For the nerds - treble bleed models in LTSpice: Duncan, Kinman, no-resistor

After auditioning half a dozen treble bleed candidates, I thought I'd work smarter not harder on getting what I wanted from it.

So I learned how to model a guitar circuit in LTSpice and figured out what the different parameters do to the frequency response of the RC network.

Now I've got graphs to share, showing what changing the values of the capacitor, the parallel resistor, and the series resistor does.

This definitely helped me zero in on what to change as I used my ears to test each RC network in my guitar.

It's kind of a generic model so I won't get into recommending any specific values. Every guitar is different, every player's pair of ears is different. But this guided me in landing on a treble bleed I'm super pleased with. I definitely heard a LOT of damn stinkers during this process and can see why people who don't methodically figure out what works in their guitar wind up just deciding that treble bleeds suck and they hate them.

Red represents full volume - all the options overlap here because the treble bleed is not in effect at full volume. Blue is volume turned down about 2/3 of the way (with an audio-taper volume pot), and green represents volume turned down to the lowest useful setting - almost-off but without losing legit, listenable signal.

First, here's what changing capacitor values does: Larger values shift the resonant peak to lower frequencies and make the peak higher. The next two images after this include a single capacitor value, chosen to line up with the full-volume resonant peak.



Next, here's what changing (Duncan-style) parallel-resistor values does: Smaller values result in less lows being blocked from the bleed, so it turns the treble bleed into a bit of a treble-and-everything-else bleed. It has the effect of bringing all the frequencies closer to the level of the resonant peak - as well as lowering the peak, both of which mellow the peak treble intensity but you can see how this makes the volume taper unpredictable:



Finally, here's what changing (Kinman-style) series-resistor values does: Larger values block signal from being bled off, reducing the height of the resonant peak and widening the remaining treble peak's band. This also mellows the peak treble intensity in away that is slightly different (rounder, softer) than the parallel capacitor does, and in a way which keeps the volume taper sounding pretty normal.



Stand by for graphs of commonly-recommended treble bleeds, as well as the one I settled on after auditioning about 10 more after the first half-dozen random choices.

 

Good point!

The cable capacitance will move the natural resonant peak to a different frequency, so the treble bleed's peak won't line up with it anymore.

I'm not 100% sure what the effect of impedance would be, but I'm guessing that it wouldn't do anything but affect signal volume - after it leaves the guitar. Not altering the relationship between the guitar's peak and the treble bleed's peak. Educate me if there's more to it than that.

 

So if you have looked at that Premier Guitar article on treble bleeds which I linked above, that's where I got the "recommended values" from. That, and also Fender's own page about their Tonesaver.

Here they are. Same three volume levels as described in my first post above. Blue is the treble bleed, green is without the treble bleed. Of course they overlap at full volume so you can't see the green.

Plain cap-only (1200pF) treble bleed. Super intense! To the eyes and the ears too. Anyone's ears, any guitar, I dare say.



Duncan style - 1000pF cap, 150K parallel resistor. Also pretty intense but less sharp and high of a peak. Plus that volume-taper problem is obvious here.



Kinman style - 1200pF cap, 130K series resistor. Bleeds a wide, low band of treble. You'd think the absence of a sharp, high peak would sound pretty different from the un-attenuated, full-volume tone but the difference isn't as glaring as the image suggests. Also note the volume-taper is normal again, no surprises from the pot.



Fender Tonesaver - 1200pF cap, 150K parallel resistor, 20K series resistor. They decided to blend Duncan and Kinman (series and parallel) resistors in the same network but this still sounds a bit too treble-intense at low volumes.



Last, here's my own treble bleed, after testing about 15 candidates. 680pF cap, 220K parallel resistor, 100K series resistor. A subtler, mellower version of the Tonesaver. A more narrow-banded, slightly peakier version of the Kinman above.



In this model (it's only a model), the treble bleed’s resonant peak lines up perfectly with that of the model's hypothetical pickup at full-volume. In my actual guitar, the cap I chose splits the difference between the resonance peaks of my two pickups (not that different), my middle position (not that different), and my phase-reverse switch (quite different, but least important as far as volume rolloff/tone integrity goes).

The band is wide enough that none of those pickup selections sound "off" and the bleed is quiet enough that low-volume tone is mostly preserved. It's a tiny bit more sparkly than full-volume, but not thin, and to me it just highlights that the full-volume sounds just a little fatter by comparison - mostly in the midrange, which is where I want it when on loud. It doesn't get to that "Oh damn that treble bleed!" that people hate on low volume.

I had the tone-on-volume-output mod (50's wiring, Fezz Parka mod) and I liked it for a while, but it made the tone control less useful. It wouldn't turn all the way down. I do like me some jazz tones from the neck pickup, so that wasn't working. It also caused the tone at full volume to be too midrangey. It's like I couldn't turn the tone all the way up either, at least at full volume.

With the treble bleed, the tone controls work great (I have both treble and bass controls). The difference in tone between full-volume and low-lower-very-low is noticeable when listening carefully, but it's really small. It's very satisfying to me and I'm glad I put in the effort to land on a decent config for this guitar and my rig.

 

Yeah, having this information helped me get closer with each iteration. Before I made the model and saw what each parameter affects, my changes each time I heard a treble bleed I didn’t like and tried to make a better one were pretty random and wouldn’t have gotten me a satisfying result.

Here’s hoping it helps others.

 

Unrelated tangent, I got 14 slightly used treble bleeds for sale! CHEAP.

Seriously though, for $10 I got way plenty of assorted resistors and capacitors from Amazon, more than more than enough to experiment and listen till the done indicator pops.

 

And visualize what we want to hear, amirite

 

Right, that's the whole thing. Every guitar's different, every rig's different, every player's pair of ears is different. Everyone's taste is for sure different.

That's why experimentation and iteration are the key.

Enjoy.

 

Oh yeah? You used the "Improved" design? Looks like Beanluc/Tonesaver-type (both parallel and series resistance) but with variable resistance in the parallel spot?

How'd that work out?