|
Post by ashcatlt on Mar 5, 2012 6:35:01 GMT -5
I would think that the new mid scoop is way too wide. It will probably sound a lot like turning down the treble cut (just compare the two graphs). Sure, it starts to slope upward past 5K, but no guitar amp is going to reproduce those frequencies, and I would submit that there's nothing in that range that any of us want to hear out of an electric guitar.
|
|
|
Post by reTrEaD on Mar 5, 2012 8:45:29 GMT -5
Ash, I don't think you should dismiss the last couple of octave in the frequency response so easily. Even though the highest fundamental frequency a guitar can produce is much lower than that, harmonics are an important part of the tone. The width of the cut seems a little wide to me, too. But if it's made too narrow it will sound a bit notchy.
Newey, I found some old notes of mine regarding LC circuits.
Resonant frequency = 1/2π√LC
- Increasing inductance decreases resonant frequency. - Increasing capacitance decreases resonant frequency. - Increasing inductance while decreasing capacitance increases "sharpness" of frequency response curve. - Decreasing inductance while increasing capacitance decreases "sharpness" of frequency response curve.
So when you're fiddling around with the component values, decrease (or increase) the inductance and capacitance by the same factor to shift the frequency without changing the width.
If you multiply the inductance and divide the capacitance by the same factor, you'll keep the same resonant frequency but narrow the width.
I don't have any notes regarding a resistor in parallel with just a capacitor. If the inductor and cap were in parallel with each other, a resistor would broaden the resonance. But since they are in series with each other, I don't know what that resistor in parallel with just the cap accomplishes.
|
|
|
Post by newey on Mar 5, 2012 12:22:44 GMT -5
Definitely needs more playing around with the values then. I seem to either have a "spike" or a "basin" with nothing in between.
Just off the top of my head, for a true "mid scoop" effect, I would think a fairly well-defined notch covering about 750Hz to 2KHz would be what we'd want to see- and so far I'm not getting anything remotely like that with this type of circuit (I do understand that it's not going to look like the cut of a graphic equalizer or anything, but still . . .)
|
|
|
Post by ashcatlt on Mar 5, 2012 15:09:30 GMT -5
Ash, I don't think you should dismiss the last couple of octave in the frequency response so easily. Even though the highest fundamental frequency a guitar can produce is much lower than that, harmonics are an important part of the tone. Well, I suppose it depends on your definition of tone. Some people like the sound of a cheap Ovation plugged straight into the PA with the treble cranked... But I don't want to hear that crap up around 10K. By that point we're way up into higher order harmonics. The actual multiples of the string frequency will be pretty far down in amplitude compared to the fundamental. What is pretty strong in this region is the non-harmonic content created by the near step function which we call pick attack. That is, it's mostly noise, and sounds like it. Oh, then there's the actual noise...
|
|
|
Post by sumgai on Mar 5, 2012 16:26:08 GMT -5
|
|
|
Post by reTrEaD on Mar 5, 2012 17:10:38 GMT -5
Well, I suppose it depends on your definition of tone. Some people like the sound of a cheap Ovation plugged straight into the PA with the treble cranked... It doesn't seem like that would sound very good. But I don't want to hear that crap up around 10K. By that point we're way up into higher order harmonics. The actual multiples of the string frequency will be pretty far down in amplitude compared to the fundamental. What is pretty strong in this region is the non-harmonic content created by the near step function which we call pick attack. That is, it's mostly noise, and sounds like it. Oh, then there's the actual noise... 10k is where the pretty stuff lives. The sparkle. Without that, you miss a lot of the good shiz. But it's all about moderation. Too much and it sounds brittle, especially if there's also peak at 5k or so. Given the choice between too much and none, I'd choose none. But what I'd really like is just the right amount. Just off the top of my head, for a true "mid scoop" effect, I would think a fairly well-defined notch covering about 750Hz to 2KHz would be what we'd want to see- and so far I'm not getting anything remotely like that with this type of circuit (I do understand that it's not going to look like the cut of a graphic equalizer or anything, but still . . .) I think you might be happier with the notch centered a little lower, say at 800Hz or so. An octave on either side of that puts you at 400Hz to 1600Hz. But it's all a matter of personal preference. Definitely needs more playing around with the values then. I seem to either have a "spike" or a "basin" with nothing in between. I reckon one thing that's complicating matters here is the large amount of internal inductance in the pickup. That's skewing the "dip", asymmetrically. Just for craps and giggles, change the 2.5H inductance in the source to 2.5uH and run the sim again. In addition to getting rid of the treble peak, the dip should be more symmetrical. At least that would be my guess. I realize you can't just remove the inductance from the pickup. But it will help understand why you are having trouble getting a "normal" looking curve. At the end of the day, you'll use the sim (with the correct 2.5H value for source inductance) to get something that looks as reasonable as you can. But you'll still probably have to experiment in the real world with values for C6 and R5, to make the best of a difficult situation.
|
|
|
Post by roadtonever on Mar 5, 2012 19:00:37 GMT -5
I'm with ashcatlt, 10KHz resonances aren't that useful IME. Classic bell tones have a resonance between 4-5 KHz, around 6KHz sounds pretty nice for strumming rhythm tones. Many guitar speakers drop off around 5KHz in fact.
May I suggest having trying a mid cut around 1.2Khz? I've found it retains the punch needed for live situations. Actually a cut at 700Hz sounds nice, especially if followed with a 4-5kHz peak, just difficult to cut through with. I play through a Sansamp directly to board so YMMV.
|
|
|
Post by newey on Mar 5, 2012 23:15:22 GMT -5
OK, Rt's suggestion of changing the pickup inductance to 2.5 microHenries (.0025H) yields this: Point taken, this is a nicer profile for the scoop, but of course not real world for the pickup inductance. sg- I read a lot about Q resonances and my head is swimming. I am paying attention, but this is well beyond my comprehension of the theory at this point. I don't rightly know how to translate that information into a useable set of parameters for this circuit. I guess I was assuming that, if'n the scoop didn't extend too high, the higher-order stuff would still be there just as if there was no mid cut at all.
|
|
|
Post by reTrEaD on Mar 6, 2012 0:29:38 GMT -5
I am paying attention, but this is well beyond my comprehension of the theory at this point. I don't rightly know how to translate that information into a useable set of parameters for this circuit. I don't either. If we were dealing with something less complex, like a filter between stages of an amplifier, it might be worth wading through that. But when you're dealing with additional inductances and capacitances in the source that will skew the results, I think it would get more than hairy trying to set up appropriate equations. What make matters worse, this is a generic model for one SC pickup. Assuming you would be using a single coil pickup and only one at a time, this still doesn't necessarily tell us how the circuit is going to act. Your pickup might have substantially more or less inductance than this. And if you use two SCs in parallel, that would mean half the inductance. And two in series would be twice the inductance. So this thing is likely to vary substantially from what is being modeled. You might be better off to set up your guitar as a testbed for experimentation. Pick a value for the inductor and install a rotary switch for the caps ala varitone and audition them to find which one sounds best. idk. *shrug*
|
|
|
Post by sumgai on Mar 6, 2012 4:16:22 GMT -5
You might be better off to set up your guitar as a testbed for experimentation. Pick a value for the inductor and install a rotary switch for the caps ala varitone and audition them to find which one sounds best. Hands down, the best answer so far. Look, all this circuit theory and SPICEing is fine, and a really cool way to waste some time, but when the chips are down, the buffalo is empty. Meaning, a computer program just isn't gonna be able to tell you what the combination of components is actually gonna sound like. Besides all the other varying factors, there's your own guitar, with it's pickups and strings, there's your amp, and maybe a cable (if'n you ain't usin' an RF jobbie), and the final arbiter of all things audible, your ears. Not a single one of those things can be quantified into a program in a meaningful way, sorry to say. So, I say reTrEaD's right - bust out the finger-burning device, and raise some smoke. Who cares what the PhD's and the EE's say - they ain't bringin' home your bacon, are they? Now hop to it! ;D And post sound files! sumgai
|
|
|
Post by JohnH on Mar 6, 2012 14:33:34 GMT -5
A strongly held belief of mine:
The truth is that if you want to work out something a bit complex, you need to have theory, then prototype, test, analyse and evaluate, then back to theory again etc. This thread is a good case for showing what you can work out by theory, and if it is decided to move to testing, the theoretical results will inform the range of design values. Ive found this process to be the key to several things that I have played with here, such as treble bleed, piezo sytems and cab design.
Its a nice change from my day job of designing building structures, where the design is all theory up untill construction drawings, and I cant usually build them first to see if they stand up!
John
|
|
|
Post by sumgai on Mar 6, 2012 22:19:40 GMT -5
Well, John.....
What you say is true, but I'm now coming at it from a slightly different perspective.
We've done all the necessary theory, three pages worth. And in the end, we didn't go all-out gang-busters to teach/learn theory, we used common sense based on many years of experiece, spread across both our own membership, and other references around the web. That's why reTrEaD and I say, enough head-scratching!, let's bread-board it, and see what our ears think! Then we can go back and examine our results in a more meaningful light (i.e. closer experience). Theory won't be worth any less, but it won't make-or-break our results, either.
HTH
sumgai
|
|