Fixing the thin tone of the Ibanez JP20 - or any other guitar

[Veit]

I own a vintage Ibanez JP20 from the the 80s. Like so many others I also never really liked the sound, but I also just did not want to give up on it, because I liked the look and the playability. After a long, long battle I finally found an explanation and a solution for the thin sound. Here is, what I did, essentially applying some proven engineering methods.

I also have a Telecaster clone which has a rather warm sound, so I used this guitar as a reference. With a signal generator and a digital oscilloscope I then measured the impedance curve of the pickups of both guitars. By comparing the curves I found that the JP20 has a pronounced resonance peak at around 3.5kHz which was very poorly damped. It appears that the inductivity of the pickup together with the capacitance of the wiring and the cable produce a very unfortunate frequency response, that boosts the frequencies leads to the nasal, thin and somewhat piercy sound.

When I had that I could immediately design a very simple compensating filter to load the pickup and reshape the frequency response, the filter is essentially a resistor and a capacitor in parallel. This pulls the resonance peak down to about 2kHz and flattens the frequency response curve, thus removing the nasty frequencies and evening out the spectrum. For a permanent solution you can easily solder these two parts into the plug of a guitar cable. To be more flexible I ended up building a dedicated JP20-Fixer pedal with a simple circuit in it. The input and output jack of the pedal are directly connected through. Between the two leads I have a 500k potentiometer which allows me to dial in a resistive load to dampen the resonance peak. The pedal then has four capacitors of 470pF, 1nF, 2.2nF and 4.7nF between the two wires, which I can switch on and off seperately. If I switch on 470pF and 2.2nF I can have 2.67nF etc.. Now I can modify the tone within a certain range and compensate for different cables. You first keep the potentiometer to max and play around with different combinations of capacitors, until the tone sounds about right. You then try different settings of the potentiometer to flatten the frequency response even more, until you are happy.

With this fix the guitar finally became a proper jazz guitar for me. My conclusion so far is, that there is not much wrong with the build of the guitar and I love the playability. But the electric wiring is not so well designed. I am now very much convinced that the influence of the electric parameters of the pickup is very much underrated. I see lots of threads about guitar cables, changing pickups, humbuckers vs. single coils etc.. What I see is that the combination of the very basic parameters (inductivity, capacitance and resistance) shape the tone so much and can be fixed so easily, that I suggest to play around with simple compensating filters first, before spending a lot of money and going through the hazzle of rebuilding half the guitar.

In case you are interested: Based on my measurements the JP20 pickup circuit can be modeled as an inductivity of 3.9H and a resistance of 7kOhms. The parasitic capacity is 500pF and the volume potentiometer is 120kOhm. With those parameters you can create an equivalent circuit to model the guitar and find tweaks.

[GuyBoden]

"For a permanent solution you can easily solder these two parts into the plug of a guitar cable."

Good stuff, I use a 8nf cap, but what value resistor an capacitor do suggest use?

[frabarmus]

I've had a JP20 back in the eighties and loved "the object" and thin neck... not so much the thin sound, 25.1/2" scale length and pickup placement, almost central (distant from the neck)... didn't that, also, contribute to the thin sound (I don't know much about the electrical wiring etc. I'm not an expert in that)? Also, as I'm mostly a plectrum user, my pick often stumbled on the pickup and fell on the ground (rather annoying).

[Veit]

"For a permanent solution you can easily solder these two parts into the plug of a guitar cable."

Good stuff, I use a 8nf cap, but what value resistor an capacitor do suggest use?

Well, there is a theoretical answer and a practical one. The problem is that your starting point is not well defined and you can also not really physically describe the desired outcome, which means that you would have to describe the 'optimum' sound as a desired spectrum. You have a very particular guitar with its tone and volume knobs, a specific cable with its capacitance, an amp with an input impedance etc. Theoretically you can measure all these details and then calculate the best values for the resistor and the capacitor to match your desired frequency response. That is a lot of work with questionable outcome. Standard values for the resistor will most likely not fit your case.
My recommendation is to set up a very simple test rig and play with it - which is actually a lot of fun and you will learn a lot. Sacrifice one of your guitar cables and cut it near the plug. You can then use a simple strip board on which you can change the two parts easily and try them out. I ended up building a couple of little circuits like the one that you see on the images: With the dip switches I can combine the three capacitors and the 500k Potentiometer allows me to play with the resistor load. When you finally know which setting you need you can make yourself the final permanent cable. Just measure the value that you dialed in on the pot and use that value for the resistor.

[John A.]

I own a vintage Ibanez JP20 from the the 80s. Like so many others I also never really liked the sound, but I also just did not want to give up on it, because I liked the look and the playability. After a long, long battle I finally found an explanation and a solution for the thin sound. Here is, what I did, essentially applying some proven engineering methods.

I also have a Telecaster clone which has a rather warm sound, so I used this guitar as a reference. With a signal generator and a digital oscilloscope I then measured the impedance curve of the pickups of both guitars. By comparing the curves I found that the JP20 has a pronounced resonance peak at around 3.5kHz which was very poorly damped. It appears that the inductivity of the pickup together with the capacitance of the wiring and the cable produce a very unfortunate frequency response, that boosts the frequencies leads to the nasal, thin and somewhat piercy sound.

When I had that I could immediately design a very simple compensating filter to load the pickup and reshape the frequency response, the filter is essentially a resistor and a capacitor in parallel. This pulls the resonance peak down to about 2kHz and flattens the frequency response curve, thus removing the nasty frequencies and evening out the spectrum. For a permanent solution you can easily solder these two parts into the plug of a guitar cable. To be more flexible I ended up building a dedicated JP20-Fixer pedal with a simple circuit in it. The input and output jack of the pedal are directly connected through. Between the two leads I have a 500k potentiometer which allows me to dial in a resistive load to dampen the resonance peak. The pedal then has four capacitors of 470pF, 1nF, 2.2nF and 4.7nF between the two wires, which I can switch on and off seperately. If I switch on 470pF and 2.2nF I can have 2.67nF etc.. Now I can modify the tone within a certain range and compensate for different cables. You first keep the potentiometer to max and play around with different combinations of capacitors, until the tone sounds about right. You then try different settings of the potentiometer to flatten the frequency response even more, until you are happy.

With this fix the guitar finally became a proper jazz guitar for me. My conclusion so far is, that there is not much wrong with the build of the guitar and I love the playability. But the electric wiring is not so well designed. I am now very much convinced that the influence of the electric parameters of the pickup is very much underrated. I see lots of threads about guitar cables, changing pickups, humbuckers vs. single coils etc.. What I see is that the combination of the very basic parameters (inductivity, capacitance and resistance) shape the tone so much and can be fixed so easily, that I suggest to play around with simple compensating filters first, before spending a lot of money and going through the hazzle of rebuilding half the guitar.

In case you are interested: Based on my measurements the JP20 pickup circuit can be modeled as an inductivity of 3.9H and a resistance of 7kOhms. The parasitic capacity is 500pF and the volume potentiometer is 120kOhm. With those parameters you can create an equivalent circuit to model the guitar and find tweaks.

The most common criticism of the JP-20 is the pickup position -- it's further back toward the bridge than the typical neck pickup, and is therefore sampling a portion of the strings that has different overtones from the classic "jazz tone". It sounds more like the middle pickup on a 3-pickup guitar than a neck pickup. Listening to, for example, Joe Pass playing the JP20 vs his other electric guitars, the JP20 sounds quite starkly different. If you're saying that you can somehow "cure" that problem via capacitors I'm skeptical, but I'm certainly open to the possibility. Recordings would be the best way to demonstrate the effect you say your circuit has.

[Question]

I once saw this solution years ago, but didn't hear the results.

[Cunamara]

The most common criticism of the JP-20 is the pickup position -- it's further back toward the bridge than the typical neck pickup, and is therefore sampling a portion of the strings that has different overtones from the classic "jazz tone". It sounds more like the middle pickup on a 3-pickup guitar than a neck pickup. Listening to, for example, Joe Pass playing the JP20 vs his other electric guitars, the JP20 sounds quite starkly different. If you're saying that you can somehow "cure" that problem via capacitors I'm skeptical, but I'm certainly open to the possibility. Recordings would be the best way to demonstrate the effect you say your circuit has.

I can't speak for the OP, "curing" might be an excessive adjective. But if there is indeed a spike at 3.5 kHz, that's going to produce a noticeable difference versus one with the resonant peak around 2 kHz. Using capacitors and resistors to change the loading on the pickup and shift the resonant peak is not a new idea. It stands to reason that the frequency response properties of the pickup and electronics on the guitar have an impact on tone as well as the placement of the pickup along the string. Dealing with the electronics is easier than moving the pickup- you might not make an L5 out of a JP20, but you might make it a lot better to your ears.

There are extensive discussions about this sort of thing at TGP and other places. Somewhere there is an thread in reference to the Bill Lawrence/Wilde Micro Coil pickups and utilizing various capacitors to affect the resonant peak- on a Fender discussion board, maybe- that goes into this idea in some detail. I would have to have a much better understanding of electrical engineering (hmmm, maybe any actual understanding of EE...) to get all of it, however.

A few years back somebody posted about putting in a Benedetto floater between the stock pickup and the end of the fingerboard, attached to the pickguard; it sounded great, as I recall. He could use either pick up or both together, as desired

[Veit]

Maybe I should have worded my post a bit more carefully and avoided the term "fixing". You cannot fix something that is not broken and of course you cannot turn a JP20 into a L5. The guitar is what it is and even with the loading circuit it needs some additional help with a delay pedal etc.. What I tried to accomplish was to take some of the piercy overtones out and get to a smoother base sound which is a better starting point for the further steps.

Let me show you a few more details of what I did. As I described in my initial post I used my TC with flatwound strings as a role model, because it has a rather warm sound. In fact, that was the bit that really annoyed me and got me started, because I did not want to accept the fact that a solid body sounded better than an archtop that was designed for Jazz and even played by my hero Joe Pass. So, I used a signal generator with some circuitry to measure the impedance curves of both instruments, including the cables. From these curves I calculated the model parameters (inductivity, capacitance and resistance) and calculated the frequency response of this model for comparison. As you can see on the first graph the JP20 has a very pronounced peak at the wrong frequencies, while the TC is much smoother.

Using the model parameters I could derive the transfer function on the second graph, which shows how the oscillation of the string is transformed to the electric signal that is fed into the amp. You can see that the TC has a rather flat characteristic, which means that the oscillations of the string are pretty much fed to the amp as they are. The JP20 boosts the spectrum around 3kHz, which is bad. I then took the difference between the curves to compute the necessary corrections that you can dial in at an equalizer to shift the spectrum of the JP20 to the spectrum of the TC - see the third graph. This curve indicated that I should pull the nasty 3.2kHz tones down by almost 5dB while I should help the treble side. I tried that, using an equalizer pedal and found, that it did indeed help to make the sound smoother. The next step was then to build that adapter circuit and play with it, looking for futher improvements empirically. I think it really made a difference.

As for the pickup position: I know this theory and it sounds plausible, because the overtone distribution changes substantially when you change the pickup position. However, I saw that some people had tried to move the pickup further up the neck and found that it improved the situation somewhat, but not as much as they had hoped. This explanation seems to fall short and does not explain the whole situation.

My conclusion was, that we essentially have to accept the guitar as it is. However, the tone can be improved within a certain range if you add some electronic means. What I learned is that the very basic circuitry with and around the pickup form an electronic filter that shapes the sound substantially. The good thing is tat this frequency response can be modified so easily that this should always be the first focus point before deciding on changing pickups or cables. A long or expensive cable for example works by changing the capacitive load on the pickup. An additional capacitor for a few cents does exactly the same thing, but you have control over the value!

[GuyBoden]

Good stuff, similar to you, many years ago, after a lot of experimentation to remove the high frequency (Brightness) using 5 to 10nf capacitors between live and earth. I like 8nf in most archtop guitars, but there's more Bass.

I still try different values between 5-10nf capacitors, listen carefully and observe the change in sound.

It's an easy experiment using electrical connectors to try various capacitor values.

[John A.]

Maybe I should have worded my post a bit more carefully and avoided the term "fixing". You cannot fix something that is not broken and of course you cannot turn a JP20 into a L5. The guitar is what it is and even with the loading circuit it needs some additional help with a delay pedal etc.. What I tried to accomplish was to take some of the piercy overtones out and get to a smoother base sound which is a better starting point for the further steps.

Let me show you a few more details of what I did. As I described in my initial post I used my TC with flatwound strings as a role model, because it has a rather warm sound. In fact, that was the bit that really annoyed me and got me started, because I did not want to accept the fact that a solid body sounded better than an archtop that was designed for Jazz and even played by my hero Joe Pass. So, I used a signal generator with some circuitry to measure the impedance curves of both instruments, including the cables. From these curves I calculated the model parameters (inductivity, capacitance and resistance) and calculated the frequency response of this model for comparison. As you can see on the first graph the JP20 has a very pronounced peak at the wrong frequencies, while the TC is much smoother.

Googling around for "PAF pickup frequency response" I stumbled on this site:

https://www.echoesofmars.com/pickup_...ot~(~)~filter~')

It looks like a resonant peak at 3k is within the realm of normal for a humbucker. EQ-ing that down might result in a sound you prefer, but I don't think that resonant peak explains much about why the JP20 sounds the way it does. Also, given how common it is for people play jazz on teles, it shouldn't really be surprising that yours has a pleasing jazz tone.

Using the model parameters I could derive the transfer function on the second graph, which shows how the oscillation of the string is transformed to the electric signal that is fed into the amp. You can see that the TC has a rather flat characteristic, which means that the oscillations of the string are pretty much fed to the amp as they are. The JP20 boosts the spectrum around 3kHz, which is bad.

But the sound of the guitar string itself does not have a lot of content that high in the frequency spectrum, so the resonant peak's effect might not be as significant as you think.

I then took the difference between the curves to compute the necessary corrections that you can dial in at an equalizer to shift the spectrum of the JP20 to the spectrum of the TC - see the third graph. This curve indicated that I should pull the nasty 3.2kHz tones down by almost 5dB while I should help the treble side. I tried that, using an equalizer pedal and found, that it did indeed help to make the sound smoother. The next step was then to build that adapter circuit and play with it, looking for futher improvements empirically. I think it really made a difference.

If you like the effect, that's great.

As for the pickup position: I know this theory and it sounds plausible, because the overtone distribution changes substantially when you change the pickup position. However, I saw that some people had tried to move the pickup further up the neck and found that it improved the situation somewhat, but not as much as they had hoped. This explanation seems to fall short and does not explain the whole situation.

You can easily tell that the theory is real (and not simply plausible) by playing a guitar with more than one pickup, or by comparing the neck pickup on 22-fret guitar to an one an otherwise similar 24-fret guitar (something I have done directly). Pickups placed at different points along the strings' length sound different, and this difference is far more noticeable than differences in tone control circuitry.

My conclusion was, that we essentially have to accept the guitar as it is.

Yes, I agree.

However, the tone can be improved within a certain range if you add some electronic means. What I learned is that the very basic circuitry with and around the pickup form an electronic filter that shapes the sound substantially. The good thing is tat this frequency response can be modified so easily that this should always be the first focus point before deciding on changing pickups or cables. A long or expensive cable for example works by changing the capacitive load on the pickup. An additional capacitor for a few cents does exactly the same thing, but you have control over the value!

I don't find it easier to rewire a guitar than to try a different guitar cable or build an eq device, but to each his own on that front. I also question whether the sorts of changes people look for when swapping pickups can be achieved via eq or changes in guitar wiring. But again, if this experimentation is informative and useful to you, that's great.

[Veit]

Hi John, thank you for your elaborate response! I really appreciate that and enjoy this discussion on these ideas! So, let me respond in turn ...

The relation between the spectrum and the perceived sound has been studied extensively and sound engineers use these insights intuitively when they master a new record. When you browse the internet or do some AI-research you find many tables like this one:

Frequency Range......Tonal Character........Action for JP20
100–250 Hz...........Boom / Mud.............Keep neutral; too much makes the archtop ""woof."""
300–500 Hz...........Warmth / Body..........Boost slightly for a ""Jazz Hero"" sound."
800–1.5 kHz..........Honk / Nasal...........Cut (This is what the Mid knob usually does).
3–5 kHz..............Piercing / Clang.......Cut heavily (add capacitance).

So, taking such a table as a starting point gives you some ideas what to shoot for. If the JP20 has a rather piercing sound and can also be nasal then you see which areas you should probably attenuate or boost. It appears that the resonance peak of the JP20 is exactly where you do not want to have one.

As for the parameters I tried to get hold of the datasheet for the pickup, but skipped that idea quickly. Even if I had found the data, I still would not know how the capacitance of all the wiring or the parasitic resistors change the whole game. So I found it easier and more reliable to just measure the whole thing. Fortunately I have the means to do that - see below.

The little excercise with the equalizer pedal was only for a plausibility check, I needed to see if I was heading down a promising path. When I saw (or heard) that manipulating the problematic zones did effect the perceived character I knew that this approach might get me somewhere.

As for the pickup position: A while ago I created an Excel spreadheet to calculate the gain of the first 10 overtones of a string as a function of the pickup position, which can be done easily because the oscillations are simple standing waves. I could clearly see that the amplitude distribution of these overtones changes significantly with the pickup position, but I found it impossible, to interpret the results. For one pickup positon you might see that the 3rd, 5th and 7th overtone had a higher amplitude when you played at the fifth fret while they were lower when playing at the tenth fret. For another pickup position it was the other way around. When I got myself totally confused I gave up on this and simplified my life by looking at the whole spectrum as a continuous distribution.

You mentioned that you find it easier to simply swap a cable, rather than adding a small circuit (which is really only a matter of adding a resistor and a capacitor to the cable plug, which takes about 15 minutes). The problem is not that little technical modification, but to find out, what component values you need. In other words, how do you know which cable you should purchase? By buying a bunch and trying them out, hoping for the best? Or asking your buddies? From an engineering perspective you don't need to guess: If the tone is piercy, you know that you need more capacitance. So, you either use a longer cable (capacitance is proportional to the length) or chose one with a larger coefficient for the insulator capacitance.

Just so that you know where I stand and where I am coming from: I do love jazz and jazz guitar in particular, in all aspects. For me understanding the chord changes of Giant Steps is as much fun and part of this hobby as is the never ending hunt for the best gear or tone. Now, in another parallel universe I happen to have a PhD in electronics engineering, so I have a well-stocked toolbox of engineering methods at hand and I also have a well-equipped electronics workshop. With these options analysing the simple pickup circuits is really trivial, after all it is nothing more but just a handful of capacitors, coils and resistors. So, instead of guessing or relying on trial-and-error approaches I can take shortcuts. (To be honest: For an engineer it can sometimes be rather frustrating to read the discussions in dedicated guitar forums, there is A LOT of snake oil around...)

Again, thanks for your time and your comments!