New Blues Blaster Overdrive Distortion

[Mark Hammer]

I've seen a lot of squaring/fuzz that have low input impedance. It must help reduce the "harmonics of harmonics" fizz in the fuzz and let the squarers track the fundamental better. But it ceases to help if you have any buffer between it and the guitar.

Which is pretty much why I suggested the additional cap.  Trimming the high end is the goal to achieving what you describe.  One can do that by tweaking the input impedance, OR by trimming the treble from the signal that drives the clipping mechanism.  Tone-shaping in the op-amp is key here.

[pinkjimiphoton]

but there's a bit of a dif in application still. one is more of a tone control, the other affects the interaction of the guitar and effect more.
to me, the goal is as much interaction READ:dynamics as possible. the "feel" is affected as well as the tone i think with the impedance more, cap alters tone by filtering the feedback somewhat. . different ways of accomplishing similar things.
remember, we're purposely distorting stuff, not necessarily trying to make it happen in a traditional or linear way. we're seeking to abuse the stuff to find that tone that makes your pants too tight.

sorry if i'm questioning everything but i'm questioning everything, and have been learning a lot more about implementing high and low pass filters... especially in guitars.  it's all about the tone, so if it ssounds good and fills a pocket...

[Ben Lyman]

Seems I've opened a can of worms here 
Sorry, not my intention but thank you all for the helpful hints... now I'm very eager for Poppy to unwrap his gift so I can plug into it again!

[Gus]

Ben Lyman
Look into what you should do with the not connected pins of the other opamp in the dual opmap
Look into noise and opamps.
Think about noise and gain of the stage

[Ben Lyman]

OK Gus, it looks like I am supposed to ground all 3 unused pins. That should be easy with my layout because the ground already links with pin 4.
Thanks

Let me run this one by you all, I want to see if I am understanding any of what you are saying.



Oops, I think I didn't mean to put that extra 10k at the input... or did I?

[R.G.]

The extra 10K at the output is what gives you treble rolloff with that cap across the volume control. As it sits now, it starts cutting at about 4.8kHz, neglecting the oddity effects when the diodes are not conducting.

Note that the 330pF and 150K feedback resistor are already starting to roll off treble from the amplifier at 3.2kHz, but that is before the diodes have clipped whatever is there. The 10K/3.3nF is rolling off both signal and clipping products.

[PRR]

> ground all 3 unused pins

NO!!

Tie Out to -In. Tie +In to Vref.

[PRR]

> imho most distortion designs i've built that use a high input impedance often have way too much high end

Granted, but....

"Low input impedance" does not equal "less highs" EXCEPT the special case of a guitar. Which maybe was not so special long ago, but today we often get buffered signals from active guitar, wireless, or other pedals.

IMHO: Take the signal verbatim (hi-Z), then whack-down the highs yourself.

This also avoids the problem of the hi-cut changing on every guitar you own. You can pick the tone for musical reasons, not compromise with your gear.

True, this may mean an added stage. However *this* plan is using half a dual op-amp.

> sometimes, you DON'T want a buffer between your guitar and effects...

OK, but....

> put a buffer in front of a true fuzz? heresy.

Because so many antique (and copycat) fuzz pedals took no consideration of buffered signals. But some of us don't like to live in 1967 any more. (Says the proud owner of a 1967 Ford backhoe.)

[Ben Lyman]

> ground all 3 unused pins

NO!!

Tie Out to -In. Tie +In to Vref.

Ok... I think I get it now... sort of. I can do all that easily enough but I am unclear what is happening if I do it. Why does the unused half of the op-amp need to have VR to +In?

[pinkjimiphoton]

>

This also avoids the problem of the hi-cut changing on every guitar you own. You can pick the tone for musical reasons, not compromise with your gear.

but.... i must counter, why would anyone want all their guitars to sound the same thru their rig? i appreciate generic rock distortion tones as much as the next guy, but i always prefer pedals where you can  tell which guitar you're playing with it. personally, i
WANT every guitar i own or play to sound different, that's the point of maintaining so many of 'em

[PRR]

An un-used not-connected opamp *may* find a way to make trouble. They have large gain far beyond the audio band. They will amplify any stray signal, even their own self hiss, to infinity (or till slamming the power rails).

It's like having dogs. Sometimes they just lie there. Sometimes they go berserk. It is good practice to set things up so they don't make trouble: leashes, fences.

By wiring Out to -In, you say "do what the +In does!". By wiring +In to Vref, you say "Stay at +4.5V and don't move!"

Vref is not critical. It can be any steady voltage within the input common-mode range. However most opamps don't handle zero or +9V well. Some even do a wrong thing. Vref is handy and safe.

Actually, if this is a '4558 project, the un-biased bipolar inputs will probably drift to an extreme and jam-up fairly harmlessly, unless there is huge stray signal which brings them out of the jam. But a distortion pedal has large signals to go astray. And you might want to try a BiFET or CMOS opamp, and these can hover in the "active" zone pretty freely.

[duck_arse]

An un-used not-connected opamp *may* find a way to make trouble.

isn't there any single op-amps equivalent to the 4558? if there is only one already distracted amplifier, it can't possibly get up to hijinks.

[samhay]

>Let me run this one by you all, I want to see if I am understanding any of what you are saying.

Your gain control isn't going to work very well now. At the moment, it will be almost max to max gain.
To get the same voltage divider effect, you would have to scale the value of the pot up by 100 x (470k vs 5k), which isn't feasible.

[Hatredman]

sometimes, you DON'T want a buffer between your guitar and effects... put a buffer in front of a true fuzz?
heresy.

buffers often tend to make stuff sound very hard and very bright. they are by all means useful, but too many can destroy your tone.

Hummm... you can use buffers to take the impedance issues out of the equation, and then work on the equalization network inside the Fuzz circuit to get  rid of the brightness and harshness.

We should also remember that the great Fuzz Face, Muff & etc tones we hear on our beloves 60s and 70s records (Hendrix anyone?) are in fact heavily post-processed by the recording engineers and producers, so we just DONT HEAR the true, raw Fuzz Face sound on those records. Modern-day FF builders should take this into account and maybe re-create this whole chain of processing inside the pedal. This way we can get the tones we hear on the record when playing live.

Just my tuppence.

[pinkjimiphoton]

>Let me run this one by you all, I want to see if I am understanding any of what you are saying.

Your gain control isn't going to work very well now. At the moment, it will be almost max to max gain.
To get the same voltage divider effect, you would have to scale the value of the pot up by 100 x (470k vs 5k), which isn't feasible.

sam, couldn't he just swap out the 150k pot in the feedback loop for a 100k or 250k pot to vary the gain?

[Gus]

The interaction with the guitar cable and effect can make a differences as well as if a effect is between them.

I did tests years ago with friends and found under 220K input resistance and some RC highpass often did not sound good with an overdrive or buffer this was a compromise for a possible guitar/bass and cable or another effect between with say a 470K input resistance
NOW with that posted
Fuzz and distortions are a little different depending on what could be before them a guitar/bass or another effect of unknown output Z and voltage and current drive 

[pinkjimiphoton]

sometimes, you DON'T want a buffer between your guitar and effects... put a buffer in front of a true fuzz?
heresy.

buffers often tend to make stuff sound very hard and very bright. they are by all means useful, but too many can destroy your tone.

Hummm... you can use buffers to take the impedance issues out of the equation, and then work on the equalization network inside the Fuzz circuit to get  rid of the brightness and harshness.

We should also remember that the great Fuzz Face, Muff & etc tones we hear on our beloves 60s and 70s records (Hendrix anyone?) are in fact heavily post-processed by the recording engineers and producers, so we just DONT HEAR the true, raw Fuzz Face sound on those records. Modern-day FF builders should take this into account and maybe re-create this whole chain of processing inside the pedal. This way we can get the tones we hear on the record when playing live.

Just my tuppence.

well stated on the post processing, but that's more true with the ff full on with the guitar cranked. to me, that's only a barely useable sound most of the time. with me it's all about the interaction of the guitar and fuzz. there is every single shade of classic rock and blues distortion you've ever heard capable with one that's good. (when it works, of course, lol).
i've built a bunch of silicon ones with buffers and various trims and filters but none of them sound the same or right... they all lose the interaction with the instrument. a buffered signal is gonna do that... it's the reason for the season, to electronically isolate the two circuits or parts of the circuit.  but a proper fuzzface or tonebender needs to be driven right from the guitar itself, so the fuzzed guitar can drive the amp. i dunno about you, but i tend to play my amp and effects more than i play my guitar itself.
i have yet to find a solid state fuzz i've been happy with. i realize my tastes in tone and sound are different from a lot of peeps. i'm not interested in the sound of the fuzz cranked up with my guitar up all the way luxuriating in square wave mud... i like the sound of the fuzz cranked, and my guitar between barely on and maybe 6 or 7.  freakin magical. that's why i tend to cobble together things the way i do so they suit my ears, i make no pretense about being any form of EE, i am strictly a 10 toed freak of a monkey playing with a breadboard, but i do tend to find some cool sounds out of things and def take the left hand path of circuit molestation and abuse. i'll smoke some components if i have to til i figure out how to get it to sound right.

my two cents

[amptramp]

zero problems i've encountered with this. 15.4k is pretty high.... guitars don't have much above maybe 6k, so i don't see what the point would be of needing the filter cut off so high?

The 15,400 Hz filter is not there so much for the effect it has on the audio, it is there to ensure you do not get into RF oscillation that may not propagate through your system but may limit the dynamic range and voltage handling capability for itself and the next pedal.  You need a scope to spot it properly and sometimes oscillation only occurs on part of a waveform.  The 47 pF capacitor as suggested cuts the gain of the amplifier stage at high frequencies and keeps the phase response from going to 180 degrees before the gain drops below 1.  This is the condition for oscillation and since the 5534 works at RF frequencies, you may not be aware of it - it may just cause some muddiness at high inputs that you can't find without a scope.

[Phoenix]

>

This also avoids the problem of the hi-cut changing on every guitar you own. You can pick the tone for musical reasons, not compromise with your gear.

but.... i must counter, why would anyone want all their guitars to sound the same thru their rig? i appreciate generic rock distortion tones as much as the next guy, but i always prefer pedals where you can  tell which guitar you're playing with it. personally, i
WANT every guitar i own or play to sound different, that's the point of maintaining so many of 'em

The points that you Pinkjimiphoton, and Paul (PRR) are trying to make, are actually the same, just two very different approaches.

When I'm designing effects, I like to make sure that they have a high input impedance (~1MEG), with minimal band limiting until after the input buffer so that I can then tailor any frequency shaping to a known source impedance (that of the buffer). This means that my designs will sound the same regardless of whether they are connected directly to the guitar, or through a buffer.
Many classic designs such as the Fuzz Face or Dallas Rangemaster use the source impedance as part of the frequency shaping, and this means that if they follow a buffer, this frequency shaping is effectively eliminated, and they sound different to how they were intended (usually terrible).
A solution to this is to add a buffer in front of the Fuzz Face, but give the buffer an output impedance that resembles a guitar! To really do this properly, it's a complex impedance, and requires inductance, series resistance, parallel capacitance... etc. That can be done http://www.muzique.com/lab/pickups.htm, but is usually more effort than you want (or need) to go to.
Instead, if we're designing from scratch (or just tweaking an existing circuit), we can just set all these tonal characteristics after the input buffer, and the circuit will behave as expected regardless of the source (guitar or buffer).

You've made the argument that you want to hear the difference between different guitars when plugged into your circuit, with the implication being that the low input impedance of the posted circuits helps with that. In fact, it does exactly the opposite. The complex impedance of a guitar has a strong resonant peak, and the Q and frequency of this peak is the most defining characteristic of different pickups. However, as a guitar is more heavily loaded, this peak is damped, and the frequency response of even very different pickups quickly starts to look very much the same much below about 300k. At 9.9k like the input impedance of the circuit posted at the start of the thread by Ben, pretty much all pickups are going to sound approximately the same.
Now, he could have used the other half of the 4558 as a high input impedance buffer (although the 4558 being a bipolar input opamp can have a minimum input impedance of 300k, but typically 5Meg, it's maybe not the best choice for a high input impedance buffer compared to a jfet input opamp like the TL07x with typical 1TeraOhm input impedance), and followed that with a 10k-ish series resistor on the output (EDIT: to approximately mimic the output impedance of a guitar), and the circuit would probably sound pretty much the same, but would work with a buffer in front, and would actually retain more of a particular pickup/guitars sonic characteristic.

I'm not going to say that the low input impedance trick is wrong, but it is certainly unpredictable, as you've observed. I like my designs to work exactly as intended, regardless of what's in front of it.

Anyway, hope my ramblings are of some use to somebody!

[Phoenix]

Thought I should add something to better illustrate my point.
The attached image shows the frequency response of two different (simulated) pickups, a single coil, and a humbucker.
The top teal trace is the frequency response of a single coil loaded with 1Meg.
The red trace below that is a humbucker loaded with 1Meg.
You can see that there is a fair difference between the two, with the resonant peaks having different Q's, amplitudes and frequencies.
Below that is the green trace, the same single coil circuit loaded with 10k, and the bottom is the humbucker loaded with 10k.
As you can see, the response of the previously fairly different pickups now looks surprisingly similar to one another. There are differences, sure, but it's mostly in the very high frequency roll off, far beyond the frequency range of a guitar or amp, and the resulting pickups end up sounding very similar. This is one good reason to use a high input impedance buffer to deal with unknown source impedances, and then do your tone shaping post buffer, when you have a known and fixed impedance.

Hope this helps.

EDIT: Here's the plots overlaid on top of each other for a more direct comparison.