This Week on the Breadboard: Boss OD-3

[Chuck D. Bones]

This little gem was designed after the BD-2 Blues Driver. Both pedals are still in production. I found this schematic over on FSB. It shows everything except the bypass part of the circuit. I believe that Q8's source resistor is typo'ed, it should be 10K. The OD-3 contains the same discrete FET-input opamp gain block as the BD-2. Unlike the BD-2, it employs a single gang pot to control the gain. Although there are two gain stages, same as the BD-2, the OD-3's overall gain is lower. Both pedals contain tone-shaping filters distributed throughout the circuit. The biggest functional difference between the two circuits is the OD-3 uses diodes in the feedback loop to perform clipping; none of the transistors are driven to saturation.
Here's a quick circuit walk-thru. Q8 & Q10 are unity-gain buffers. The input impedance is 1M (R33). The first filter is an 88Hz high-pass made up of C29 & R46. This will have very little effect on a guitar's tone, but will attenuate the bottom octave of a bass guitar. R34, R35, R48, C22, C31 & the parallel resistance of R36 & R49 form a -4.5dB notch filter centered at 500Hz. This provides a mild de-emphasis of the midrange before going to the first gain stage. Both gain stages contain diodes in the feedback loop to provide soft clipping. D10 & D11 provide clipping in the 1st gain stage. R51 & C33 roll-off the bass below 720Hz. D9 & D12 are protection diodes that limit the input signal to prevent overdriving and latching-up the FET diff amp (Q13 & Q14). Unless you drive the shit outta this pedal with another pedal, D9 & D12 will never turn on. The DRIVE pot varies the gain of the 1st stage from 16dB to 43dB at 1KHz. The 2nd gain stage is a simple JFET booster (Q15) with diode feedback (D6 & D7). This stage is reminiscent of a BMP clipping stage. The gain is a modest 11dB. C34 rolls-off the treble above 5.3KHz. R26 & C16 roll-off more treble above 2.8KHz. R26 and C16 also load the output of the 2nd stage, further reducing the treble content. The 1st half of the M5218AL opamp is a 15.5dB gain stage. What comes next is intriguing. The 2nd half of the M5218AL opamp looks like the treble half of an active James tone stack with the treble control set at noon. The only parts that are not balanced are C7 & C8. What we end up with is a -4.4dB notch centered at 1.6KHz, which de-emphasizes the lower part of the treble band. Next comes the same TONE control as the BD-2, with slightly different capacitor values (C13 & C17). After the TONE control comes the LEVEL control, followed by the output buffer (Q1). Q4 and Q9 are part of the soft-switching.

What we end up with is a light overdrive. The modest gain, soft clipping and frequency shaping produces a very smooth, natural sounding distortion. The diode clipping is symmetric and produces only odd-order harmonics. Even with DRIVE at zero, there is still some noticeable clipping in the 2nd stage. The only way to play clean is to back-off on the guitar's volume. Maximum output is just under 1.8Vp-p.

One other comment re: the difference between the OD-3 & the BD-2. I consider the BD-2 to be more of a clean boost because it only sounds good (to me) when it's running clean. The OD-3 is definitely an overdive in my book because it sounds good running dirty.




As you may surmise, I have a few mods in the works. Some minor circuit clean-up and I converted the fixed tone shaper in the last stage to an active James tone stack. The TREBLE control is tuned to mimic the original TONE control. The added BASS control provides -7dB cut to +12dB boost at 100Hz. The LEVEL control was moved ahead of the tone stack. No output buffer req'd. Overall gain, clipping and output level are the same as the original pedal. "Schematic Coming Soon"

 

[Chuck D. Bones]

Here's a trace from last year by Kanengomibako. This circuit is a little strange because the two clipping stages are powered from a regulated +5V rail. Not sure why Roland did that, but based on my simulations it works fine as long as the JFETs have a low enough Vp. Aion's Heliodor schematic shows those two stages running on +9V, same as the trace above. That's how I did it on my breadboard. BTW, "kanengomibako" roughly translates as "dumpster fire."

 

[jubal81]

Think it'd be worth it to put a switch on C35 for a boost/OD option?

 

[Chuck D. Bones]

Shorting R54 would bump up the gain on the 2nd clipping stage by 10dB. Opening R54 would lower the gain, but it's already pretty low. Either way, it's still gonna clip at the same level, so no significant volume change. You can also increase the gain of that stage by reducing R39. Then the volume will go up a bit too.

One of the things I did on my breadboard was lower R38 to 1K so I could dial the gain a little lower if I wanted to.

 

[jeffwhitfield]

Built the AionFX Heliodor which is an OD-3 clone. Really nice overdrive for sure. I kind of think of it as a mish-mash between the BD-2 and OD-2. I found it to be identical to the Brown drive setting on my OD-200. Kind of odd when you think about it...the OD-3 being what BOSS considers the "brown sound". I mean, it is quite amp like and does sort of have that kind of tone about it. Damn nice sounding OD regardless.

 

[Chuck D. Bones]

Yeah, Kevin makes some nice boards.
The OD-2, OD-3 & BD-2 were all cut from the same cloth. They all employ the discrete opamp building block and perform tone shaping within each stage and between stages. I see the OD-3 as a refinement of the OD-2. For the OD-3, Boss deleted the Turbo channel and beefed up the Normal channel a bit. It's like they were going for a tone somewhere in between Normal & Turbo. The BD-2 is very similar to the OD-2's Normal channel with its cascaded discrete opamp stages and hard clipping. The tone shaping is a little different, but it's the same basic concept.

 

[MattG]

@Chuck D. Bones, thanks for diving into the OD-3! I don't have anything useful to add, just wanted to say thanks for putting this on the breadboard and sharing your findings. I was semi-obsessed with this pedal for a while, I still think it's kind of special. Here's a related link that may be of interest: Guitar Pedals Visualized: Boss OD-3. User "11 Gauge" at TDPRI forums clearly has a fondness for the OD-3, and has written about it extensively, often in technical detail (he's the guy behind Machine Head Pedals).

 

[Chuck D. Bones]

Thanks for pointing me to 11 gauge's comments re: the OD-3 on TDPRI. Good stuff.

I've been messing with the Tone control circuit on my breadboard, and at some point I'll have to decide if my mods made it better, worse or just different.

Here's a peek at how it's wired right now. I renumbered all of the ref des. Some of my mods are about using fewer parts to accomplish the same thing. I also used different transistors throughout because I don't have any 2SK184s or 2SC1048s.

J112 has a high Vp, so I can bias it from ground in stead of Vref. I used that for the input buffer (Q1).

PF5102s are high-gain and have Vp in the right range, so I used those for Q2, Q3 & Q5. Plenty of other JFETs will work as long as Vp is in the right range. I got lucky and the first two PF5102s I tried had the circuit at the correct bias point, so I did not need to adjust R9.

I used 2N5087 for Q4 because it's low-noise and high-gain. 2N3906 is perfectly acceptable.

The original pedal has two input buffers to support the buffered bypass soft switching. One is enough with true-bypass, so I ditched one of the buffers and configured the remaining one to have the correct freq response and input impedance.

I added a switch (not shown on schematic) to select 450Ω, 3.3K (stock) or 11.5K for R7. This moves the mid notch up or down about 3dB. Not sure this is needed, but I wanted to experiment.

I moved the protection diodes from the input of the 2nd stage to the input of the 1st stage. Their purpose is to keep the Q2-Q3 diff pair from going haywire if they're driven too hard. Saved one capacitor (C32, above).

In the original circuit, two 220K resistors (R36 & R49) bias Q2's gate at 1/2 Vcc. But the circuit has Vref, a clean source of 1/2 Vcc, so I used that instead. Saved one resistor.

The discrete opamp circuit (Q2-Q4 and associated parts) has bandwidth up into the hundreds of KHz. That's fine if you're building a radio, not so good with a guitar pedal. I added C17 to roll-off the highs above the audio range. I reduced R11, below, to 1K so I could dial-in lower gains.

Other than subbing a different JFET for Q5, I left the 3rd stage alone. I used 47nF for C5 because it's close enough to 56nF and it's not part of a filter.

The final two stages is where I started messing about. I moved the LEVEL control between the 3rd and 4th stages. I subbed LM833 for the dual opamp, although any dual opamp would suffice. I left the 4th stage gain at 17.8dB. I replaced the filter network surrounding the 5th stage with an active James tone stack. As I mentioned in an earlier post, BASS provides -7dB cut to +12dB boost at 100Hz. The TREBLE control is tuned to more-or-less mimic the original TONE control. It has a little more range and tracks the original freq response curve within a couple of dB.

With the active TREBLE control, I didn't need the passive TONE control any more, so that got tossed overboard.

Because the output comes right off of the 5th stage, no output buffer is req'd.

There are many places to fiddle the gain, clipping and freq response in this circuit. The OD-3 works so well right out of the box it's hard to justify tweaking any of that.

 

[fig]

Thank you.

 

[music6000]

Keith Aviles ''11 Gauge'' is Machine Head Pedals, I have his 98.6. from 10 years ago which is a higher gain version of his 72 Degrees.
He helped me a few times with my Builds, Great wealth of knowledge!

 

[Chuck D. Bones]

Cool! Any idea where I can get schematics?

 

[music6000]

Cool! Any idea where I can get schematics?

Out of respect for his large Input to the DIY pedal building community, no one has released his designs.

If you login to freestompboxes.org & type MHP 72 in the Search box, you can read about it from Keith.
Jubal81 is there for the talk!

 

[Chuck D. Bones]

I'll do that.

I personally am proud to show off my designs. I think it's folly to think that an electronic design can be concealed. Just look at Finnegan's Klon. Someone will trace it, someone will share it, and everyone will see that 99% of pedal designs are based on the work of others. Klon is definitely in the 1%, yet it's cloned for $30 and the originals are sold for 500x and higher.

I study circuits because I'm interested in how other people do it. I appreciate clever designs and I respect those who pull it off. I'm not going to buy an MHP pedal just to trace it, so I guess my interest ends here.

 

[Chuck D. Bones]

I'm calling this breadboard done. The OD-3 is a great pedal in stock form. My mod adds a little more control in the Tone dept.

 

[MattG]

Chuck, do you have any thoughts on simple mod(s) to cut bass while retaining the core tone/voice of the OD-3? Your active tone controls certainly look appealing, but that looks a little tricky to incorporate into an existing OD-3. I'm thinking of something like a three-way switch, with "stock, humbucker, stage" modes, where "humbucker" rolls off a bit of bass and "stage" rolls off even more.

A while ago, based on 11 Gauge's suggestions on TDPR, I reduced C35 and C14 (using the first/J Luha schematic you posted) from 10uF to a few different lower values. The bass and gain were indeed reduced, but it also seemed to remove the "depth" of the tone... which is perhaps a silly statement to make, since I am trying to reduce the bass. But I'm (perhaps wishfully) thinking that it's possible to keep most of the delightfully "thick" tone of this pedal, while making it a bit more humbucker and/or band-mix friendly.

Another suggestion 11 Gauge made, which I haven't yet tried, is to reduce C12 and C15. Maybe it's enough just to reduce C12?

 

[Chuck D. Bones]

If you want to reduce the muddiness caused by too much bass going into the distortion stages, then try reducing C33 instead. Do it in small steps because C33 has a lot of leverage. Leave all of those other caps at the stock values.

These are untested suggestions; you need to find out if they work for you.

I hope you're doing this on a breadboard because that Boss board will not tolerate repeated soldering & unsoldering.

 

[scheffehcs]

Revisiting this post after finishing up my Aion Heliodor build tonight. Thanks for the great write-up Chuck.

I really love this pedal, it just has such a pleasant break up. For me, having the gain all the way down hits a nice sweet spot - just a little added dirtiness. That being said, I might go back and make the change you suggested and lower R38 (R14 on Aion’s schematic) just to have the option to run it clean.

Would love to see a PPCB board for this one!

 

[jimilee]

^^^ This but with a bass switch option for the ODB3. It was my first bass overdrive. I wish I hadn’t sold it…now.

 

[Chuck D. Bones]

^^^ This but with a bass switch option for the ODB3. It was my first bass overdrive. I wish I hadn’t sold it…now.

The OD-3B is a similar, but different circuit. Different enough that it would take major surgery to make an OD-3 behave like an OD-3B. The OD3-B has a clean blend and a (barf) TS-type tone stack. One can certainly fatten up the OD-3, but it won't be an OD-3B.

 

[MattG]

Chuck, I just did your r38 mod (reduce 3k3 to 1k) on my OD-3. I’ve done (and undone) and handful of random mods to this pedal over the years, but nothing really hit home. Until now!

Such an absurdly simple mod, but solves what was for me a frustration with the OD-3: there was little to no dirty boost/edge of breakup available. I always wanted a lower drive setting than what was available. Now I have it!

Thanks again for the great analysis and simple r38 tweak!