|
|
Post by cynical1 on Jan 18, 2009 21:49:35 GMT -5
Yep. In theory, the ratio of the frequency ranges between the two instruments should be used to scale the treble bypass cap. If it's an octave lower, the cap should be an octave larger (Q.E.D.). The same metric could apply to the tone control cap(s) as well. Now, if you CAN find caps measured in octaves, please do let me know.... I got a guitar wiring book for Christmas that indicated that a "stronger" cap should be used. WTF  Guess I could always use the cap stretcher... Well, I spent hours looking for anything on bass treble bleeds, but came up short. And no one sells caps based on octaves either...what a surprise... Seems this isn't a valued mod for bass players... I did read enough to learn that raising the cap value high enough basically makes the tone control null and void... If you find anything definite please feel free to let me know. Otherwise I'll stick with what I have, and modify it once it's built if it just doesn't work. A few alligator clips and a bag of caps and resistors should do the trick. Thanks again for insight. Happy Trails Cynical One P.S.: On the tone mod...I was thinking of swapping the 0.047 pF cap to a 0.033pF and keep the 0.022 pF cap. This would give me slightly brighter then what I have at .033 pF, exactly what I have now at 0.022 pF, and nice and dark at 0.0132 pF. |
|
|
|
Post by ashcatlt on Jan 19, 2009 1:04:54 GMT -5
In a simple RC high pass filter the cutoff frequency (F) is inversely proportionate to the time constant(RC):
F = 1/(2pi*RC)
To double the cutoff frequency (transposing up an octave) you'd halve the time constant. To drop an octave (halving the frequency) you'd double the time constant. This can be accomplished by doubling either R or C. In this case, C is easier.
Of course, this isn't a simple RC filter, but it gives you somewhere to start.
|
|
|
|
Post by ChrisK on Jan 19, 2009 2:01:45 GMT -5
...an octave lower....an octave larger...
Sorry, I was being too clever in a clue.
If the frequency range is an octave lower (divided by 2), then, for an equal response effect (albeit an octave lower), the caps double (times 2) in value.
Yep.
Yep.
[/quote]
No, this will be the brightest.
BTW, uF (10^-6 Farads) and not pF (10^-12 Farads)
|
|
|
|
Post by cynical1 on Jan 19, 2009 12:49:50 GMT -5
In a simple RC high pass filter the cutoff frequency (F) is inversely proportionate to the time constant(RC): F = 1/(2pi*RC) To double the cutoff frequency (transposing up an octave) you'd halve the time constant. To drop an octave (halving the frequency) you'd double the time constant... Okay. Not being a math genius I looked up RC high pass filters on-line and found a Java High Pass Filter App. Starting with the currently specified 1000 nF capacitor in parallel with a 220 K resistor I took the lowest and highest notes from a guitar to see what it returned. Now this is not terribly scientific, as the applet rounds the frequencies, but It looks like this: GUITAR: Open Low E = 82.4 hz  GUITAR: 24th Fret High E = 1318.5 hz  Now for the bass, 4 string, lowest and highest note, with the same 1000 nF capacitor in parallel with a 220 K resistor: BASS: Open Low E = 41.2 hz  BASS: 24th fret High G = 392 hz  Okay, so that looks like about half to me. So, doubling the cap value from 1000nF to 2200 nF, (because that's where the applet rounds it off to...) and keeping the same 220 K resistor, we get this: BASS: Open Low E = 41.2 hz  BASS: 24th fret High G = 392 hz  Seems closer to the same curve for the guitar, to my uneducated eye... Now, I noticed that if I raised the resistor to 270 K, and kept the same 2200 nF cap, I got even closer to the curve:  Okay, that was fun, heh... Now, at what point does the values from the caps and resistors begin to effect the functionality of the pot? Does the rating of the pot (IE: 250K, 500K, 1meg...) need to change to accommodate the treble bleed module? Sorry for all the images, but the pictures make up for my babbling 1000 words... Happy Trails Cynical One |
|
|
|
Post by cynical1 on Jan 19, 2009 13:18:26 GMT -5
Cool Cool, again... Okay, unless your speaking metaphorically, you lost me. When I went looking to see what the "Woman Tone" was (pardon the bass players ignorance on that one) I noticed that in every example the tone was turned all the way down on the neck pickup to simulate the tone. I have also labored under the impression that the lower the cap value, in relation to the tone control, the more high end it rolls off. Granted, I'm not the sharpest tool in the shed when it comes to electronics, but throw me a line on this one... Happy Trails Cynical One |
|
|
|
Post by ashcatlt on Jan 19, 2009 15:07:12 GMT -5
That little tool looks like fun. I'd like to mention that the 24th fret G on the bass is not an octave lower than the 24th fret high E on a guitar. Sounds obvious, but it sounds like you might have lost sight of this while comparing the curves.
I find it interesting that the phase shift in all of these pictures make it so that the peak of the filtered wave seems to hit exactly where it crosses the original wave. Don't quite know what to make of that. It could well just be an aberration, either an accident of the numbers or just the way this applet draws things.
|
|
|
|
Post by ashcatlt on Jan 19, 2009 15:20:32 GMT -5
I have also labored under the impression that the lower the cap value, in relation to the tone control, the more high end it rolls off. But you just proved yourself that this is not the case! The relationship of time constant to cutoff frequency in a low pass filter is exactly the same as in a high pass. In this case the frequencies which are passed through the capacitor go "down the drain", while the rest of the signal goes on to the amp. |
|
|
|
Post by cynical1 on Jan 19, 2009 15:52:16 GMT -5
I'd like to mention that the 24th fret G on the bass is not an octave lower than the 24th fret high E on a guitar. Yeah, I noticed that, but I've been a few semi-tones off most of my life... I just used the doubled value for reference. If you drop the cap down to 1800 nF then you need a 390 K resistor to effect the same curve as the guitar at 1000 nF/220 K resistor.  Now, again, during my long surfings on the net to gain a better understanding of all this, I came across this:
...a lower resistor value will accelerate the taper more and more, and it will tend to be less and less frequency selective as the resistor value becomes smaller. Increasing the value of the cap, if you start with a very small one, will make the sound brighter and brighter until you reach a certain point. After that, the cap will start passing midrange frequencies as well... And in the extreme case, if you put in a large enough cap it will be equivalent to short circuiting the input and wiper lugs on the volume pot.
So, if I read this right, I need to be careful not to put too large a cap on this module or it will defeat its own purpose. And, if I follow this, the larger resistor will retard the taper, again defeating its purpose. Just need to figure out where the wall is and back off accordingly... Happy Trails Cynical One |
|
|
|
Post by flateric on Feb 17, 2009 17:35:12 GMT -5
Not sure I can get my head round all this electronics. The info I have is that a 100k resistor and 1 or 2 nF (0.001 to 0.002uF) is a common value range to use for a guitar treble mod. You're using 1000 times bigger cap value, but is this just for the tone pot or as a treble bleed?
What is likely to be a good value set for a bass?
|
|
|
|
Post by cynical1 on Feb 17, 2009 19:05:18 GMT -5
This is the cap value for the treble bleed. Tone pots use the Woman Tone Mod so graciously shared by ChrisK.
I haven't built this yet, but for the cost of a few caps and resistors...and a few alligator clips, I figure I can work out something acceptable.
Oh, and somewhere in the Archives here there's a series of posts from JohnH on the 100k versus 220K resistor for the treble bleed mod. It doesn't really apply 100% to the bass, but I trust his logic.
Setting the values for the treble mod on a bass is virgin territory, as I have yet to find it in any bass diagrams posted anywhere. Once we get this nailed it'll probably be one of the few references out there.
Shout back on what you try and how it works.
Happy Trails
Cynical One
|
|
|
|
Post by flateric on Feb 18, 2009 9:28:26 GMT -5
Heres some graphic I prepared earlier...  Shows how increasing the resistor in a parallel treble bleed enhances the treble response at low volume. Now i need to find out L,C and R values for P-bass and Jazz bass pickups  |
|
|
|
Post by flateric on Feb 18, 2009 10:50:38 GMT -5
According to the table of pickup values John posted earlier (buildyourownguitar.com/resources/lemme/table.html)
a P-bass pickup will be approx 6H, 15pF winding capacity and you should expect about 6K resistance.
Jazz bass is 3.6H, 150pF and about 5.5K resistance.
You can plug these results in and get a good treble bleed using 100 - 200k R and 0.6 to 2.2nF cap in parallel.
I will make use of a spare DPDT mini switch and try wiring in a 0.6nF and a 2.2nF cap into the circuit to give them a try with about 180K resistor.
|
|
|
|
Post by ChrisK on Feb 18, 2009 17:37:26 GMT -5
No, methinks it's just a matter of using nano Farads (nF) when pico Farads is correct.  |
|
|
|
Post by flateric on Feb 19, 2009 11:34:54 GMT -5
Oops, that embarassing!  |
|
|
|
Post by sumgai on Feb 20, 2009 4:46:20 GMT -5
C1, The rule of thumb when designing with capacitors is "the larger the value, the lower the frequency it will pass". Stated conversely, which is handy sometimes, you'd say "the smaller the cap's value, the higher the frequency must be in order to pass through". You could re-phrase it numerous other ways, I'm sure. What's not obvious here is that the capacitor does not hit a magic frequency (for its value), and then pass all higher frequencies at the same signal strength. There is a slope here, and it will tend to roll off and eventually appear to be flat, but that's covering several decades. (To keep the math simple, we can use musician's terms, so substitute octaves for decades - the final numbers will change a tiny bit, but the overall effect is the same, a sloping response curve.) The essence of a treble bleed is to allow some of the higher frequencies (the "treble") to bypass the volume control. How much bypassing you get is affected by several factors not yet mentioned. First, remember that no signal is "pure", there are lots of harmonics involved. That said, when you have no reactive components in a circuit, you don't modify the relative strengths of those harmonics (unintentionally or otherwise). But by introducing a capacitor, you now affect the relative strengths of those harmonics. Remember, the rule above says that higher harmonics (aka higher frequencies) will tend to flow more easily through a cap (for any given value) than the lower ones, not to mention the fundamental itself. This means that you've now re-shaped the signal itself, hopefully for the better, or at least acceptably. But in point of fact, the higher harmonics are now stronger, Q.E.D. Ergo, your graphing will only get you to a ballpark starting figure, and as you rightly point out, the bass end of this mod has not gotten a lot of "press", so you're really "plucking numbers from the air", aren't you? Experimenting after the build is completed is gonna have to be the order of the day, I'm afraid. So my mouth's been running, now you wanna know where my money is on all this, eh? Glad you asked. First, recall that all values you see in example circuits are dependent on the volume control's resistance value - the three components each react to value changes in the remaining two parts. There's a lot of room to play, but generally, if you have a 500K pot in a humbucker circuit, then you'd probably want at least a 150 to 180KΩ resistor, paired with a 0.001µf cap. For a 250KΩ pot, drop down to a 100KΩ resistor, leaving the cap the same. Now, those values are all for guitars. For your bass, double the cap values, and see how that sounds. Don't change the resistor values for awhile, just play with the caps. The reason is, the resistor has an affect on the pot's overall taper. Monkeying around here is OK, but just be aware that you might make things worse, before you make them better.  You can learn a bunch more from your friend, Mr. Google. I used the search terms "guitar treble bleed circuit" Substituting the word 'bass' for 'guitar' brought almost identical results, meaning, as you first surmised, basses have not been treated equally on this topic. HTH sumgai |
|
|
|
Post by cynical1 on Feb 20, 2009 11:48:42 GMT -5
sumgai -
Thanks for coming out of the shadows on this one.
Thanks for the insight on the resistor. I've read a good deal of what's out there on on the Internet regarding treble bleed mods. I've read about how raising the resistor value too high will effect the taper of the pot...but nothing hard and fast on where this begins to occur. The rule of thumb on 100K for a 250K pot and 150K-180K (and JohnH's explanation of up to 220K) on a 500K pot helps quite a bit...especially for the empirically rooted types...such as myself...
I've played with the spreadsheet from JohnH a little more and I'm starting to get a better mental image of how all of this interacts. The fog is beginning to lift.
Since beginning my minor odyssey into digital recording I notice the overtones and harmonics being the first things that suffers from rolling off the tone on the bass. As I mentioned before, I don't expect too many changes in the fundamental range of the bass, but I would like to preserve the ability to pull harmonics with the tone turned down...
I'm also intrigued by flaterics observation on the overall change in tone with the treble bleed mod on the bass. Some sound samples would be real nice (...hint...hint...)
I've got some time before I can get to my bass mods and this is all very helpful in the planning stages. Everyone's input is greatly appreciated.
My, how this posting has wandered...
Well, back to the taxes...
Happy Trails -
Cynical One
|
|
|
|
Post by ashcatlt on Feb 20, 2009 11:49:16 GMT -5
First, I'd like to say that it's always great to see you come around sumgai, however briefly.
Second, I'm getting a little dizzy trying to follow this conversation, since it spans two separate threads.
Third, I'm still not convinced - especially considering the empirical observations we've got from flateric's build - that this mod has anything to do with the tuning of the strings. The roll off for which we're trying to compensate doesn't care about the tuning, so why should the compensation mechanism?
Last, an on-board or cable-borne buffer ala JohnH would more or less obviate the need for this.
|
|
|
|
Post by JohnH on Feb 20, 2009 16:02:38 GMT -5
Hello, nice to see you Mr S!
I just had a scan back through this thread, and Ive lost track of what it is about, but basses and treble bleeds seem to feature more than volume controls, so thats what im talkin'about.....
I never paid much attention to basses until recently, but Im learning that the upper frequencies can be very important, maybe even more than for guitar. Some bass cabs have tweaters, while guitar cabs do not. Also the, Boss 7-band EQ pedals have sliders up to 6.4kHz for guitar, while the bass version goes up to 10kHz. I suppose that, while there may not be much signal up there, what there is can be very important to the percussive character of the bass sound. Also, we are generaly keeping our bass signals clean, so theres no nasty distortion harmonics up there, while overdriven guitar depends on that high roll off to make it bearable.
If you really want to keep the full range from the bass, and are up for experimenting, the buffer cable or built in buffer is the way to go. It doesnt just keep the full volume sound intact - it will extend the range upwards to provide everything that is available. On my spreadsheet, you can test it by pressing the 'buffer' button. Its particularly true if you set lemmes values for the precision bass pup, with the very low winding capcitance of 15pF = 0.015nF (I wonder if that value is true?).
With a treble bleed (and no buffer) on the P-Bass pup, I had a fiddle to see what values looked best. I do this by looking at the full-volume response shape, and noting the height and frequency of the peak response. Then I try to match that shape with a reduced volume say -6db down for the low frequencies. You can't match exactly, but I found with a 10' cable, 220k and about 1.5nF was looking good. The resistor, which help to keep the shape correct, change the pot taper, so that -6db point happens at about 5.5 on the volume pot turn (audio pot) , instead of about 8, which I think is a good change.
cheers
John
|
|
|
|
Post by flateric on Feb 20, 2009 16:35:39 GMT -5
I'd go along t=with that conclusion on the treble bleed values - this is a parallel treble bleed. When I've figured out a way of posting sound clips I'll give it a go. The higher frequencies add all the flavour to bass notes - the percussive twang of the fingers on strings and strings grazing frets, without these, the sound is like an old upright with rubber bands. Just turn off the top eq slider of a bass amp if you have one handy and listen how flat the tone is.
The treble bleed does modify the tone ctrl taper, adn I tried to put it into clumsy words earlier, but its not easy to do. The difference between a 650pf and a 2.2nf cap is hard for my ears to pick out, but the difference in tone sweep is there.
|
|
|
|
Post by newey on Feb 20, 2009 20:12:05 GMT -5
Several have made the point that this thread long ago ceased to be about 3 Vol pots. And that we have 2 simultaneous discussions of the bass treble cut topic. And also, that the rest of the net is eerily silent on the topic. All true and valid points.
I would therefore invite the principal posters to corral their collective findings, although preliminary, into a single coherent thread to assist those who may be interested- and who are not likely to find it here in the coffee shop. This would logically go in the design modules section, since it is, errrr . . .a module.
I also note that this discusssion has been confined to a parallel wired module. Has any one considered wiring a series treble bleed? I don't know what the pros and cons might be, obviously the values would need to be different. But I seem to recall some sort of a series-wired bleed circuit on old Ric basses?? I could be misremembering that, lot of years ago when I looked inside one.
|
|
|
|
Post by JohnH on Feb 20, 2009 20:39:13 GMT -5
A module section on treble bleed circuits would be useful.
The Series treble bleed, attributed to Kinman (famous Aussie pickup guru), usually comprises 130k or 120k in series with about 1.2nF. It does not affect the volume pot taper, unlike the parallel version. You can test on my spreadsheet - theres even a button for it. I think it is less good than the parallel version at keeping he shape of the frequency response intact however, since it also adds some extra mids to the response peak.
John
|
|
