|
|
Post by JohnH on Feb 19, 2014 2:47:18 GMT -5
Moved, as previously posted. This post is about tone capacitors in a normal treble cut guitar tone control. Tone CapacitorsThis simple component can cause confusion, and expense for those who pursue vintage mojo. Personally I think the type of cap makes very little difference, only the value is relevant, and that is all we can analyse here. One of the things we know about caps in a guitar tone control is that at maximum settings of the tone pot, the actual cap value makes very little difference. How little? with 250k pots and a typical single coil pickup, less than 0.25db of difference anywhere over the full spectrum of an electric guitar, even over an extreme range of values from 0.002uF up to 0.1uF. Let's consider a typical 0.022uF cap, at a frequency of say 3kHz at which we expect tone controls to be doing something. The impedance of that cap at that frequency is 2.4k, which is much less than that of the 250k pot that it might be connected to. And the way of combining the two is such that we take the root of sum of squares of the two values, like on a right angle triangle. So the net impedance of the cap and pot at max is 250.01k, ie the same 250k to all intent. But once we turn down the tone pot, the cap starts to be more significant. But when and how much? Here is the first output. It is for a guitar with a single coil pickup and two 250k pots for tone and volume, testing the most common range for tone caps of between 0.01 and 0.047uF (10nF to 47nF). The key to the chart shows knob setting/cap value. Knobs are assumed to be on log pots with 10% at midway, over a range 0 to 10. The chart shows decibels v frequency, of the guitar acting as an electrical filter:  The top two traces relating to 10 and 8 on the knob, are in fact, each two traces with 10nf and 47nF, and for reasons described above, the results from different cap values effectively coincide. At 6 on the knob, there is a slight visible difference, shown by two pale blue lines. Below about 5 in this case, the difference starts to become more marked, and very significant at tone-0. Below about 2, the low midrange peak starts to emerge, which is where the cap value makes most difference. The three red lines relate to 10, 22 and 47nF, at minimum tone setting (10nF is highest).[/p]
So tone cap values are best picked to find your favourite sound at low tone settings, knowing that they have almost no effect at high settings
Here is the same set of readings using 500k pots, and the PAF humbucker. With this, there is even less effect at high tone setting and the knob needs to be below 5 before the cap value has an effect. also see how all the peak frequencies shift down the scale, due to greater pickup inductance.
|
|
|
|
Post by JohnH on Feb 19, 2014 2:55:47 GMT -5
Small Tone CapsCarrying on from above, with the tone pot at minimum (or not there at all), varying the tone cap creates a moveable peak in the response that can either slightly darken the natural pickup sound, or bring the resonance right down into the mids. Here is a range of small tone caps, with 250k Volume pot and SC pickup:  The right-most trace has no added cap and shows the natural pickup output, as the tone pot is 'no-load' THe left trace represents 10nF, which matches the lowest of the values tested in the previous post I've tried a few extra caps like this, and find in practice that about 2.2nF makes nice difference, and smaller than that, there's not as much audible change that you might expect from these plots. So it's worth a try but important to experiment with the real thing.
|
|
col
format tables
Posts: 474
Likes: 25
|
Post by col on Feb 21, 2014 16:01:32 GMT -5
Hi John,
These are useful plots. Thanks for posting this.
Although I wouldn't expect it to be as interesting, might you consider posting plots for bass cut circuits?
Col.
|
|
|
|
Post by ozboomer on Feb 21, 2014 17:48:52 GMT -5
Although I wouldn't expect it to be as interesting, might you consider posting plots for bass cut circuits? I think it would be useful as well.. 'coz then we'd see the response curves for passive low-pass filters (LPF = treble-cut), high-pass filters (HPF = bass-cut) and of course, the combination we might have ( with my pet G&L tone control  ) a band-pass filter (BPF). Now, if only we could easily change the slope of the response curves on the fly (without any active electronics)  ... |
|
|
|
Post by JohnH on Feb 21, 2014 18:21:23 GMT -5
Although I wouldn't expect it to be as interesting, might you consider posting plots for bass cut circuits? I think it would be useful as well.. 'coz then we'd see the response curves for passive low-pass filters (LPF = treble-cut), high-pass filters (HPF = bass-cut) and of course, the combination we might have ( with my pet G&L tone control ) a band-pass filter (BPF). Now, if only we could easily change the slope of the response curves on the fly (without any active electronics) ... Bass cut and G&L PTB is the next subject, will be posted probably today. |
|
|
|
Post by sumgai on Feb 21, 2014 23:01:42 GMT -5
.... Now, if only we could easily change the slope of the response curves on the fly (without any active electronics) ... Well, the easy way would be to install additional cap(s) with a switch to control it/them. Basically, a single cap and resistor setup is a second-order filter (regardless of which portion of the frequency spectrum it acts upon). That said, the slope for such is generally considered to be 3dB per octave. By adding additional sections, one can double, triple, quadruple, etc the slope. Although it's pretty rare for someone to "need" a slope greater than 24dB per octave, that's pretty steep by most accounts. The down side is that you also move the "knee" of the response curve - the drop-off frequency will either double or be cut in half, depending on whether the caps are in series or parallel. But at least it's all passive.  Any switching can also be setup to control the variable resistance. Perhaps adding a second resistance element from a dual-concentric pot, or switching over to that second element alone, because it's not the same as the nominally normal pot value.... there's room for a lot of experimentation here! HTH sumgai |
|
|
|
Post by ozboomer on Mar 12, 2014 6:00:22 GMT -5
.... Now, if only we could easily change the slope of the response curves on the fly (without any active electronics) ... [...] Basically, a single cap and resistor setup is a second-order filter (regardless of which portion of the frequency spectrum it acts upon). That said, the slope for such is generally considered to be 3dB per octave. I've read about that 3dB/octave figure... but I'm curious now... (for those who came in late...) When we play a note of a certain frequency and then play a note with double the frequency, we say the interval between the notes is an octave. Combining these 2 thoughts, why is it that JohnH's plots show that, say, when we look at 1000 Hz compared with 2000 Hz (an octave), the difference in response is more like 15dB rather than 3dB? ...*thinks*... Maybe the pickup in-circuit is making a difference? |
|
|
|
Post by JohnH on Mar 12, 2014 7:22:55 GMT -5
My plots are based on the convention whereby if the voltage halves, that will result in 1/4 the power when the signal finally gets to an output, and that is -6db. So in the way of my plots (and usual convention I think) a basic RC filter would have an output that falls 6db per octave. The top end of the plots in these threads fall much faster than that however, since they are not just RC networks but also have inductance active including resonant effects near the peak
|
|
|
|
Post by sumgai on Mar 12, 2014 12:03:44 GMT -5
ozzy, John's correct, of course, although we're talking about miniscule amounts of power here, not the big honking stuff coming out of an amp. Just remember that as we affect the voltage of something, so do we also affect the current involved in that circuit. It is fortunate for the engineering world that by simply stating a linear voltage value in logarithmic terms, we have shortcutted the massive amount of math needed to represent how current and voltage interact at any given frequency when passing through a cap or a coil, or both. (In my usual manner of attempting to obey the KISS principle, I'm leaving out any references to phase shift, power factors and other esoteric considerations. After all, that's why I just said that converting to the log scale takes care of that for us, in a most handy fashion.) Additionally, the 3dB figure is axiomatic - it never changes, because it's a ratio, and ratios don't change due to frequency, no matter if they're represented in linear or log fashion. But.... Both you and John have hit it right, the coil (inductance) is in play as well as the simple cap-and-resistance part of the circuit. The interplay between all three elements is what's driving the circuit behavior, and now, where we might expect to see a 6dB (simple) voltage drop, it's gonna be more like 12dB in log-speak. And that's only good for a certain portion of the frequency spectrum, near the resonance point. Let's not get any closer than that for the moment. KISS and all that. HTH sumgai |
|
|
|
Post by ozboomer on Mar 12, 2014 18:05:11 GMT -5
Both you and John have hit it right, the coil (inductance) is in play as well as the simple cap-and-resistance part of the circuit. [...] Let's not get any closer than that for the moment. KISS and all that. Ahh, Okie... Heh... I still have plenty of memories of the frustrations I had with trying to understand all the "real maths" in electrical engineering some decades ago... especially when I was trying to get a grip on the effects of a 6db/octave filter compared to a 24db/octave filter when I was building synthesizer modules... ...but the notion of "a little knowledge is a dangerous thing" comes to mind a bit... and that's why I wanted to clarify that the "axiom" (3db/octave) doesn't really apply here 'coz of the pickups being in-circuit. Don't get me wrong -- I love playing about with JohnH's spreadsheet (at least at work, where I have Excel) to help "direct" some of my experimenting -- but I'm kinda nutzo about better-understanding these "computer/software tools" we use... 'coz in work, as well as the IMPORTANT stuff (music and such), I see people frequently taking the tools' output as gospel instead of using their eyes 'n ears; in our case, it's all about us understanding how we can get the sounds we want rather than whether it's "right" by the books(!) |
|
nikogo
Meter Reader 1st Class
Passerby
Posts: 51 
Likes: 0
|
Post by nikogo on Feb 21, 2015 15:37:19 GMT -5
"Tone CapacitorsPersonally I think the type of cap makes very little difference, only the value is relevant, " Actually different types of capacitors of the same nominal value give a bit different tone that is hard to describe. Some people use description like "harsh", "clean" and so on. There is video on YouTube with a demo. The paper-in-oil and mylar based capacitors are voted as preferred. I believe those film/paper wound capacitors have higher internal inductance that helps to reduce the micro-resonances in the "pickup coil - capacitor" contour and eliminate unwanted oscillations. |
|
nikogo
Meter Reader 1st Class
Passerby
Posts: 51
Likes: 0
|
Post by nikogo on Feb 21, 2015 15:42:20 GMT -5
Sorry for some mess.
I am a newbie here.
Please remove my first flub
|
|
cooltone
Apprentice Shielder
Posts: 27
Likes: 6
|
Post by cooltone on Mar 26, 2015 17:41:12 GMT -5
Hi John, Apologies for not replying for so long! .... Your plots are great and exactly what I have found (See thread on Parallel Humbuckers). A quick update: I implemented a double pole-six throw switch with five caps + no cap on a parellel HB, with no resistor between the caps and the HB, on both pups. (Actually, I modelled the whole lot on LTSpice, even developed a model for a 6PDT switch to plot all switch positions ) I have been playing with this set up for a couple of years now (pubs mostly). The result ..... it is a fabulous set up. Before this I almost never used the tone controls, now I use them all the time. Here's my experience of use: 1) The characteristic sound is set by the resonant peak: I can go from shimmer to jazz sounds easily 2) The pups are more usable at low settings - amp noise is usually the limiting factor at 3 and below 3) Distortion control isn't just good, it's brilliant Regarding distortion here's what I believe is happening. It is the resonant peak that reaches the amp non-linear region first. The amp non-linear region introduces higher frequency harmonincs, which makes for a harsh sound. When a larger cap is switched-in the res. peak is at a lower frequency, so lower harmonics are generated. The result is cleaner 'crunch' sound and 1970's LP lead sounds - as a blues player I'm in heaven! GTG - still no time  , Have fun and thanks for all the work you put into this board. Happy to send you LTSpice models if you want just need an addy. Cooltone! PS: I believe that adding extra caps to increase slope is unlikely to work in a passive resonant circuit because the L can only resonate with one cap. This is why there are two peaks in the first graph, the right peak is resonance with the HB capacitance, the left peak is resonance with the tone cap and in between, as the two caps try to resonate at different F's, the different phases of both kills the resonance frequencies of both. |
|
|
|
Post by Ro_S on Nov 4, 2018 2:39:22 GMT -5
JohnH - You wrote above that, "at maximum settings of the tone pot, the actual cap value makes very little difference." So is that insufficient affect for one to perceive audibly the differences between different tone cap values (when at nil attenuation on the treble cut tone control)?
|
|
|
|
Post by JohnH on Nov 4, 2018 6:32:53 GMT -5
Yep, the difference even with a large cap value change is way too small to hear when tbe pot is at max.
|
|
mongoose
Rookie Solder Flinger
Posts: 3
Likes: 0
|
Post by mongoose on Jan 29, 2021 12:30:24 GMT -5
Ok Ok... so if capacitors have so little involved with the tome, why are they installed in tone circuits....??
So heres my take. The signal that comes out of a transducer is or should be digital. Thats because thats how transducers work... analog voltage controls the driver and digital signal is processed to control the output. Now I know this is sounding funky... with humming and crackling and hissing sometimes, but what we want is as clear a digital signal as we can get.
Assuming the amp is dumping some 46v into a set of transducers that are registered at about 50v, this is just not acceptable in digital signal processing because the transducers max rating is just to close to the input volts... and this amount of power will make the low end of the transducers impossible to reach--- because they will only 'scream' feedback often and the sustain on the strings will probably be infinite.
This is only an example of why tones don't work sometimes... there must be a +/- center tap and the connection here can and will change the circuit capabilities.
The caps are supposed to make the power into a digital signal by firing at a specified rate, where a .0022uf(22nf) cap placed inline will make a digital signal that is very fast, it just makes pulses until it is divided by the resistance to ground. That makes a variable RC circuit that can be set to a specific digital frequency. The circuit needs to be placed right... in the PathOfLeastResistance, for it to work, but it does work.
The problem I am having with the Baxandal is that the resistance is so high that it lowers the volume too much for these low power transducers, to compensate. Even when I raise the transducers to the strings, they just aren't enough to fire the drivers.
I am beginning to think that I am going to need a book on advanced electronics.... or a raise.
|
|
veliko
Rookie Solder Flinger
Posts: 7
Likes: 0
|
Post by veliko on Apr 19, 2022 17:19:56 GMT -5
JohnH there is something I don't understand. In your first 2 graphs where we observe gradually turning the pot down, we see peaks flattening and eventually disappearing. Why is the pickups' response peaking again when the pot is at 2 and 0 when there were no peaks when the pot was at 4? Also why is the the change of the dB axis so sudden when going from 2 to 0 on pot?
|
|
|
|
Post by JohnH on Apr 20, 2022 8:29:28 GMT -5
JohnH there is something I don't understand. In your first 2 graphs where we observe gradually turning the pot down, we see peaks flattening and eventually disappearing. Why is the pickups' response peaking again when the pot is at 2 and 0 when there were no peaks when the pot was at 4? Also why is the the change of the dB axis so sudden when going from 2 to 0 on pot? Hi veliko, the peaks that we get at max volume/tone, and also when tone is at minimum are resonances. These occur when the inductive coil is directly connected in parallel with a capacitance. Its analogous to a mass bouncing freely on a spring, with no damping. At min tone the capacitance is the tone cap and and max volume and tone, it is the capacitance of the guitar cable which is much less so frequency peak is higher. At other settings of the pots, there is no direct cap to coil connection, except via resistances in the pots. These damp the free resonance squashing the peak, like shock absorbers controlling a car bouncing on its suspension. In between, the transition is set by the maths of circuit. It doesn't take much resistance to kill the low peak, dononly a small pot turn does it. |
|
carlosg
Rookie Solder Flinger
Posts: 17
Likes: 0
|
Post by carlosg on Nov 21, 2022 9:22:03 GMT -5
Great chart with small tone caps. We can change resonance peak with soldering cap between hot and ground. Can you do chart with caps in hot wire, between wire and pot lug? It's bass cut, I'm wonder it only bass cut, on maybe move resonance peak to the riht too.
0.047uF Duncan's recomend to muddy neck pickup. It cuts bass a little. Smaller cap value eg 0.01uF cuts more bas.
|
|
|
|
Post by gckelloch on May 6, 2023 14:58:30 GMT -5
|
|
|
|
Post by sumgai on May 6, 2023 23:17:18 GMT -5
I'm going to leave those loooooong links alone, for reference purposes. But in my time, I've learned that Amazon (and other sites) like to add a lot of info to a link that they think is necessary. It's not. Perhaps they do tracking of some kind, I don't know, but what I do know is that almost any Amazon link can be shortened, sometimes drastically. Compare the above (quoted) to these short versions: www.amazon.com/Kootage-Resistor-Assortment-Tolerance-Resistors/dp/B0BFRNTVZFwww.amazon.com/EEEEE-Multilayer-Monolithic-assortment-Individual/dp/B094HRK8QXWhat I usually do is highlight the link in the address bar, find the first sequence of random letters after the textual descrition, and delete everything thereafter. If that doesn't work, I restore the whole thing, and find the next break that looks promising. Things like a backslash, then a few characters followed the equal sign, or the question mark followed by one or two characters and the equal sign, those are prime candidates of where to start removing stuff. It certainly makes things easy to understand, but then again, there's a menu choice above the text box that allows one to insert the whole link, enter some descriptive text, and hit "Do it". Presto, the text appears as a link, and the underlying link itself does not clutter up the screen. Works wonders. However, sometimes the link is so long that the redirect chokes up. Then you'll have to shorten the link as I've described. HTH sumgai |
|
|
|
Post by ozboomer on May 7, 2023 1:36:05 GMT -5
... The Amazon links are very long: ... I'm going to leave those loooooong links alone, for reference purposes. But in my time, I've learned that Amazon (and other sites) like to add a lot of info to a link that they think is necessary. It's not. Perhaps they do tracking of some kind, I don't know, but what I do know is that almost any Amazon link can be shortened, sometimes drastically. ... sumgai Some further explanation... For folks who aren't au fait with the CGI ( Common Gateway Interface) and web programming... When you see a '?' in an URI/URL (which are different things: URI & URL), that is most often a delimiter for a {variable=value} pair. So in general, as soon as you hit the first non-quoted '?', everything else can be removed, together with the '?'. It's most often about when folks use 'dynamic web pages' on their web sites... and what you see is a function of a lot of things... and many of those things go into '?var=value' sequences. O'course, this is 'old (1990s) tech' (and unsecured, as the 'var=value' is plain text flying around TheInternet TM) these days... and APIs (Application Programming Interfaces) are hidden behind the web pages... and use encryption and other things to better secure the information... but it's still the same sort of put/get mechanism. Your browser is like a 'formatter' of what the web site sends back to you when you 'enter' a web site address. True... but note that 'hiding' the links is a mongrel for usability... However, most folks tend to not be using Lynx, Cello or similar web browsers anymore... web crawlers don't care... and visually impaired folks are using smarter tech that interpret links 'better' these days... So the argument is, perhaps, moot.. Pfft!..but WWW and how it's actually implemented and navigated is a big topic... for elsewhere... |
|
blujosh
Rookie Solder Flinger
Posts: 2
Likes: 0
|
Post by blujosh on Jan 21, 2024 16:51:26 GMT -5
since this is my first post, i'd first like to extend my gratitude for the wealth of knowledge on this thread and this forum in general.
i had a question relating to "tone" caps, so thought this would be the best place to post this. if my inquiry should be its own thread, please just let me know.
after reading all of the info above, i got to wondering it's possible to emulate a "fixed tone" setting in volume-pot-only circuit by wiring up a cap between the vol pot input lug and ground.
for example, would a circuit with a 500K vol pot + 500k tone pot w/ 15nF cap (standard/modern tone wiring) with tone pot at "0" be similar to just a 500K vol pot (so, no tone pot) with that same 15nF cap wired across the input lug and ground of the vol pot?
if not, is there a way to create a "fixed tone" setting in a volume-pot-only circuit?
thanks in advance!
_jc
|
|
|
|
Post by reTrEaD on Jan 21, 2024 18:57:00 GMT -5
if my inquiry should be its own thread, please just let me know. As long as your question(s) are tightly related to the general topic in the OP, posting here is fine. However, if you need extensive help regarding your particular issue, it might be to your benefit to create your own thread. after reading all of the info above, i got to wondering it's possible to emulate a "fixed tone" setting in volume-pot-only circuit by wiring up a cap between the vol pot input lug and ground. for example, would a circuit with a 500K vol pot + 500k tone pot w/ 15nF cap (standard/modern tone wiring) with tone pot at "0" be similar to just a 500K vol pot (so, no tone pot) with that same 15nF cap wired across the input lug and ground of the vol pot? Yes, that's certainly a viable strategy. But without a pot to control the amount of cut (or a switch to bring that cap in or out of the circuit), you're stuck with whatever you preset. So you'll want to choose a value that tames things down a bit but isn't so severe that it sounds dull. A couple of things worth noting: 1 - With no resistance at all in series with the 'tone' capacitor, you'll have a peak in the response curve (which is much lower in frequency than the peak without a capacitor) and a rather sharp rolloff above that peak. If you want to reduce the peak, a small resistor (say 2.2k ~ 4.7k) will reduce the peak but still provide a similar rolloff to the capacitor-only tone-cut). 2 - If you only have 1 hole for a volume pot and no other holes in your pickguard, you may wish to temporarily remove your volume control and install a rotary switch so you can audition several different values of tone caps (and/or tone-cap with series resistor combinations). IF you do that, just use a fixed resistor similar in value to the volume pot you intend to eventually use. This will emulate the loading of the volume pot on the circuit. The input and output will be on one end of this fixed resistor and the other end of the resistor will be connected to ground. |
|
blujosh
Rookie Solder Flinger
Posts: 2
Likes: 0
|
Post by blujosh on Jan 21, 2024 20:07:01 GMT -5
after reading all of the info above, i got to wondering it's possible to emulate a "fixed tone" setting in volume-pot-only circuit by wiring up a cap between the vol pot input lug and ground. for example, would a circuit with a 500K vol pot + 500k tone pot w/ 15nF cap (standard/modern tone wiring) with tone pot at "0" be similar to just a 500K vol pot (so, no tone pot) with that same 15nF cap wired across the input lug and ground of the vol pot? Yes, that's certainly a viable strategy. But without a pot to control the amount of cut (or a switch to bring that cap in or out of the circuit), you're stuck with whatever you preset. So you'll want to choose a value that tames things down a bit but isn't so severe that it sounds dull. A couple of things worth noting: 1 - With no resistance at all in series with the 'tone' capacitor, you'll have a peak in the response curve (which is much lower in frequency than the peak without a capacitor) and a rather sharp rolloff above that peak. If you want to reduce the peak, a small resistor (say 2.2k ~ 4.7k) will reduce the peak but still provide a similar rolloff to the capacitor-only tone-cut). 2 - If you only have 1 hole for a volume pot and no other holes in your pickguard, you may wish to temporarily remove your volume control and install a rotary switch so you can audition several different values of tone caps (and/or tone-cap with series resistor combinations). IF you do that, just use a fixed resistor similar in value to the volume pot you intend to eventually use. This will emulate the loading of the volume pot on the circuit. The input and output will be on one end of this fixed resistor and the other end of the resistor will be connected to ground. thanks reTrEaDif i'm understanding correctly, the small/fixed resistor would be taking the place of the resistance that would otherwise be generated by the tone pot? if so, would the following be (roughly) equivalent? a) a vol + tone circuit, that uses a 500k volume pot and a 500k linear tone pot w/ .022uF cap (using "modern" tone wiring) with the tone pot always at "5" (so, roughly 250k) b) a vol-only cirucuit that uses a 500k volume pot but has that same .022uf cap + a 250k resistor wired in series between the input lug of the volume pot and ground if not, what am i missing? thanks again! |
|
|
|
Post by JohnH on Jan 21, 2024 21:05:49 GMT -5
A couple of temporary flying leads to connect where a tone control would go, taken outside the guitar, then you can attach various R and C combinations, (maybe the R is a pot), to find out what suits you. Then you can hardwire your choice inside the guitar.
|
|
|
|
Post by reTrEaD on Jan 21, 2024 23:45:13 GMT -5
if i'm understanding correctly, the small/fixed resistor would be taking the place of the resistance that would otherwise be generated by the tone pot? Yes. if so, would the following be (roughly) equivalent? a) a vol + tone circuit, that uses a 500k volume pot and a 500k linear tone pot w/ .022uF cap (using "modern" tone wiring) with the tone pot always at "5" (so, roughly 250k) b) a vol-only cirucuit that uses a 500k volume pot but has that same .022uf cap + a 250k resistor wired in series between the input lug of the volume pot and ground Not exactly. Right concept, wrong values. The pots we use for tone-cut circuits in most guitars are 'audio taper'. In modern guitars we typically use A2 taper (10%). In other words, at the midpoint of the rotation, there is 10% of the total resistance between the wiper and the CCW lug. So for a 500k A2 taper, that means 500k x 0.10 = 50k. So if your aim is to emulate having the tone control at "5" you'll want a 50k resistor in series with your .022uF capacitor. And yes, that series combination would be connected between the CW and CCW lugs of the volume pot. |
|
) a band-pass filter (BPF).
...
, Have fun and thanks for all the work you put into this board. Happy to send you LTSpice models if you want just need an addy.
