Blue Pilot, an integrated Blues Driver

Original effect designs complete with schematics, layouts and instructions. Freely contributed by members or found in printed or online publications. Cannot be used for commercial purposes without consent of the respective owners of the copyright.
Post Reply
User avatar
dylan159
Breadboard Brother
Information
Posts: 124
Joined: 18 Jul 2018, 14:02
Has thanked: 39 times
Been thanked: 234 times

Post by dylan159 »

INTRO
Hello everyone. The Blues Driver always pops up among the most popular overdrive pedals, no doubt because many people like how it sounds, but also because it's easy to find and relatively affordable, being from one the biggest pedal company in the world. It's not as popular in the DIY world, aside from mods, probably because it's not the smallest circuit and uses many transistors in a discrete op-amp configuration. Also since you can get it anywhere for cheap, I guess most people just don't bother.
I think that's a pity, because it's an interesting circuit which I'll go through while showing you my attempt at making a version using IC op-amps, which results in a simpler, cheaper and more DIY-friendly circuit. Also in this more familiar-looking form, you can use it an insipiration to make your own BD-flavored distortion.
SCHEMATIC
bluepilot.png
THE DISCRETE OP-AMPS
So, about the discrete op-amps. The Blues Driver features two such stages, with JFET input devices and a PNP BJT for the second amplifier. Their open loop gain depends heavily on the JFET used, but simulating/calculating with the typical transconductance found on the datasheet, it's probably not too far off to estimate it at above 300-400. That's not much compared to any integrated op-amp, but it's not too little either, especially for small closed loop gains when the gain knobs aren't maxed out. This should result in the linearization and abrupt clipping transition associated with op-amps and large loop gain, so it's not unreasonable to use an integrated one, especially since examples of op-amp clipping are already everywhere (this included? :) ) and well liked despite their bad name. I don't have an original to take measurements from, but simulation shows very similar transfer curves for both versions.
There's other things to consider of course. Clipping is symmetrical and headroom is 8V p-p, which means basically the full supply voltage as advertised on the schematic, because of the capacitance multiplier on the supply rail. I really liked this being here, but it's one thing we can happily get rid of, since op-amps benefit from very good PSRR.
Anyway, a TL072, as many other jellybean op-amps, has at most 6V p-p of output swing on a 9v supply, so this difference in headroom has to be compensated in gain to get the same amount of clipping and same input headroom. This is one reason the values for the feedback divider are different from the original for the first stage. This doesn't apply to the second stage since the voltage at its input will be a fraction of the original's exactly because of the reduced gain and output swing of the first stage, and the two amplifiers being the same, that's already the same 6/8 ratio to apply to the second stage.
A second reason the values have been changed is to compensate for finite open loop gain. While the open loop gain in the original is reasonably high, it's still lower than that of the 072, so this needs to be taken care of to get the same closed loop gain. Capacitor values have been tweaked to get a similar high and low frequency response as the original in simulation.
THE TONE STACK
Next, let's talk about what's between the two op-amp stages in the original: what the keen-eyed will recognize as a fixed FMV tone stack, followed by clipping diodes. This looks very interesting and suggests a design based on a Fender preamplifier, with the tone stack placed between two gain stages, but in reality it turns out to not be that exciting.
First the tone stack: at first look you can already tell its controls are set to maximum middle and bass and minimum treble. What this results into is basically a shelving filter attenuating by 6 dB everything above 100Hz, with a tiny amount of ripple above that. I've analyzed this together with the RC highpass at the input of the next stage, which cuts a not negligible amount of bass, and you'll see why doing this is legitimate shortly.
Knowing this, I set myself for making a simpler filter that achieves the same response. This is composed by everything between R4 and R6. The difference is at most -1dB at 10kHz. Good luck hearing that, especially before some clipping :D .
tonestack.png
Now, the diodes. I was a bit surprised to find that these aren't doing much at all. In hindsight, it makes sense because they're after all the attenuation from the tone stack, which is also an equalization curve, so that even with the full 4V peak swing from the previous stage, only frequencies below 150Hz have enough amplitude to clip, and even then the conduction is very limited, a few μA and correspondingly low forward voltages of about 0.9V. That's why I've decided to remove these altogether. It's safe to say that the op-amps are the main source of clipping in the circuit, and the contribution from the diodes might or might not be audible. If someone has a blues driver to pull the diodes out from for an A/B comparison let me know :D.
THE TONE CONTROL
Here comes another block which looks more interesting than it is. The tone control is made from two filters, a fixed high shelf and a variable treble cut/relative boost. The result though, is very similar to that of a simple variable high shelf like those found in many pedals and guitars, save for a bit less ripple in the pass band. So I swapped it out for exactly that, tweaking values and loading from the volume pot to get a close response and good sweep across the tone control:
tonecontrol.png
I invite you to focus on the extremes of the sweep and the very small ripple of the simpler version. The center position graph is an approximate match since the taper of the two tone pots is different and the sweep slightly too. The highpass at 45Hz is negligible too, but if desired a 220nF capacitor before the volume pot takes care of that.
THE GYRATOR STAGE
The gyrator stage adds 6 dB of boost at around 120Hz with a Q of about 2.4. I didn't have to change much here except tweaking some values to achieve the same response with an op-amp buffer, which will have resulted in a narrower and bigger peak if left as it is. I've also used some more friendly values overall, in this stage and in the rest of the circuit. You can see some E12 values, but they're only used when the increased accuracy in frequency response or gain is necessary, otherwise I had no reason to use more than the humble E6 series, since some careful scaling and tweaking could get me where I wanted.
The protection diodes at the inputs have been left out, because even in the worst case of a full swing at the peak frequency (which in case you're wondering will be LOUD) the differential input voltage is within limits and there's no noticeable difference at the output.
DEMO
Here's how it sounds (warning, previous version with wrong gyrator values):
CONCLUSION
The BD-2 is one of the most popular overdrive pedals around, but hasn't got as much attention in terms of DIY. With this version I hope to have given both an easy to build version and a starting point for people to start experimenting with this kind of circuit.
Last edited by dylan159 on 15 Oct 2021, 07:20, edited 2 times in total.
Find me at https://discord.gg/NSveWjbdxs Audio Electronics discord server.

User avatar
asbestosaurus
Information
Posts: 17
Joined: 27 Mar 2020, 00:55
Has thanked: 20 times
Been thanked: 29 times

Post by asbestosaurus »

Fully prepared to be wrong on this, but it looks like your gyrator stage is based around a version of the schematic where the feedback cap (C12 in your schem, C16 or C18 in the boss schem - its a little hard to read) is 10x smaller than the series cap (C11 in yours, C9 in the original).

In the Boss factory schem these appear to be the same value (56nF) which I'm fairly sure I visually confirmed for myself a few months ago with a real unit from around 2010, according to the serial number. This moves the resulting peak to about 120 Hz and bumps the Q up for a better response from the transistor used in the Boss version. I've encountered both versions all over the Internet so I don't know if earlier pedals have a different cap in them or if it was a tracing error.
Last edited by asbestosaurus on 15 Oct 2021, 19:34, edited 1 time in total.

User avatar
dylan159
Breadboard Brother
Information
Posts: 124
Joined: 18 Jul 2018, 14:02
Has thanked: 39 times
Been thanked: 234 times

Post by dylan159 »

asbestosaurus wrote: 15 Oct 2021, 06:02 Fully prepared to be wrong on this, but it looks like your gyrator stage is based around a version of the schematic where the feedback cap (C12 in your schem, C16 or C18 in the boss schem - its a little hard to read) is 10x smaller than the series cap (C13 in yours, C9 in the original).

In the Boss factory schem these appear to be the same value (56nF) which I'm fairly sure I visually confirmed for myself a few months ago with a real unit from around 2010, according to the serial number. This moves the resulting peak to about 120 Hz and bumps the Q up for a better response from the transistor used in the Boss version. I've encountered both versions all over the Internet so I don't know if earlier pedals have a different cap in them or if it was a tracing error.
you're probably right, thanks for checking. the weird blend of popular and neglected reflects in the fact that I was able to find only two schematics: the factory one, which I followed even with its readability issues, and an unofficial one which was full of errors. Easy enough fix though. Updated.
Find me at https://discord.gg/NSveWjbdxs Audio Electronics discord server.

User avatar
asbestosaurus
Information
Posts: 17
Joined: 27 Mar 2020, 00:55
Has thanked: 20 times
Been thanked: 29 times

Post by asbestosaurus »

After doing a bit more digging I've been able to confirm that the series cap C16 is indeed 56n. C18 is a 100n near the input. I've found a few gutshots, just on google images, of popular boutique mods or reverb listings etc. where the label on C16 is clearly visible and reads 563.

It could be cool to have both responses available. I'm sure a non-zero quantity of clones or mods or diy builds exist with that smaller gyrator feedback cap, just based off of the prevalence of the alternate schematic. so it's probably been heard before.

User avatar
dylan159
Breadboard Brother
Information
Posts: 124
Joined: 18 Jul 2018, 14:02
Has thanked: 39 times
Been thanked: 234 times

Post by dylan159 »

asbestosaurus wrote: 15 Oct 2021, 13:01 After doing a bit more digging I've been able to confirm that the series cap C16 is indeed 56n. C18 is a 100n near the input. I've found a few gutshots, just on google images, of popular boutique mods or reverb listings etc. where the label on C16 is clearly visible and reads 563.

It could be cool to have both responses available. I'm sure a non-zero quantity of clones or mods or diy builds exist with that smaller gyrator feedback cap, just based off of the prevalence of the alternate schematic. so it's probably been heard before.
Clones? pcb or kits seem pretty rare, or I'd have used those schematics too as comparison. In any case a 120 Hz boost seems much more desirable than a bigger 380 Hz one. Also if I were to follow that other schematic there was a missing cap on the tone control and maybe even more.
C18? You mean the one part of the fixed high shelf that is marked as 5.6n on the schematic?
Find me at https://discord.gg/NSveWjbdxs Audio Electronics discord server.

User avatar
Frank_NH
Solder Soldier
Information
Posts: 196
Joined: 12 Jun 2013, 14:18
Has thanked: 40 times
Been thanked: 87 times

Post by Frank_NH »

Very coo design,l Dylan159 - would love to build this. I was wondering if a rail-to-rail op amp such as the TLC2262 would work better given that the BD-2 relies on op amp clipping. Also, would it be beneficial to run at 18V?

I've experimented with discrete op amps in the past (made a tube screamer from discrete op amps) and to my ear the resulting tone was a little less harsh (smoother) than with an IC. I used a JFET discrete op amp similar to the BD-2, and it seemed to work OK even with unmatched JFETs. In any case, I like the use of ICs in your design as it provides a lot more repeatability than using discrete components.

User avatar
dylan159
Breadboard Brother
Information
Posts: 124
Joined: 18 Jul 2018, 14:02
Has thanked: 39 times
Been thanked: 234 times

Post by dylan159 »

Frank_NH wrote: 15 Oct 2021, 19:29 Very coo design,l Dylan159 - would love to build this. I was wondering if a rail-to-rail op amp such as the TLC2262 would work better given that the BD-2 relies on op amp clipping. Also, would it be beneficial to run at 18V?

I've experimented with discrete op amps in the past (made a tube screamer from discrete op amps) and to my ear the resulting tone was a little less harsh (smoother) than with an IC. I used a JFET discrete op amp similar to the BD-2, and it seemed to work OK even with unmatched JFETs. In any case, I like the use of ICs in your design as it provides a lot more repeatability than using discrete components.
In both cases you're changing the headroom, which all other things being equal has the same effect as changing gain except for the small but real effect on clipping behavior (the "knee" size is relatively smaller to the signal swing with higher voltage and gain). In any case a clipping op amp is an op amp with a feedback not working as intended, so that's what matters also for the second part. I think you're right in saying matching isn't critical for some things, even if it is for others like DC precision and common mode rejection. A discrete op-amp for sure is worse in some ways, especially open loop gain and CMRR, a fact Boss knew in using a cap multiplier. Anyway let's not forget how much more compact this makes the circuit too!
Find me at https://discord.gg/NSveWjbdxs Audio Electronics discord server.

User avatar
bmxguitarsbmx
Resistor Ronker
Information
Posts: 308
Joined: 25 Mar 2010, 21:15
Has thanked: 401 times
Been thanked: 121 times

Post by bmxguitarsbmx »

Sounds real good Dylan. Impressive work again!

Post Reply