Treble bleed comparison - frequency response curves

Discussion in 'Tele-Technical' started by BobUrban, Jul 15, 2013.

  1. BobUrban

    BobUrban Tele-Meister

    Posts:
    367
    Joined:
    Jun 23, 2009
    Location:
    Cleveland, OH
    I've been toying with circuit simulation using LTSpice, and I simulated a guitar circuit to compare different treble bleed circuits. This simulation includes the pickup (this case, a Telecaster bridge pickup), two 500K pots (volume and tone), the treble bleed circuit, the guitar cable, and the impedance of the amplifier.

    The attached figure compares the frequency response of the following treble bleed circuits with the guitar at "half volume", which in this case would be at 5 on a linear volume pot, or about 8 on a log pot.

    1) No treble bleed
    2) 0.001 uF capacitor
    3) 0.001 uF capacitor in parallel with a 220K resistor
    4) 0.001 uF capacitor in series with a 130K resistor (the "Kinman Mod")
    5) Input to the tone circuit is taken off the wiper of the volume pot rather than the input lug (the "Fezz Parka" or "50s wiring" mod)

    As you can see, all are effective in boosting treble frequencies. This figure supports what people say about the Fezz Parka mod, which is that it is the least "coloring" of the treble bleed options. There is a decrease in bass frequencies, but a clear boost in treble. A simple and effective treble bleed.

    The standard 0.001 uF treble bleed greatly boosts treble frequencies. This supports the conclusion that this treble bleed is very bright. However, I've also heard people say that it decreases bass, but that is not shown here. It might be perceived as less bassy since there is so much more treble.

    The 0.001 uF in often placed in parallel with a 220K resistor to also boost bass frequencies. This is clearly the case, but you can see that it really "colors" the unmodified frequency response.

    The "Kinman Mod" is something between the two extremes. It is a more tame version of a lone 0.001 uF cap. I personally like this treble bleed mod because I feel like it gives a brighter tone without sounding unnatural.

    Anyways, I thought this might be of interest to some of you techies. Keep in mind that frequency response is a "steady state" measurement, meaning the transient or dynamic response isn't captured. This is harder to measure and compare but equally as important, as it defines the character of sound.

    [​IMG]
     
    ftbtx likes this.
  2. BobUrban

    BobUrban Tele-Meister

    Posts:
    367
    Joined:
    Jun 23, 2009
    Location:
    Cleveland, OH
    Here is the same figure, but with a curve corresponding to the circuit at full volume.

    [​IMG]
     
  3. notyalcer

    notyalcer Tele-Meister

    Posts:
    132
    Joined:
    Nov 6, 2005
    Thanks for posting this great information. I also prefer the Kinman treble bleed over the other methods you've examined here. Just feels and sounds the best to my ears.
     
  4. Brooks A Hood

    Brooks A Hood Tele-Meister

    Posts:
    148
    Joined:
    Jan 26, 2011
    Location:
    Reykjavik, Iceland
    Hey Bob - Thanks for doing the work and posting this. Can I ask a few questions ? Why the 500k pots vs. 250k that most people have in their Tele´s ? And am I not seeing this correctly that the curve most closely resembling the full volume curve is the parallel 220k/.001μF ? The Fezz curve is lower between 2-4khz as well as the peak looking to be shifted from ≈4.2k to ≈5.7k ? Is there something I am missing ?? Also, how much capicitance was factored in from the cable ?

    I like the parallel style because it has the effect I want on the slope/taper of the pot. (I am not a fan of 10% audio taper pots) I usually do not use the same resistor value you did to achieve what my ears tell me sounds best.
     
  5. BobUrban

    BobUrban Tele-Meister

    Posts:
    367
    Joined:
    Jun 23, 2009
    Location:
    Cleveland, OH
    I used 500K pots because that is what I like to use in my guitars. The general shapes should stay the same. I also used a 0.022 uF tone cap, which I forgot to mention. If I get some time, I'll re-simulate using the traditional 250K/0.047 uF Fender combination.

    I assumed a 30 pF/foot capacitance for the guitar cable, and assumed a 6 foot cable.

    You are right, the parallel combination most resembles the full volume curve in terms of shape. But it's the most different from the non-treble bleed at half-volume (not in terms of shape, but in terms of dB over the frequency range).

    There is a cool Excel macro that lets you play around with all sorts of different values to see the effect on frequency response.

    http://guitarnuts2.proboards.com/thread/3627
     
  6. jamieorc

    jamieorc Tele-Holic

    Posts:
    733
    Joined:
    Apr 22, 2011
    Location:
    Virginia
  7. GCKelloch

    GCKelloch Tele-Afflicted

    Posts:
    1,240
    Joined:
    Apr 10, 2011
    Location:
    MA USA
    The cap value recommended is based on the circuit capacitance, and the resistor value: on the pot value, being ~1/3. The pickup impedance and/or inductance may be a factor, but I'm not sure. I've used the BL treble bleed on a few low/mid inductance pickup guitars with a ~200pF cable. It sounds very natural. 0.001uF/1nF/1000pF is a pretty high cap value, but it might be good with a similar C value cable. High C cables just limit treble response, and can end up placing a strong peak in harsh frequency range. Different pickups will therefore have a different response. Wilde pickups don't generally need a treble bleed, because they have very low internal capacitance, and comparatively low impedance to other Hi-Z pickups. I use one on my Wilde NF series pickups, but rarely use the volume anyway. I generally like the difference in sound when my audio taper pot is down at ~5. It's bright, but sweet. Seems like what is expected by a guitar player at half volume.


    The treble does increase as the volume knob is lowered with the 50's wiring, so the tone knob has an unusual response. Some people like that. I suppose it depends on where in the spectrum that treble peak is located, which depends on the pickups and the cabling.
     
  8. Ricky D.

    Ricky D. Doctor of Teleocity

    Age:
    70
    Posts:
    11,339
    Joined:
    Oct 22, 2006
    Location:
    Marion, VA
    Thanks, BobUrban!
     
  9. murrayatuptown

    murrayatuptown TDPRI Member

    Posts:
    12
    Joined:
    Mar 10, 2009
    Location:
    Holland Michigan
    Hi:

    Someday I'll try simulating this myself, but there would be several steps and opportunities for error, and I have a variation question, so I'll tack it on here.

    I saw someone's online evaluation of a 'common' treble bleed mod with test clips for quick evaluation. IIRC , the component values were something Stew-Mac sells.

    The author was mostly happy with results on a Strat & Les Paul, even with their differences. He found it to be unusable on an (old?) Tele with a 1 M pot. The other guitars had 250k & 350k, IIRC (and I may not).

    I had a hare-brained idea, and came up with an alternate series RC pair, despite not knowing enough to run a simulation. I think I rationalized scaling the resistor and capacitor, increasing the resistor proportionally to the LARGER pot, or maybe the average of the two pot values, then scaled the capacitor down so the end result had the same corner frequency or time constant...really a haphazard rationalization, considering I didn't have a 'model'.

    I made up three experimental series RC sets with mini-'gator clips, same resistor for all three, a single 'new' C value for the first, two of that value paralleled, and three, and sent them to the author, unsolicited, no obligation, of course.

    He did receive them & thanked me, but had lost interest and quite reasonably had put his guitars back together. Can't blame him. Someday there may be a follow-up discussion...but who knows? He didn't ask me to do it and pulling something apart again is a nuisance...although a Tele is easier...I completely understand not wanting to get inside a Strat again until changing strings!

    I recently got curious whether simulation would answer my curiosity of whether I solved a problem, or wasted TWO people's time with an irrational half- (or quarter?) assed solution.

    I haven't found my spreadsheet to review my rationale, but I came up with 820 pF and 391k (series) for the 'scaled' treble bleed for Tele with 1 M volune pot, and (1640 pF + 391k) and (2560 pF + 391k) for the 'band-aid' alternates.

    I had a box of 820 pF polystyrene capacitors since probably the late 70's and several 391k resistors left over from who knows what project, so that's why I forged ahead and made up some Franken-tone bleeders...I conveniently had the parts & alligator clips. Got no Tele, so I thought I had a solution but no problem (well, obviously I have others, like time management!).

    I'm curious if any of these pairs are useful info to pass on (or warn others to stay away from), and if someone with simulation skills might humor me...or help me scramble up the learning curve a bit.

    I won't ask why there was a 1M volume version Tele...maybe an old one...,but I didn't ask. Uou didn't see that.

    Thanks

    ...anyone else hate typing on a 'smart' phone? I wouldn't call it that...
     
  10. The Ballzz

    The Ballzz Tele-Afflicted

    Age:
    66
    Posts:
    1,976
    Joined:
    Apr 11, 2016
    Location:
    Las Vegas, NV
  11. murrayatuptown

    murrayatuptown TDPRI Member

    Posts:
    12
    Joined:
    Mar 10, 2009
    Location:
    Holland Michigan
    I just looked back & saw that 1/3-pot rule of thumb, so for a 1M pot, my 391k was in the ball-park, perhaps by luck, as I didn't have that wisdom when I came up with my values (alchemy, perhaps?).

    Ah well, short question...if anyone is willing to share simulation info for a Tele, I'd appreciate it.

    Thanks

    Murray
     
IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.