BOSS DS-1: modding my own mod

In my previous post https://electric-safari.com/2019/03/17/turn-a-boss-ds-1-into-a-nice-overdrive/ I converted the DS-1 distortion pedal into something very close to an overdrive. One of the things I changed was the clipping section; one of the diodes was replaced by a red led. I tried some other combinations, but not very scientifically, honestly. The final one was chosen by ear, and that’s the one I published. But the doubt about some other combinations remained in my head after that.

Recently I got another DS-1 in ebay, and after modifying it (following my own instructions), I decided to give it another twist: let the user decide what diode to use (or no diode at all) for each side of the clipping section. In the next diagram I show how I connected two rotary selectors for choosing the diodes:

The rotary selectors are these ones: https://es.aliexpress.com/item/32947552565.html?spm=a2g0s.9042311.0.0.6e1663c0haFbmP

The actual devices with the diodes connected:

The X and Y terminals from the diagram above have to be connected to one of the removed diodes terminals in the PCB (see the white and blue cables):

It is not necessary to connect anything to the other diode because the two diodes are connected cathode to anode and anode to cathode. Don’t connect one of the terminals to ground, as you could do with some other hard clipping designs (ProCo Rat, for instance) because that is not how they are connected in the DS-1.

The result:

The two selectors act (as I have connected them for my own convenience) as two “taps” for the two sides of the signal (positive and negative), as seen from the supply jack side of the pedal. Open the tap and you have a more open (less clipped) signal, close the tap and you have a more clipped signal. The sequence of clipping goes like this:

  • wide open, left position: no diode for clipping. The clipping comes from the transistor section and the operational
  • second position from the left: blue led, Vf=2V
  • third position from the left: red led, Vf=1.5V
  • fourth position from the left: original diode: Vf=0.5V
  • fifth position from the left: schottky diode: Vf=0.2V

Seen from the front, as you play, you can think about left=less volts / right=more volts.

You could say: there are repeated combinations, because for instance “red led + schottky = schottky + red led”, so I could have chosen a 15 positions (I think) rotary switch for all the valid combinations. Apart from the fact that a 15 positions switch is hard to find and it would have complicated the cabling, that hypothesis (“red led + schottky = schottky + red led”) is not completely true in the real world, because of two reasons:

  • Let’s face the truth: two different diodes with the same name do not have exactly the same Vf, not to mention the rest of the specs.
  • Clipping the positive part of the signal is not the same as clipping the negative part of the signal, because of the non linearities of the circuit design and components.

I think that setting two selectors is (as least for me) more intuitive. This is mostly an experiment that can be used also for gigging.

How does it sound

Yes, an experiment, but also a gigging pedal I am planning to use. In spite of the flexibility given by the two selectors, it doesn’t lose its original character. Selecting the combination: position 2 + position 3 (original diode + red led) it sound almost like the other DS-1 I modified, as expected, given the intrinsic differences of every unit components.

Selecting different positions for each selector, especially if they are very extreme, accentuates the asymmetrical character of the distortion (even harmonics, more on that here: https://blackstoneappliances.com/dist101.html)

The blue led combined with any other diode gives a very mid rangy character for my ear.

As you “open” the “tap”, the signal is louder. Selecting the open – open combination creates, as expected, a very loud but distorted signal, not bad given the fact that it is created by the transistor and the operational. Cannot be used as a clean boost pedal, but can crank a tube amp. 

At the other extreme, selecting the two schottky diodes and maxing the distortion control creates an almost fuzzy distortion. If you max the three pedal controls, select a neck humbucker pickup and lower the guitar tone control, the fuzz is there for you to enjoy. That way all the treble harmonics that you hear come from the pedal itself, not from the pickup.

Please let any thoughts in the comments section, and if you try this mod please let me know.

The DS-1 mod I mention above (Turn a BOSS DS-1 into a nice overdrive) has been implemented at least by two friends of mine (thanks Pepe and Juanjo for trying it out), and they are very happy users of it, I think they are even gigging with it. They call it “warrior mod” because of my last name in spanish.

 

 

DS-1 vs modded DS-1: frequency response

Let’s compare the original BOSS DS-1 with my modified DS-1, so we check out the difference in tonal character between a distortion pedal and an overdrive pedal (or whatever DS-1 has become after being modified in https://electric-safari.com/2019/03/17/turn-a-boss-ds-1-into-a-nice-overdrive/).

I have tested them using a Linux computer, its sound card and qloud software. I know the measurement is not very precise, but the real thing is not going to vary very much, I think. All controls were at noon during the measurement.

Original BOSS DS-1

Original DS-1 frequency response

As expected from a pure distortion effect pedal, frequency response is mid scooped, with a bump at 100Hz.

Modified BOSS DS-1

Modified DS-1 frequency response

As I expected after the modifications in the tone stack, frequency response has much more mids. The maximum is not at 800Hz, just because the circuit is not only the tone stack. By touching the tone control, the curve can be modified (I leave it for a next article).

As you can see also, at aproximately the same “level” value, the overall volume is higher (over -60dB for a wider range in the test), without changing any measurement parameter, even if the perceived distortion is higher in the original unit. The reason, I guess, is the red led in the clipping section instead of a common diode, see the article https://electric-safari.com/2019/03/17/turn-a-boss-ds-1-into-a-nice-overdrive/. Since the Vf of the led is higher than that of the diode, it clips at a higher voltage, giving the circuit more headroom. This is very common when doing this kind of modifications. I think the tone stack change alone does not explain that higher volume.

Turn a BOSS DS-1 into a nice overdrive

Español

The fence that separates distortion and overdrive pedals is very thin. Usually pedals based on hard clipping, high amount of distortion and mid scooped response curve are called distortion and those ones based on soft clipping and mid boosted are called overdrive. But some of the more famous overdrive pedals, known for being “transparent”, “clean” or “dynamic” are based on hard clipping (clipping diodes at the output of an operational amplifier or transistor). Think of Klon Centaur, for example (mid boosted, by the way).

Hard clipping seems to have usually a more dynamic overall response to the guitar touch than soft clipping, although it depends on the circuit design. Some soft clipping pedals are very dynamic while others are too compressed, but most of them seem to be compressing and distorting all the time.

Hard clipping, theoretically limits drastically the signal when it reaches a certain level, so it creates those high harmonics that give it its character. Soft clipping limits the signal more gradually, giving that “soft”, characteristic valve-like distortion sound.

Having taken all these considerations into account, let’s review my last pedal modification. I wanted to experiment those concepts with a BOSS DS-1 pedal I purchased broken and then fixed. I liked as it was, but that kind of distortion is not for me, frankly.

BOSS DS-1 is based on classical symmetrical hard clipping, two diodes at the output of an operational, and the tone stack tends to mid scoop the response curve, so it should be considered a distortion pedal. It is orange and is called “Distortion”, so in this case it was easy.

My experiment had these objectives:

  • Reducing distortion (amount of clipping)
  • Give it an asymmetrical clipping character, more pleasant to my ears and more valve-like distortion
  • Mid boost the response curve
DS-1 original schematic

For reducing distortion, I have made these two changes:

  • R9 at the emmiter of Q2 controls the gain of the first stage, previous to the opamp stage. With a value of 22 Ohms, the output is so high that the signal at the output of the operational is almost always clipped. I tried with different values and for me 100 Ohms gave the sweet spot for having the right amount of distortion with the gain control at 12 o’clock, for both single coil and humbucker pickups.
  • LED diodes have a Vf higher than common bipolar diodes. Bipolar diodes have Vf=0.7V and LED have Vf > 1.5V, depending on the color. Vf is what causes the clipping, and the higher Vf is, the higher the signal can be before it starts to clip. Replacing only one of the clipping diodes (D5 for instance) by a red LED yields more headroom and at the same time serves my second purpose of having asymmetrical clipping. By the way, this is one of the changes that the famous Keeley Ultra Mod makes to the DS-1.

For the mid boost thing, I changed several components of the tone stack. DS-1 tone stack is based on the Big Muff tone stack. It is passive and tends to mid scoop the tone. By simulating the circuit with LTSpice, I found the spot where the signal is mid boosted without loosing the ability to control the tone (although it is somewhat reduced compared to the original DS-1 design).

Modified DS-1 tone stack

The bump happens at 800Hz, which for me is the best point to get a nice tone that is capable of cutting through the mix in a band context. It is a simplification because the response curve of the pedal depends on the rest of the circuit as well, but I can tell you that the result is quite acceptable.

The components replaced in the tone stack are C11, C12, R15 and R16. As you can se, only six components have been replaced. I have prefered to keep it as simple as posible, but the change in tone and behavior is drastical. Now the pedal is a very dynamic, mid boosted overdrive, with a wide range of distortion. With gain at max level, it reveals it true distortion spirit. Put a compressor before and an equalizer that boosts the 100Hz frequency after it, and you’ll get the real sound of rock and roll.

As a summary, these are the replacements:

componentoriginalmod
R922 Ohms100 Ohms
D51N4148Red LED
C1122nF33nF
C12100nF82nF
R152.2K3.3K
R166.8K1K

For C12, 82nF is not a very common value, but you can put 47nF and 33nF in parallel to get 80nF, enough given the usual tolerances of capacitors. Use a capacitor tester if possible to get a close value. The board has plenty of space to put the two capacitors in that position. For C11 and C12, any kind of capacitor will do, but the values are very important (test if possible).

I hope you enjoy this mod, please let me know of any improvements or mistakes you may find. If you are a good guitar player, please record a demo and upload it to youtube. If you are rich, please send me a Klon Centaur to compare with my DS-1 modified pedal.

PD: if you make this mod, please comment below, I would like to keep track of people enjoying (or not) my mods.

Convierte el BOSS DS-1 en un buen overdrive

La frontera entre pedales de overdrive y de distorsión es más bien fina. Normalmente los pedales basados en hard clipping, mucha distorsión y una curva de frecuencias “mid-scooped” (frecuencias medias rebajadas) son denominados “distorsión”, y aquellos basados en soft clipping y con resalte de medios son llamados “overdrive”. Pero algunos de los overdrives más famosos, conocidos por ser “transparentes”, “limpios” o dinámicos” están basados en hard clipping (cortar la señal en los extremos mediante diodos puestos a tierra a la salida de un operacional). Pensad en el Klon Centaur, por ejemplo (con los medios resaltados, por cierto).

El hard clipping parece tener una respuesta general más dinámica al toque de la guitarra que el soft clipping, aunque depende mucho del diseño del circuito. Algunos pedales con soft clipping son muy dinámicos mientras que otros comprimen demasiado, y la mayoría parecen comprimir a la vez que distorsionan.

El hard clipping, teóricamente limita la señal cuando alcanza cierto nivel, de manera que crea los armónicos en altas frecuencias que le dan ese carácter. El soft clipping limita la señal más gradualmente aportando ese sonido más suave que generalmente asociamos con la distorsión de válvulas.

Teniendo en cuenta todas estas consideraciones, veamos mi última modificación de pedal. Quería experimentar estos conceptos con un BOSS DS-1 que compré averiado y que arreglé. No me disgustaba tal cual, pero ese tipo de distorsión no es para mí, francamente.El BOSS DS-1 está basado en el clásico hard clipping simétrico, dos diodos a la salida de un operacional, y la sección de tono tiende a dar una respuesta “mid scooped”, así que debe ser considerado como un pedal de distorsión. Es naranja y se llama “distortion”, así que en este caso es fácil de clasificar.

Mi experimento tenía estos objetivos:

  • Reducir la distorsión (cantidad de “clipping”)
  • Darle un carácter asimétrico, más agradable a mis oídos y más parecido a la distorsión de válvulas
  • Resaltar los medios en la curva de respuesta de frecuencias
DS-1 original schematic

Para reducir la distorsión, he hecho estos cambios:

  • R9 en el emisor de Q2 controla la ganancia de la primera etapa, previa al operacional. Con un valor de 22 Ohms, la salida es tan alta que la señal a la salida del operacional es limitada por los diodos prácticamente siempre. Probé con diferentes valores y el de 100 Ohms parecía ser el punto justo para tener la cantidad de distorsión requerida con el control de ganancia a las 12 en punto, tanto para single coils como para humbuckers.
  • Los diodos LED tienen una Vf (corriente directa) más alta que los diodos bipolares comunes. Los diodos bipolares tienen Vf=0.7V y los LED tienen Vf > 1.5V, dependiendo del color. Vf es lo que provoca el clipping, y cuanto más alto es Vf, más alta puede ser la señal antes de que empiece a limitar. Reemplazar solo uno de los diodos (D5 por ejemplo) por un LED rojo deja más espacio para la señal, y al mismo tiempo cumple con el otro propósito de tener clipping asimétrico. Éste es, por cierto, uno de los cambios del famoso Keeley Ultra Mod.

Para el resalte de medios, cambié varios componentes de la sección de tono. La sección de tono está basada en el control de tono del Big Muff. Es pasiva, y tiende a atenuar los medios. Simulando el circuito en LTSpice, encontré el punto en el que se resaltan los medios de la señal sin perder la capacidad de controlar el tono (aunque esta capacidad se ve algo reducida respecto al DS-1 original).

Modified DS-1 tone stack

El resalte se produce a los 800Hz, que para mí es el mejor punto para conseguir un tono agradable y capaz de ser oído en una mezcla en el contexto de una banda. Es una simplificación, ya que la curva de respuesta del pedal depende del conjunto del circuito, pero os aseguro que el resultado es bastante aceptable.

Los componentes reemplazados en la sección de tono son C11, C12, R15 and R16. Como podéis ver, sólo se han reemplazado seis componentes. He preferido mantenerlo lo más simple posible, pero los cambios en el tono y el comportamiento son drásticos. Ahora el pedal es un overdrive dinámico muy rico en medios, con un rango amplio de distorsión. Con la ganancia al máximo, revela su verdadero espíritu de distorsión. Coloca un compresor delante y un ecualizador después que resalte los 100Hz y conseguirás el verdadero sonido del rock and roll.

Como resumen, éstos son los componentes reemplazados:

componentesoriginalmod
R922 Ohms100 Ohms
D51N4148Red LED
C1122nF33nF
C12100nF82nF
R152.2K3.3K
R166.8K1K

Para C12, 82nF no es un valor muy común, pero puedes poner 47nF y 33nF en paralelo para tener 80nF, suficiente dada la tolerances habitual de los condensadores. Usad un medidor de capacidad si es posible para obtener un valor lo más cercano posible. La placa tiene sitio de sobra para poner dos condensadores en esa posición. Para C11 y C12, cualquier tipo vale, pero los valores son muy importantes (comprobadlos si es posible).

Espero que disfrutéis esta modificación, por favor hacedme llegar cualquier mejora o error que podáis encontrar. Si sois buenos guitarristas, grabad una demo y subidla a youtube. Si sois ricos, enviadme un Klon Centaur para compararlo con mi DS-1 modificado.

PD: si hacéis ese mod, ponedlo en los comentario abajo. Me gustaría saber si la gente disfruta (o no) de mis mods.



BOSS OS-2 Frequency response in different settings

BOSS OS-2 is a good platform for illustrating the difference in frequency response between overdrive and distortion. I have tested my modified OS-2 (see BOSS OS-2 Overdrive/ Distortion MOD) with REW software and my laptop soundcard. It is very simple, you just have to connect the speakers jack to the pedal input and the mic jack to the output. And power the pedal, of course. To make it a little more complex and accurate, I made an adaptor that connects the left channel directly from speaker jack to mic jack and the right channel passes through the pedal. You can configure the software so that the left channel acts as a reference and discount the effect of the frequency response of the sound card itself. Output level has to be kept below some level, in order not to saturate the BOSS buffer, check it with an amplifier connected to the output of the pedal or with an oscilloscope or software tool (more on this in a future post).

REW stands for Room EQ Wizard, and is intended to measure the acoustics of a room and help in equalizing the sound in it. But one of its features is generating a sweep of sine signals and measure the frequency response of a system in front of those signals. While measuring you can see the harmonics, the average 2nd and 3rd harmonic level, and at the end you see the average frequency response for the whole sweep. The software is “donationware” and works in Windows, Linux and MacOS.

Below you can see the different settings I put under measurement:

os-2_blue
Overdrive at max gain, blue

os-2_purple
Distortion at max gain, purple

os-2_green
Distortion at medium gain, red

os-2_red
Blended overdrive and distortion, medium gain, green

And the frequency response of each of the settings:

allfr

As can be seen, distortion settings have a very characteristic “scoop” at middle frequencies (800Hz) that make them more flat sounding. A peak around 100Hz can be seen (and heard as I mentioned in a previous post).

In overdrive settings, on the opposite side, a prevalence of middle frequencies around 500Hz can be seen, what makes it more pleasant and cutting through the mix, as they say.

Combining both (green curve), you have less scoop and the prevalent frequency can be higher, giving more presence to the tone.

 

BOSS OS-2 Overdrive/ Distortion MOD

I have been playing for two weeks with a BOSS OS-2 overdrive/distortion pedal I found in ebay for repair. Repairing was easy, just needed some pots cleaning.

OS-2 with modified knobs and circuit
OS-2 with modified knobs and circuit

The concept is nice: there are two blended circuits:

  • Overdrive, by asymmetrical soft clipping in the feedback loop of an operational amplifier, like most overdrives out there.
  • Distortion by symmetrical hard clipping in the output of the other operational amplifier of the same chip.

You can blend the two by rotating a “color” control, from overdrive on the left to distortion on the right. After the blend section, there is a tone one. The “Drive” pot is a dual 270k potentiometer, each unit connected to the feedback loop of each of the amplifier units.

Out of the brown box (it didn’t come with the original BOSS box) I found some things I didn’t like:

  • Little bass response in the overdrive mode, too much treble for my liking. Usable for me, but could be improved (from my point of view, for my own needs).
  • I usually don’t like pure distortions and hard clipping, so I didn’t expect to like this side too much. I found it too flat and hissing, with little mids, but having a nice boost in the 100Hz (aprox) freq.
Original OS-2 circuit
Original OS-2 circuit

I think there are almost no “bad designs” or “bad pedals” out there, it is just a matter of taste. Many like pedals that I hate (don’t want to give names). Maybe the best overdrive or distortion pedal is no pedal at all, but for people like me that usually don’t play on stadiums, distortion pedals are a good tool to get close to the tone of your favorite player.

Moreover, when you modify a pedal, you are not improving the design, but improving your particular unit for your particular taste. Most components vary a lot in value from unit to unit, and manufacturers have to take many constraints into consideration in their designs. Therefore I think we have to be humble when “improving” a device.

That said, in this case, I considered these objectives:

  • Distortion side:
    • Raising mids and cutting that hiss
    • Trying asymmetrical clipping in this kind of circuits, just for fun
  • Overdrive:
    • Raising mids too, and add some more bass response
    • Trying leds clipping for a supposedly more natural overdrive, and also for fun

After some tweaking and some regrets, I performed to the following changes:

component old value new value why
U2 (opamp) JCR1458D JCR4558D JCR4558D has better characteristics:
1458D
Input Resistance = 1M
Slew Rate = 0,5V/uS4558:
Input Resistance = 5M
Slew Rate = 1V/uSSlew rate affects the circuit bandwidth higher limit, in this case from 8KHZ to 16KHz. Maybe it is overkill, but I think this gives more freedom at adjusting the frequency response of the circuit.
C6 1nF remove More bass in the output of the tone stack
C27 47nF 220nF Lowers freq. in high pass filter at the input of the overdrive section -> more bass
R39 100 150 Lowers gain in overdrive. See C23
C23 4.7uF 2.2uF In combination with R39, it forms a high pass filter, attenuating frequencies below the cut-off frequency. The modification changes the pass freq from 338Hz to 482Hz. In combination with the C27 change, it results in more mids
D7,D8,D9 Junction diodes D7,D8=Red LED – D9=BAT46 (Schottky) It changes the form of the clipped signal. Red LEDs have Vf=1.8V (instead of 0,7) and different I/V curve. I put a schottky just to experiment, another kind of diode can be used, or just a cable for symmetrical clipping. Another LED would be too much Vf and can result in no clipping
R2 20K 68K This resistor is part of the circuit that balances overdrive and distortion. Since LEDs are used for clipping, the output voltage of the operational is too much when compared to distortion output. Raising the value of this resistor lowers the output of the overdrive section. R13 at the end of the distortion circuit can be lowered too, but that raises the cut-off frequency of the low pass filter formed by R13 and C8, not contributing to eliminate the hiss
C16 18nF 22nF Lowers the cut-off frequency of the high pass filter after the hard clipping section, raising mids in the overall circuit
D3,D4 Junction diodes D3=1N4148 – D4=BAT85 (Schottky) + 10 Ohms resistor It changes the hard clipping section from symmetrical to asymmetrical. The schottky + resistor gives a smoother I/V curve than the diode alone. Just an experiment (successful for my ears), as in the soft clipping section
C8 820pF 4.7nF Lowers the cut-off frequency of the high pass filter after the hard clipping circuit. This is key to cut the hiss
BOSS OS-2 modified circuit
BOSS OS-2 modified circuit

Lessons learned

At first I tried LEDs also in the hard clipping section, getting a not so nice result. Probing the circuit with the oscilloscope, I discovered that it was not clipping at all, you could remove the diodes and get the same output. Forward voltage is so high, even for red LEDs (different color LEDs have different Vf), that it didn’t clip at all. The distortion came from the saturated transistor and was not very pleasant.

Then I tried different combinations of Schottky and junction diodes (I like Schottky diodes lately…) until I got to the above blend.

I tried green and blue LEDs and combined LEDs with junction and Schottky in the soft clipping section, but I didn’t like the results. If you have read my other post about SD-1, green and schottky was my final combination in the BOSS SD-1, but it does not seem to work in the OS-2. The final combination was the nicer for my ears, just that.

I put a trimpot instead of R13, in order to adjust the output of the distortion circuit, but the result was catastrophic: more hiss and even oscillation when the trimpot got near zero ohms. So I changed R2, getting much better results.

As you can imagine, the values of the capacitors and resistance are not casual, I have tried many combinations and calculated some filter frequencies to get to those values. Some starting points came from forum posts and some other pages, and I changed some components and at the end returned to the original values (C26 for instance) . The lesson here is: calculate values for the filters involved and act with a purpose. I took some ideas from this post: https://www.roboticbeast.com/modification-de-la-boss-os-2/ but some didn’t wotk for me or with my unit. Another lesson (I already knew, of course) is that every modification affects the whole circuit in some measure, so you shouldn’t change a single component and see if you like it.

Also found a very useful tools for analyzing frequency responses by generating signals and capturing the output of the pedal with a computer and its sound card, more on this in some future post.

PD: if you make this mod, please comment below, I would like to keep track of people enjoying (or not) my mods.

BOSS SD-1 SUPER Overdrive MOD

Out of curiosity, I bought a second hand SD-1, just to try it as is, and try some of the mods people have been doing to this pedal for the last decades. Some mods try to transform it into a TS9 or even a TS808, so you can turn this relatively cheap pedal into a more expensive one. But this box has its own personality, tsd-1he design is very similar to TS9 but with some remarkable differences:

  • Clipping is asymmetrical, two diodes forward and one backwards (in the direction of the operational). Asymmetrical clipping sounds different than symmetrical one, some people describe it as harsh or hairy of fuzzy compared to the latter.
  • There is no capacitor in the feedback loop of the clipping amplifier. This capacitor smooths the clipping a little, giving the TubeScreamer part of its particular tone.
  • Tone control is very similar but with different component values and with a capacitor (C6) in the feedback loop of the tone control operational amplifier. This capacitor is the first thing I have seen every mod removes because it sucks a significant amount of bass frequencies to the signal.

After trying the pedal for a while, I quickly noticed what most people complain about, this makes the guitar sound thinner. This has not to be bad in every ocasion, especially in live gigs situations where you want to sound in a different spectrum space than the rest of the instruments. But ok, I want more bass too. In this clip I recorded the original sound of the pedal, playing with a RS420 (humbuckers) and a Fender Blues Deluxe amp (please don’t pay attention to the music, just the sound 😉 ):

To correct this defect or feature, there are multiple mods out there, usually people remove capacitor C6 and change values of R6 and C3, which connect the feedback loop and the negative operational input to a voltage divider, in a similar way to he TS9 connects those to ground. lowering R6 and raising C3 values results in more bass response from the clipping circuit. See below the schematic:

boss-sd1-super-overdrive
BOSS SD1

I chose to make this changes in order to get a nice tone from the pedal:

  • C6: instead of removing, changed it with a 100pF capacitor from its original value of 10nF.
  • C3: raised from 47nF to 100nF.
  • R6: lowered to 3K3

The other section where people make changes is the clipping circuit. Many mods aim in the direction of getting symmetrical clipping, but I didn’t want another TubeScreamer. Instead of that, I wanted to enhance the personality of the effect by exaggerating the asymmetry of the clipping section. Asymmetry comes from using a different quantity of diodes in each direction, or from using diodes with different specifications, especially with different forward voltages (Vf). Vf is different for the following types of diodes:

  • Ge diodes: low Vf, around 0,3V
  • Si diodes: usually around 0,6V
  • LED diodes: depends on the color, 2V for green ones, 1,6 for red ones.

So using a Ge diode in one sense and a LED in the other, we get asymmetrical clipping, right? Right, but Ge diodes are expensive, hard to find and unstable. I read about simulating Ge diodes with schottky diodes in this article:

http://rezzonics.blogspot.com/2014/01/germanium-diodes-vs-schottky-diodes-for.html

Ge diodes are used in some mods also because of the smooth I-V curve the exhibit. Schottky diodes have a low Vf too, but the I-V curve is more abrupt than the Ge one. By adding a resistor in series, you can get a similar response, at a lower cost and with more stability.

In my mod, I have replaced D5 and D6 (in series in the original circuit) with a resistor (10 Ohms) and a schottky diode (BAT46) with a Vf of 365mV. And D4 has been replaced with a green LED with VF=2V. You can see the measures below:

BAT46green_diode

The components used:

componentes

And the changes to make:

PCB-antes

For the clipping section, that I suppose will be subject of further modifications, I have replaced the original components with sockets. In the picture below you can see the other components replaced too:

sockets

Here you can see a detail of the changes in the clipping section:

detalle_clipping

And the result in the following record. I was not trying to play a song, just to get all the possible tones from the pedal. Please forgive me if I get sloppy:

I can perceive these differences after playing for a while:

  • More volume at the same “level”.
  • Much more bass frequencies, without being too much for a band situation, I think (has to be tested in a band situation).
  • More of the asymmetrical character: “harsher” when playing harder, keeping dynamics (I think more than the original, maybe I am somewhat subjective).
  • I think the sound changes more than the original when moving the drive control: from a smooth, almost TS9-esque overdrive at low gain levels to almost fuzzy and even octave-like at maximum gain.

If you are interested in this mod, please try it, test it and give me your opinion in the comments section. If you improve it, tell me how. If you like it, give me a “like”, and if it’s possible for you, mention my blog to yout friends 🙂 Mods have no copyright, but we like some recognition, don’t we?


Edited on March 16 2019:

Some days ago I put a red led instead of the green one. Now the overdrive is more usable and versatile, although the octave effect is less pronunciated.

PD: si hacéis ese mod, ponedlo en los comentario abajo. Me gustaría saber si la gente disfruta (o no) de mis mods.

BOSS FW-3 sweep range switch (mod)

In a previous post (BOSS FW-3 Mods) I told how I did some basic mods to a BOSS FW-3 Wah pedal, which is basically a Cry Baby design with a potentiometer for varying the Q.

For the capacitor that controls the sweep range (C12), I chose a value of 0,022uF, to give the effect a very personal lower sweep range. I found it problematic in some situations, because the effectiveness of the variable band pass filter as a musical device is very dependent on the color of the signal that it encounters at its input. For instance, if I am using the brigde pickup in an already bright sounding guitar (tele or strat), playing high pitch notes, a low frequency sweep will let the sound almost without volume at the low end of the range. The same happens in the opposite situation – high sweep range with dark sounding pickups, or lower notes.

So, in order to have an all terrain wah effect, I have put a switch for alternating between three different values:

  • 5nF – higher frequencies sweep range
  • 10nF – the original one
  • 20nF – lower frequencies sweep range

Given the behavior of a DPDT on-on-on switch, you can get those three values and their corresponding sweep range values (see https://www.electrosmash.com/crybaby-gcb-95) using three 10nF capacitors, this way:

sweep_switch

Other values can be chosen, of course, to obtain more radical or subtle changes. For instance, with 10nF in the center, 56nF at right and 4.7nF on top you get:

  • 3.2nF – very high range, good for a very bright guitar and higher notes
  • 10nF – original sweep range
  • 66nF – very low sweep, suitable for a bass guitar

The real life device, as built for this project:

IMG_20180604_102036_HHT

As I mentioned in the previous post (BOSS FW-3 Mods), this pedal, apart from being a tank, has space enough inside for storing all the money I have. In this picture you can see how the DPDT switch fits inside the case:

IMG_20180604_131340_HHT

In the main PCB, the cables from the switch are soldered instead of the original capacitor:

IMG_20180604_132724_HHT

And the final result from the outside (the scratches and dents were already there, they are natural relic as this pedal has more than 20 years):

IMG_20180604_133345_HHT

If you have any advice about the execution of the project, or simply have tried it, or in case you had any problem, please let me know in the comments below.

 

PD: si hacéis ese mod, ponedlo en los comentario abajo. Me gustaría saber si la gente disfruta (o no) de mis mods.

BOSS FW-3 Mods

I have a BOSS FW-3 pedal bought in the 90’s when I started to play guitar, looking for a Jimi Hendrix or SRV sound. Why this pedal instead of a Cry Baby or VOX? Because back then you bought what they had in the shop.

20170727_155340

Some months ago I decided to sell it because I didn’t like its sound and behavior any more. Too subtle when placed before dirt and too uneven in volume when placed after. But nobody seemed to be attracted to this squared tank in spite of the low price I asked for it, so I kept it.

It is not a bad product compared to other wah pedals, it is just you get sick of the same tone and problems, and want to get a different tone (and problems). In fact I like its size and robustness, it is solid like a tank and fits very well under my feet.

So I decided to tweak it a bit in order to find a different sound, and started to look for possible mods in the web: nothing at all.

After studying its circuit, I found that the BOSS FW-3 wah pedal is essentially a modded Cry Baby. A very clear and in depth analysis of the Dunlop Cry Baby can be found here: https://www.electrosmash.com/crybaby-gcb-95

There is plenty of information about mods for Cry Baby. In many sites they talk about these particular mods, in Cry Baby terminology, for a GCB-95 PCB:

  • True bypass (more details in the referenced sites)
  • Mid range response: R1 resistor change from  1.5K to 2K-2.7K
  • Q: R5, from 33K to a bigger value, up to 100K. This mod is already done in a BOSS FW3, there is a pot in the pedal panel for adjusting this value.
  • Sweep Range, change C5 from 0.01uF to a bigger value for lower frequencies and a lower value for higher frequencies range. Typically 0.068uF for a bass wah.
  • Gain and bass response: changing R9 from 390 to a lower value will raise gain and add bass content, while a higher value will reduce the gain

In order to apply this valuable information to modify a FW-3 and find our personal tone, we need to map these components to those found in the BOSS PCB. I didn’t fight with the true bypass mod because I like the BOSS buffer, and it is useful for me in most situations.

From the schematics:

FW3

Mod Cry Baby (GCB-95) FW-3
Gain R9 R6
Sweep range C5 C12
Mids response R1 R4

See below the pictures for FW-3 pedal PCB. In the top picture I have marked these components in the board.

IMG_20180201_143735_HHTIMG_20180201_143846_HHT

Which values to choose depends a lot on your personal preferences. I have experimented with some values and ended choosing these ones:

  • R4 (Mids): 2.2K, gives some more mids content.
  • C12 (Sweep range): .022uF, lowers the sweep range in the frequency spectrum, giving a very pronounced WOOOO effect in the lower frequencies end.
  • R6 (Gain): 390. It lowers the gain a bit, making it more even with the effect not engaged

These values give a very personal tone to the FW-3 wah. Even placed before dirt, the effect is very pronounced and significantly different from the original, specially in the lower end of the sweep range.

In a next post I will try to record the sound of the modified pedal.

I’m planning to add some pots in the future to regulate the mids and gain, and even a selector to change among different values for C12. The unit has space in the panel to place the pots and enough space inside to place another two PCBs, so I think it is a perfect platform to play around a bit.

References

http://stinkfoot.se/archives/549

http://www.wah-wah.co.uk/diy.html

https://www.electrosmash.com/crybaby-gcb-95

PD: si hacéis ese mod, ponedlo en los comentario abajo. Me gustaría saber si la gente disfruta (o no) de mis mods.