Measuremnt (Agilent/Keysight) and then at Tektronix. Couldn't help myself. 😆 The other thing I wanted to do, watching this was segment memory and look at each capture individually with its time stamp to understand the waveform distortion. And maybe even trigger on it. But just because you can, doesn't mean it's needed. 😎
@@LantertronicsI remember one of the Tek engineers recounting their use of an IBM fab and the IBM engineers not knowing how they achieved what was being created on their own equipment. When your challenge is to build equipment that can measure technology that hasn't been invented yet with technology you've just inventing or even existing technology, it's a fun place to be! I feel so incredibly blessed to have worked at Keysight when it was Hewlett-Packard Test & Measurement, and Tektronix, before Danaher bought them and stripped out cash, resources, and talent. I do have an affinity to those products I sold... I recently bought another Tek instrument that was new when I started there in 2000. It's also an interesting thought when you think "If someone rear ended my car in this traffic, they could destroy 1.5M in instrumentation in my trunk." I felt fortunate to have the customers I had, to be able to see what they were working on... BlackBerry, ATI/AMD, Motorola, Ciena, and innovate start up companies, university professors. OMG, for those of us curious, it was a wonderful job to understand their challenges and assist in finding solutions. So when you do things like this, I flash back to sitting with associates and customers, working through how do you measure it and what does it mean. I apologise for the long response, I really do find a lot of joy in stuff like this, so as a result, I appreciate your videos. 😊
So the exciter coil does basically what the strings and plucking a note does? Except it's constant instead of starting and stopping? So you get a response from the tested pickup with the exciter coil? I'm not sure I understand unless that's how it works lol I actually only just started getting into this. Been reading a lot of the pickup analyses done by Antigua, a guy on a bunch of the forums thats tested A LOT of pickups. You can look up "Antigua pickup analysis" and you'll see all of his stuff. He gets EXTREMELY in depth. I think he even did a test to measure how pickup height of the pickup that's OFF affects the tone of the pickup that's active. He's tested different pole pieces and base plates and all, sorts of stuff. There's one test where he tests a Donlis PAF against a couple Dimarzios and when he swapped pole pieces between the 2 pickups, it made zero difference. Meaning Donlis uses the same parts Dimarzio does, more or less.
It's something like that. The pickup is responding to a change in magnetic field. Here I'm changing the magnetic field using the excitation pickup. When you plug a guitar string, the strings become magnetized by being in the magnetic field of the pickups magnet, and moving them disturbs the field. It's similar in that the pickup responds to the VELOCITY of the string, not its absolute position. I'm not familiar with Antigua, but will check out their work! Thanks for the tip!
So then, would it be safe to say the first plot (with the excitation coil) is an analog of the pickup’s reluctance Bode plot? Would it be safe to assume that aside from some offset, the relative measurement in dB between given frequencies is directly correlated? This is something I’ve been pondering how to measure - really exciting to see the quality of your results from even a rough setup like this As for the jank - maybe you’re experiencing non-linearity due to saturation? Especially as the first test has a strong static field imposed on the DUT
I'll be honest, I don't actually remember enough EM to make a statement about reluctance, or exactly what it is that I'm plotting. I'd have to do some reviewing and some thinking...
Nice experiment! Usually the volume pots on guitars are 250K for single coils and 500K for humbuckers, sometimes 1M for some Telecasters, afaik. There's also an interesting idea of running the humbucker outputs through an opamp inverting and noninverting outputs directly, like an "activator" (making passive pickups behave more like active pickups), which, in my tests, increase the "bite", probably because of the effect you demonstrated of moving the peak and treble roll off higher.
Yeah, in lab and homework I sometimes have my students use an artificially low value for the volume control just to exaggerate the effect and make it more obvious.
@@Lantertronics Just to be clear: Not my idea. That's how (afaik) all active pickups are made. There's an interesting article on ElectroSmash website (which I think I'm unable to link, I guess) about the EMG81 (a famous active pickup) about the disassembly and analysis of the pickup itself and the circuit inside, which works as I pointed out. The most interesting part to me is the fact that you can use different resistor values on the opamp inputs (in series with each coil) to swerve the pickup response to be more single-coil-ish. You get less of the noise-cancelling benefit inherent to humbuckers as a tradeoff, I guess. You can also set the gain separately on the feedback loop to "match" volumes between bridge and neck pickups, which is nice. Anyway, thank you for sharing valuable knowledge online.
Your exciting one coil with another which is kind of like a tank circuit. I wonder what the resonant frequency of the guitar pickup coil is? Is this resonant frequency within the audio range and if not; does it have any sub harmonic frequencies within the audio spectrum we can detect by ear?
The underlying tank isn't really one coil vs. the other, it's the inductance of the pickup (which is something the designer can readily control) combined with the interwinding capacitance of the pickup (which is harder to control and just kind of winds up however it winds up). The resonance bump is going to be close to 1/(2*pi*L*C). It will be less than that because of the resistance. It's well within the audio range, and it's really what gives each pickup its characteristic voice. I have a bunch of lectures on this in my ECE4448 "Guitar Amplification and Effects" class playlist I'd invite you to check out -- you can readily find it by searching on TH-cam (I'm asking you to search instead of adding a link since TH-cam sometimes deletes links.)
Are you using obsolete TL072 cause it's the cheapest high impedance buffer and is good enough? More expensive OPA2134 and plenty of CMOS opamps have better noise specs.
Oh yeah, there's definitely better spec op amps out there. I used the TL072 because I have a drawer full of them. ;) They're generally a reasonable workhorse op amp for synthesizers and guitar pedals and such.
Humbucker-equipped guitars will typically have 500k volume and tone pots. Would that further shift the frequency plot as compared to your 100k resistor?
Using a higher resistance pot wouldn't shift it as much as the 100K pot I used. Think about it this way, using a higher resistance pot eventually is approximately like not using a volume control at all.
Helmuth Lemme has charts showing the effect on the resonant frequency of various resistance loads. Also the effect of capacitance (for example from the guitar cable) on the Q factor (height of resonant peak ~= treble emphasis). Google his "The Secrets of Electric Guitar Pickups" and also his "Pickup measuring techniques". The standard in pickup testing is not only to measure the unloaded response, but also that with 250k total resistance and 470 pF capacitance. But of course you can use any loading more typical of the conditions you want to put the pickup into. Given their typical difference in resistance loading, it would make sense to have one standard for single coil guitars and one for humbucker guitars.
500k will raise the resonant peak compared to 100k, yes. The other very important thing to remember is there will be tone controls also connected in parallel with both the volume control and the pickups, so those also will impact the resonant peak (though not that much when they are set to max level).
@@danielsanichiban yep! that's great. but i'd still love to hear what this setup sounds like, because it's unique :) would love to see how it reacts when the items are being moved around in 3d space etc.
@@cheater00The sound output can be pretty ok except lacking bass with programme like you suggested. Varying load impedance and/or tone pot/tone cap and listening and looking at the FR plot is revealing. For proper measurement/analysis the drive coil needs to be constant current source. A power amp driving the coil through a high value resistor approximates constant current source well enough.
Here I want to get the actual data to load into Excel or MATLAB and process. There's a way to do it, I just need to spend some quality time with the manual.
@@Lantertronics I've been there and seen similar things. Be careful, you might get deep into a rabbit hole. On my recent built guitar FX I noticed a permanent noise, which could come from the circuit, but it was there. Turned out, it came from the switching power supply. After using a ring transformed PSU, the noise was gone.
The guitar pickup is picking up 60hz hum and other EMI interferences that is what you're seeing I'm sure. You oscilloscope impedance is set to 50 ohms which a guitar pickup impedance output is NOT 50 ohms so that is a problem you have to change the Oscilloscope port impedance. Try to make another video about measuring the frequency response, Pickups peak frequency and Pickups impedance
Have you seen the huge database of pickups analyzed by that guy Antigua? He's on a bunch of different forums and has analyzed probably hundreds of pickups by now.
The TL072 output impedance is not going to be ideal 0 ohms but I think close enough with the oscilloscope set to 50 ohm input impedance. Would be good to measure to confirm.
It would be interesting to get some good frequency response plots for guitar pickups, and to see the effect of adding a volume knob with a certain impedance, even if it is up at full.
One of my homework assignments in my "Guitar Amplification and Effects" class has the students make some plots showing how things change with different choice of the volume control (with the volume up at full), to show how the volume control loads everything down.
Great video. About the frequency response measurement, have you considered an op-amp integrator after the TL072 buffer? That should cancel the differentiating effect of the exciter/pickup coils and restore a normalized frequency function.
I have a test rig to approximate the frequency response and output impedance of the whole guitar wiring, which is essentially an HSS strat pickguard, but instead of connecting the start of the coils to ground, they are connected to an input jack. This allows me to send a sweep through the test rig before it goes into other devices (mainly pedals I am designing) in order to check the interaction between the pedal's input stage and the guitar wiring.
To negate the effects of the derivator zero, can't we just use an OTA as a Variable Gain Integrator between the oscilloscope and the excitor coil (with another buffer) and get an integrator effect to cancel the zero with unity gain?
I have a feeling that may be overkill and there's probably a simpler solution, but I haven't really thought about it much.
@@Lantertronics We can also consider not using excitation by flux, but rather inducing vibration on an actual string to take the frequency response.
@@Eren_Sahin_ I have an eBow. It's hard to use consistently, though. I think doing this with lab experiment level consistency would be tricky.
@@LantertronicsI recently found out that the integrator idea has been done and produces great results, a youtube channel called "Guitar MD" has a comprehensive guide on how he measures pickups that's the same idea just executed with a pc instead of an oscilloscope.
@@Eren_Sahin_ Thanks, I will check it out!
Without the buffer I would suspect your output wave from would mirror the input wave form but have a different amplitude.
An audio listen to source and output would confirm or discredit the visual on the scope.
Exciting the pickup with an Elbow through the scope would also be revealing.
Oh! I do have an Ebow, that's a good idea, I should try that.
The buffer is just, well, buffering. There's no getting around the derivative -- that's coming form the physics.
Can you select the gen as a trigger source in the scope? If not, use second channel to trigger off gen source ( you can turn the trace off and still use trigger). This will create a rock solid trigger.
An FFT will help you find the interference frequencies, causing that distortion.
Ooooh! Yeah... I could have totally triggered off the input, and usually do that, I just wasn't thinking.
@@Lantertronics I spent a couple of decades showing people how to get rock solid triggers, having worked for HP Test & Measuremnt (Agilent/Keysight) and then at Tektronix. Couldn't help myself. 😆
The other thing I wanted to do, watching this was segment memory and look at each capture individually with its time stamp to understand the waveform distortion. And maybe even trigger on it. But just because you can, doesn't mean it's needed. 😎
@@GrinRepair The quality of electronics -- and the cleverness -- of high end measurement equipment blows my mind.
@@LantertronicsI remember one of the Tek engineers recounting their use of an IBM fab and the IBM engineers not knowing how they achieved what was being created on their own equipment.
When your challenge is to build equipment that can measure technology that hasn't been invented yet with technology you've just inventing or even existing technology, it's a fun place to be!
I feel so incredibly blessed to have worked at Keysight when it was Hewlett-Packard Test & Measurement, and Tektronix, before Danaher bought them and stripped out cash, resources, and talent.
I do have an affinity to those products I sold... I recently bought another Tek instrument that was new when I started there in 2000.
It's also an interesting thought when you think "If someone rear ended my car in this traffic, they could destroy 1.5M in instrumentation in my trunk."
I felt fortunate to have the customers I had, to be able to see what they were working on... BlackBerry, ATI/AMD, Motorola, Ciena, and innovate start up companies, university professors. OMG, for those of us curious, it was a wonderful job to understand their challenges and assist in finding solutions.
So when you do things like this, I flash back to sitting with associates and customers, working through how do you measure it and what does it mean.
I apologise for the long response, I really do find a lot of joy in stuff like this, so as a result, I appreciate your videos. 😊
So the exciter coil does basically what the strings and plucking a note does? Except it's constant instead of starting and stopping? So you get a response from the tested pickup with the exciter coil? I'm not sure I understand unless that's how it works lol
I actually only just started getting into this. Been reading a lot of the pickup analyses done by Antigua, a guy on a bunch of the forums thats tested A LOT of pickups.
You can look up "Antigua pickup analysis" and you'll see all of his stuff. He gets EXTREMELY in depth. I think he even did a test to measure how pickup height of the pickup that's OFF affects the tone of the pickup that's active. He's tested different pole pieces and base plates and all, sorts of stuff. There's one test where he tests a Donlis PAF against a couple Dimarzios and when he swapped pole pieces between the 2 pickups, it made zero difference. Meaning Donlis uses the same parts Dimarzio does, more or less.
It's something like that. The pickup is responding to a change in magnetic field. Here I'm changing the magnetic field using the excitation pickup. When you plug a guitar string, the strings become magnetized by being in the magnetic field of the pickups magnet, and moving them disturbs the field. It's similar in that the pickup responds to the VELOCITY of the string, not its absolute position.
I'm not familiar with Antigua, but will check out their work! Thanks for the tip!
So then, would it be safe to say the first plot (with the excitation coil) is an analog of the pickup’s reluctance Bode plot? Would it be safe to assume that aside from some offset, the relative measurement in dB between given frequencies is directly correlated?
This is something I’ve been pondering how to measure - really exciting to see the quality of your results from even a rough setup like this
As for the jank - maybe you’re experiencing non-linearity due to saturation? Especially as the first test has a strong static field imposed on the DUT
I'll be honest, I don't actually remember enough EM to make a statement about reluctance, or exactly what it is that I'm plotting. I'd have to do some reviewing and some thinking...
Nice experiment!
Usually the volume pots on guitars are 250K for single coils and 500K for humbuckers, sometimes 1M for some Telecasters, afaik.
There's also an interesting idea of running the humbucker outputs through an opamp inverting and noninverting outputs directly, like an "activator" (making passive pickups behave more like active pickups), which, in my tests, increase the "bite", probably because of the effect you demonstrated of moving the peak and treble roll off higher.
Yeah, in lab and homework I sometimes have my students use an artificially low value for the volume control just to exaggerate the effect and make it more obvious.
Your suggestion with the humbucker and op amp is really interesting!
@@Lantertronics Just to be clear: Not my idea. That's how (afaik) all active pickups are made. There's an interesting article on ElectroSmash website (which I think I'm unable to link, I guess) about the EMG81 (a famous active pickup) about the disassembly and analysis of the pickup itself and the circuit inside, which works as I pointed out.
The most interesting part to me is the fact that you can use different resistor values on the opamp inputs (in series with each coil) to swerve the pickup response to be more single-coil-ish. You get less of the noise-cancelling benefit inherent to humbuckers as a tradeoff, I guess.
You can also set the gain separately on the feedback loop to "match" volumes between bridge and neck pickups, which is nice.
Anyway, thank you for sharing valuable knowledge online.
Your exciting one coil with another which is kind of like a tank circuit. I wonder what the resonant frequency of the guitar pickup coil is? Is this resonant frequency within the audio range and if not; does it have any sub harmonic frequencies within the audio spectrum we can detect by ear?
The underlying tank isn't really one coil vs. the other, it's the inductance of the pickup (which is something the designer can readily control) combined with the interwinding capacitance of the pickup (which is harder to control and just kind of winds up however it winds up). The resonance bump is going to be close to 1/(2*pi*L*C). It will be less than that because of the resistance. It's well within the audio range, and it's really what gives each pickup its characteristic voice.
I have a bunch of lectures on this in my ECE4448 "Guitar Amplification and Effects" class playlist I'd invite you to check out -- you can readily find it by searching on TH-cam (I'm asking you to search instead of adding a link since TH-cam sometimes deletes links.)
Sub harmonics aren't really an issue here.
Are you using obsolete TL072 cause it's the cheapest high impedance buffer and is good enough? More expensive OPA2134 and plenty of CMOS opamps have better noise specs.
Oh yeah, there's definitely better spec op amps out there. I used the TL072 because I have a drawer full of them. ;)
They're generally a reasonable workhorse op amp for synthesizers and guitar pedals and such.
Sweet, thanks for confirming. I got a few sitting around too. Nice to use a 25 cent component when I can. Sure is in many audio circuit examples.
Humbucker-equipped guitars will typically have 500k volume and tone pots. Would that further shift the frequency plot as compared to your 100k resistor?
Using a higher resistance pot wouldn't shift it as much as the 100K pot I used. Think about it this way, using a higher resistance pot eventually is approximately like not using a volume control at all.
Helmuth Lemme has charts showing the effect on the resonant frequency of various resistance loads. Also the effect of capacitance (for example from the guitar cable) on the Q factor (height of resonant peak ~= treble emphasis). Google his "The Secrets of Electric Guitar Pickups" and also his "Pickup measuring techniques". The standard in pickup testing is not only to measure the unloaded response, but also that with 250k total resistance and 470 pF capacitance. But of course you can use any loading more typical of the conditions you want to put the pickup into. Given their typical difference in resistance loading, it would make sense to have one standard for single coil guitars and one for humbucker guitars.
500k will raise the resonant peak compared to 100k, yes. The other very important thing to remember is there will be tone controls also connected in parallel with both the volume control and the pickups, so those also will impact the resonant peak (though not that much when they are set to max level).
i actually understood some of what you said in this one! Normally its all over my head. haha
:)
ok, but what does it *sound* like? run some audio through it! human speech, guitar, and a fully mixed electronic song for a few points of reference
He’s just demonstrating how pickups behave when you load them, how dynamic they are and their dependence on guitar wiring etc, in general
@@danielsanichiban yep! that's great. but i'd still love to hear what this setup sounds like, because it's unique :) would love to see how it reacts when the items are being moved around in 3d space etc.
@@cheater00The sound output can be pretty ok except lacking bass with programme like you suggested. Varying load impedance and/or tone pot/tone cap and listening and looking at the FR plot is revealing. For proper measurement/analysis the drive coil needs to be constant current source. A power amp driving the coil through a high value resistor approximates constant current source well enough.
@@ericmc6482yea i'd love some demos :)
That sounds like a lot of fun actually -- could be an interesting effect.
When you say 80hz is the lowest frequency on a guitar, i guess you mean in standard tuning and not in metalhead drop A
On my RIGOL pu put the USB stick in, it has to be FAT32 or some such, and I hit the "print" button and it saves a screenshot.
Here I want to get the actual data to load into Excel or MATLAB and process. There's a way to do it, I just need to spend some quality time with the manual.
I'm not sure, what you want to prove here, but I'm sure it's fun anyway.
I'm not sure what I was trying to prove either. :)
I'm not sure I was wanting to prove anything at all... just wanted to measure stuff.
@@Lantertronics I've been there and seen similar things. Be careful, you might get deep into a rabbit hole. On my recent built guitar FX I noticed a permanent noise, which could come from the circuit, but it was there. Turned out, it came from the switching power supply. After using a ring transformed PSU, the noise was gone.
2:29 that looks like crossover-distortion haha
Oh, it kind of does! It's a coincidence... but is interesting.
☺️ "promosm"
I like watermelons
The guitar pickup is picking up 60hz hum and other EMI interferences that is what you're seeing I'm sure. You oscilloscope impedance is set to 50 ohms which a guitar pickup impedance output is NOT 50 ohms so that is a problem you have to change the Oscilloscope port impedance. Try to make another video about measuring the frequency response, Pickups peak frequency and Pickups impedance
But he has a buffer between the pickup and the oscilloscope.
@@moroboshidan7960 yes true he does not sure if the buffers output is 50 ohms
Have you seen the huge database of pickups analyzed by that guy Antigua? He's on a bunch of different forums and has analyzed probably hundreds of pickups by now.
Another issue might be driving the source coil with the function generator output directly with its 50ohm output impedance.
The TL072 output impedance is not going to be ideal 0 ohms but I think close enough with the oscilloscope set to 50 ohm input impedance. Would be good to measure to confirm.