The Aefea Drive - a simple one-chip diversion

[commathe]

Very, very intersting! Definitely want to try this one out! I might try putting the adjusted amz body/tone control for a bit more tonal shaping control:

[Mark Hammer]

I opted for something that could yield a broader range of sounds.  The BMP-style control, as elegant as it is, operates in reciprocal fashion.  If you want more bass, it MUST be at the expense of mids and treble.  If you want more bite, it is at the expense of bottom.  That's absolutely perfect for some contexts/circuits, but I wanted something more chameleon-like that could get different combinations.

But ultimately, if the BMP-style control, with or without Jack's mods, works for you, go for it.

[Frank_NH]

Hey Mark,

Perhaps this is a good opportunity to ask if you could elaborate further on your tone circuit in this design (e.g. how did you arrive at the resistor values outside of the first order filter e.g the 6k8).  I would think that in the low pass filter mode, the roll-off at 76Hz - 1.6 kHz would be somewhat dark sounding even at 1.6 kHz (someone at the tagboard site reported this).   And I guess adding the cap in parallel brings back the treble, but the bass-mids are still attenuated by the low pass filter, hence the "brightness" in this mode.  I suppose I'm thinking of potential alternatives/mods for my build if it turns out to be too dark/bright for my tastes...

[Mark Hammer]

The rather ancient midscoop filter found in so many fuzzes, but nicely illustrated in the Shin-Ei FY-2, uses 4 components.  Here, we see a 10k and 15k resistor, a 100nf cap to ground from their junction, and a 1000pf cap that straddles the whole thing. 

How is this a midscoop?  The 10k and 100nf to ground form a singlepole lowpass filter, whole treble rolloff begins around 160hz.  That's pretty low but since it is a shallow slope, there is still a fair amount of mids that pass through.

The highs sidestep everything and pass through unaffected, via the 1000pf cap.  Since we couldn't have a lowpass function if that cap was tied directly to the junction of the 10k and 100nf, the 15k resistor serves to essentially isolate the highpass from the lowpass portion.  Its value also serves to more or less adjust the relative amplitudes of the high and low-freq content.  Voila! lows+highs and reduced mids - midscoop!

If one varies the value of the resistance leading up to the cap to ground, you change where the rolloff begins.  I employed a small value resistor in series with a pot and another fixed resistor to create a fixed total resistance that runs in parallel to the bypass cap (4700pf or the suggested 3900pf, whichever you prefer).  The pot allows for the rolloff point to be varied, with the pot value setting the range of adjustment, and the small value resistor setting the highest rolloff point.  The 6k8 resistor serves the "decoupling/isolation" function that the 15k serves in the schematic below.  I aimed for a total pot+2 resistor value that was more or less in the ballpark of the FY-2 and similar.  No wanting it too scooped, I opted for a higher-value bypass cap (4700pf instead of 1000pf).

It occurred to me that if one was going to have an adjustable midscoop, produced by the pot-adjusted lowpass section, that if the rolloff could be set high enough, AND you had sufficient mids and highs to begin with, that you could simply lift one end of the bypass cap for a "rounder", less edgy sound.

As a fan of 3-position toggles, I mused about what I might be able to do with one, and realized that if I used it to replace the bypass cap with a fixed resistor that let everything through, but at a lower level (because the resistance is larger than the 1k+pot+6k8), that the variable lowpass could simply serve to introduce additional lows and mids on top of a basic full bandwidth signal.

So, the toggle gets you standard midscoop function (cap connected), variable lowpass only (no connection), or full bandwidth+variable lowpass (30k connected).

If a person finds it too buzzy overall, one possibility is to run a small-value cap to ground from the common of the toggle - say 3300pf, for starters - to roll off highest highs in any of the modes.  Alternatively, one might stick a smallish value resistor after the 4700pf cap - say 2k2-4k7 - and stick a cap to ground (.015uf, if using 2k2) from the toggle lug where you are going to connect the bypass cap.  I stipulate that the additional cap is connected to the switch because you don't want it to present a constant path to ground even when not selecting the additional top end.  If it goes to ground from the unselected lug of the toggle, it has nothing to bleed off, and the bypass cap will still have one end "floating".  Select the bypass cap for the midscoop mode and the additional 2k2 and .015uf cap will both drop the top end a little in amplitude, when mixed in with the lows, and also shave off the higher treble.

Does that make sense to you?

[Frank_NH]

Thanks a lot Mark!  That was very helpful.  I may experiment with changing the 100nF cap to 47nf or 39nF.  That would boost the roll-off frequencies in the low-pass network, perhaps reducing "darkness".  The 3.9nF cap and 30K resistor will remain as specified.

[vigilante397]

Okay, so I finally got my Tayda order in and got this all finished up. It sounds great (even side-by-side with my modded RAT, and I LOVE my RAT. It seems to be working like it's supposed to (aside from the couple pots I wired backwards ), with the exception of my gain pot, which seems to do nothing. The shape pot on the other hand seems to change the shape and the gain. Any thoughts?

[Frank_NH]

I finally got my Aefea drive working this morning! 

Before I get to my impressions of the sound so far (I really haven't had a chance to play around with it extensively), a quick note on perseverance:

I made the vero board and populated it last Saturday and Sunday.  Unfortunately, when I fired it up...nothing.  So I checked all components, part values, placements, and orientations first.  All appeared to be OK.  Then I took my meter and checked for solder bridges - arghhh, found two, but STILL it didn't go.  So I looked at voltages and was about to chalk it up to bad ICs last night (I have both a TL071 and several OPA134s).  Then this morning at breakfast, I decided to try one more time to check continuity - what?  Another solder bridge??!!  Yep - between pin 1 of the IC and an adjacent track.  It then fired right up!  Success!!  (whew).  Sometimes, builds are like this, although I've gotten lucky and had a couple fire up the first try.

So...it sounds great.  I stuck to the original design, with the tone section modded values (30K, 3.9 nF) and a 25K tone pot.  I also used a 150 pF cap in the feedback circuit.  What I noticed is that with the hardness pot maxed the gain pot doesn't have much effect.  This is as is expected since the 500K of the hardness pot dominates.  If you turn down the hardness, then the gain responds normally.  The two controls are really interactive as many have commented previously.  There's enough drive for an overdrive, but I suppose if you want more gain you can experiment by upping the gain pot values or playing around with the 3.3K/47nF, 4.7K/220nF network.  It may also help to have log pots versus linear, but that's just a detail.

But what I really LOVE about Aefea is the tone control, particularly the mid scoop mode.  That's a really cool sound, and something I don't have in any of my other DIY overdrives.  It will be worth boxing this up for that feature alone.

My current build is a bit gnarly and has some sockets for swapping caps.  I may go ahead add some other sockets so I can dial the sounds in to my liking.  Then I'll remake the circuit and box it up.  By the way, the OPA134 sounds great with this, but I'll check out some other ICs.

Thanks again Mark for an interesting and unique project 

[Mark Hammer]

Okay, so I finally got my Tayda order in and got this all finished up. It sounds great (even side-by-side with my modded RAT, and I LOVE my RAT. It seems to be working like it's supposed to (aside from the couple pots I wired backwards ), with the exception of my gain pot, which seems to do nothing. The shape pot on the other hand seems to change the shape and the gain. Any thoughts?

The gain pot really only comes into its own when the clipping/shape pot is set to softer/softest clipping.  Once the clipping/shape pot has contributed 300k or so to the total feedback resistance, the extent of clipping is enough that adding another 250k on top of that probably won't change the sound much.  If ALL or most of the feedback resistance is coming from the Gain pot, then adjustments of it will be noticeable.

I used a TL071 but didn't really experiment with any other chips, to be honest.  Since people have tended to have preferences for chip type when clipping diodes form part of an op-amp feedback loop (starting with the JRC4558 frenzy over Tube Screamers a decade back), I imagine it to be worthwhile to try other chips and see if they appeal more.  In that respect, since the available choices increase when using duals, as opposed to singles, building it with a layout for using one half of a dual may be worthwhile.  I know that when a buddy brought over his Timmy pedal (a real one, not a clone), we were pleased by the improvement made when a 1458 was installed.

[jaredmcohen]

This is brilliant Mark. I built it from IvIark's vero layout. The range of drive from the circuit is nuts. You really have created almost everything for everyone. Great work.

[Frank_NH]

In that respect, since the available choices increase when using duals, as opposed to singles, building it with a layout for using one half of a dual may be worthwhile.  

I was wondering, Mark - would this circuit benefit from a unity gain buffer at the output?  Perhaps something that the second half of a dual op amp could be used for.  Or I suppose you could use a jfet buffer and tack it on at the end, and still use a TL071.

[Mark Hammer]

Or a TL072....

I don't know that there is a need for an output buffer, but on the other hand I don't have any sense that it would suffer from one.  And since it can be set to relatively clean, and has a modicum of tone-shaping, sticking a buffer/gain stage (say gain = 3-4x) after the volume pot wouldn't be a terrible thing.  And if it expands the possibilities for chip choice, why the heck not?

Myself, I was aiming for something as simple and stripped down as I could get.  But if complicating it a smidgen opens up more tailoring possibilities for people, have at it.

[bluebunny]

You really have created almost everything for everyone.

Almost.   

Great work.

+1

[lars-musik]

What a great design. Thank you very much Mark.

I think this might become my favourite overdrive ever. I really like the overdrive section of the Box of Rock and I own an early 80s OD-1, of course I build myself a TS (ah – and a Blues Driver), but this one here is more flexible and just sounding better than all of them. Unfortunately I already etched a King of Tone board, so that must follow, too. But the AEFEA: Do build it!  The "Everything I Need - EIN Drive" would have fitted better. Anyway, Mark designed it and he named it, so I stuck to it.

Here are some pictures of my first (maybe not the last) build. I put a dual op-amp in there, because the drawer looked fuller than the one of the  single op-amps. The fullest box contained 4558s (when I began building pedals I still thought that is the only op-amp for overdrives so I ordered too many of them), so this was it.  Maybe others sound better, but for me there's no need at the moment. I think I'll have to fiddle about with the switchable 4n7 – the sound is too brittle for my taste.



For all the 1590a fans, here's a layout.



Best, Lars

[bluebunny]

Well, a year and a half after I said I'd build this, I finally finished it!



Did a perf layout with board-mounted pots.  I have a short, because the volume pot doesn't do anything!  So I'll have to take the board out and fix that, but otherwise it was worth the wait.  Thanks, Mark.

[Mark Hammer]

Hey, thank YOU.

I'm always honoured when someone thinks that it's worth their while to invest their time and money in something that came to me on a whim.  I hope you enjoy it for a long time.

And yes, the 4n7 cap/path is a little shrill.  It may be a good idea to stick a resistor in series with it just to tone down its contribution to the mix.  Maybe try 6k8 to 12k.

[roseblood11]

I'm just planning a two-in-one pedal with the AEFEA and a distortion pedal, maybe somethings with a Marshall-like sound.
Did anybody try how the AEFEA stacks with other drive pedals? I might use the Catalinbread Dirty Little Secret (Mk1), but that's quite bright as well. And I use a AC30 Top Boost, which really needs drive pedals with the ability to tame the treble...
I'll use a varibale relay switching, that allows for both popular wirings of two footswitches (a) each circuit has its individual bypass switch, b) one switch toggles from circuit 1 to circuit 2, the other bypasses the whole pedal.)

[Mark Hammer]

I think it should probably work out nice if one uses the "middle" toggle setting (variable lowpass with no treble bypass) and sets the unit for soft clipping.

[bluebunny]

. . . something that came to me on a whim.

Nice whim.   

[cab42]

I finally got around to finish my AEFEA. What a great pedal! Thanks Mark.

It's been quite a long process. I started AEFEAing my Tube Screamer by replacing the original feedback loop with the aefea's on a daughter board. A recommended mod for a tube screamer. Unfortunately I pretty much ruined the vero board with broken and lifted pads when experimenting with the tone stack. So I decided to ditch the ts and build an aefea from scratch. But it was a fun process.

I just have one question: I have never built or played a Rat but I like the rattier sounds of the aefea. I saw the Rat has a pair of diodes to ground and I tried to add them to the aefea, but it really didn't make much of a difference. Slightly lower volume and slightly different sound. Is that expected behavior? I had expected something more dramatic ie. more drive. I tagged the diodes to pin 1 of the tone pot.

[Mark Hammer]

Thanks for the positive feedback.  I appreciate it.

The problem with your attempt to Rat-ify the pedal is that you still have the feedback diodes in place.  So when you increase the gain to get clipping out of the additional diodes, the unit is already clipping.  Because there is an asymmetrical trio of diodes in the feedback loop, and you're adding a back-to-back pair on top of that, there likely won't be much of a noticeable increase in clipping, but there will be a drop in volume because a 2+1 diode complement results in a roughly 1/3 higher output than a 1+1 pair.  The second pair of diodes reduces that 1/3.

That said, if you adjust the 500k clipping-hardness pot so that you end up with the 22k in series with the diodes, you will reduce the degree of clipping in the op-amp, via the feedback path, and provide a hotter output to your additional diodes.  Under those conditions, your added pair will be responsible for more of whatever clipping you hear.

Alternatively, just use a toggle to disconnect the feedback diodes from the output of the op-amp.

I understand why you were prompted to consider a Rat-like arrangement.  Much like the Rat, this circuit uses two "tails" to provide differential gain for the bass/mids and highs.  Of course, where the Rat has gain in the thousands, this one doesn't apply more than 230x gain for the highs; more than a TS-9, but clearly not even on the same continent as the Rat.