Adding a Fuzz to an MXR Envelope Filter? Possible?

[Freekmagnet]

I built a BassBalls clone a couple months ago, I'm really liking it. I really like the fuzz aspect of it. Would it be possible to add a little fuzz circuit to the TonePad MXR Envelope Filter? If so, how would I approach it?

[Mark Hammer]

Not really the same animal.  The Bass Balls exploits the envelope follower section as its "fuzz".  The EF adopts a different chip for its follower, so you won't get that sound out of it.

[Freekmagnet]

Well, I built an Ampeg Scrambler and placed it before the MXR EF. Really does the trick nicely. Blending the fuzz with the clean signal gives the filter plenty to chew on and it really comes alive!

[Mark Hammer]

Ideally, one does the envelope-detection *before* the signal gets compressed.... which will inevitably be the case when a distortion pedal is used.  This is why pedals like the Q-Tron+ and Meatball incorporate a send/return loop; so that the envelope follower taps the clean signal, but what gets filtered  is the distorted signal.

[Freekmagnet]

So in other words, I might try putting the fuzz after the filter? I'll check it out. It sounds pretty cool before the filter. I blend just a little fuzz with the clean signal. Produces a nice, vocal sound that's pretty responsive.

[nocentelli]

So in other words, I might try putting the fuzz after the filter? 

No, if the fuzz is to be filtered, it must come before the filter section: The meatball has a buffer first, and the buffered signal is sent separately to the envelope section AND the filter section: The loop is just an insert before the filter section but after the buffer. For the MXR EF, you could add a simple buffer to the front end, and split the buffered output signal to a) the envelope detector section and b) a switch that bypasses/engages the fuzz: The output of that switch would then go on to the filter section.

[Mark Hammer]

The filtering effect will be more pronounced if:
1) the signal you feed it is harmonically rich;
2) there is enough dynamic range in the signal to produce a wide sweep.

If you stick the fuzz before the filter, you lose #2, even though you keep #1.  If you stick if after the filter, you keep #2, but you forfeit #1.

HOWEVER, if you use the pre-fuzz signal to force the sweep, AND you use that sweep to filter the fuzz, you get the best of both worlds.  THAT's why the loop is there in better pedals.

[Freekmagnet]

Here's another question about the MXR EF:

First, let me say that I recently acquired the newer MXR Bass EF. It's a great filter and definitely delivers the goods. Having the onboard blend control is amazing, and the sweep sounds really good, and it's nicely responsive and funky.

But compared to the Tonepad MXR EF clone I built several years ago, with the range mod added (I think I added an addition 470pf cap or something like that), I gotta say that the low end of sweep that my old MXR has eats the new MXR Bass EF for breakfast.

However new MXR has this gorgeous top end of it's weep, which really shines above my old MXR. Oddly enough, when I turned the mix of the instrument down, I realized that the newer unit is really dependent upon the dry/wet mix for it's low end.

So, is there a way to add more top end, or rather, widen the range of the sweep on the old MXR while retaining the low-end awesomeness it has already?

[Mark Hammer]

It may be that this is not possible.  The old EF uses PWM to vary the time that a pair of switches remain open or closed.  This simulates the change from a high to a low resistance. in a very predictable way, without needing to match components.

But there are limits to how long one can leave a switch open or closed without producing audible degradation of sound.  In other words, it is a great trick within a limited set of conditions, and outside those conditions, it won't work so well.

I honestly can't say whether the EF exploits that PWM trick to the fullest possible extent or not.  But because of the underlying "engine" of the pedal, prepare yourself for the possibility that you won't be able to squeeze wider range out of it via a few simple changes.

As well, note that, because it uses a clock to do the switching, it needs to use some lowpass filtering on the output to remove the clock noise from the audio.  That's what the 100nf cap on the output is for.  You can try reducing the value a bit and see if that gets you a little more usable top end.

[Freekmagnet]

Figured as much. It's interesting because both pedals sound very similar - similar enough that I'd ask myself, "do I need to keep both of these?" Of course I favor the pedal I built myself however, the newer MXR is in a much smaller enclosure, which helps a lot as far as real estate is concerned. I don't have plans to get rid of either at this point. Really what I should do is sit down one day and figure out how to put an effects loop in the old pedal. That would pretty much solidify its place as the superior pedal of the two.

I'm going through a another similar debate with myself right now - I can't decide if I should build a Mutron III clone or a Meat Sphere. Similar pedals, but one has crazy capabilities while the other is smaller and probably much easier to dial in. Ah, decisions.

[Freekmagnet]

So, looking at my Agua Filter schematic here:

http://www.tonepad.com/getFile.asp?id=113

it seems pretty clear that the signal splits at the junction of R8 and R7 with R8 going into what looks like the envelope detection circuit. I'm only able to guess this because I read your auto-wah paper, Mr. Hammer, and seeing that the signal path passes through the threshold control, I'm assuming that this would be where the input from the instrument is controlled before going into the envelope detector.

This, I assume, would leave the path leading from R7 as the filter section, with R7 attenuating the signal going into the filter. This being said, if I were to build an effects loop into this circuit, would it be as simple as say, sticking an AMZ Jfet Splitter (http://www.muzique.com/lab/splitter.htm) before R7, with one lead going into R7 and another going to an output jack (send) and an input jack (return) that also feeds into R7?

[Mark Hammer]

Pretty much.  You'd want to put a DC-blocking cap before R8.  C3 is doing DC-blocking duties for both the filter and envelope-follower sections.  You'll want to make sure that both paths keep DC out.

If you wanted something fuzzierfeeding the filter, then you might want to take a dual op-amp, and use one op-amp for an input buffer.  That input buffer would have two outputs with one feeding the recitfier/follower (output of buffer goes to a cap then R8).  The other output would feed a second op-amp configured as a clipping stage.  I'll suggest something sort of like the TubeScreamer clipping stage, only using Schottky diodes instead of silicon, so you don't have to apply too much gain and don't have to attenuate much in order to feed that output to the filter.

I can't draw it out now, because it's late, but I'll see what I can do tomorrow or the day after.  But what I describe is pretty much what you want.

[Freekmagnet]

Pretty much.  You'd want to put a DC-blocking cap before R8.  C3 is doing DC-blocking duties for both the filter and envelope-follower sections.  You'll want to make sure that both paths keep DC out.

If you wanted something fuzzierfeeding the filter, then you might want to take a dual op-amp, and use one op-amp for an input buffer.  That input buffer would have two outputs with one feeding the recitfier/follower (output of buffer goes to a cap then R8).  The other output would feed a second op-amp configured as a clipping stage.  I'll suggest something sort of like the TubeScreamer clipping stage, only using Schottky diodes instead of silicon, so you don't have to apply too much gain and don't have to attenuate much in order to feed that output to the filter.

I can't draw it out now, because it's late, but I'll see what I can do tomorrow or the day after.  But what I describe is pretty much what you want.

Gosh, that'd be awful nice of you.  I barely know what's going on and any guidance would be helpful.

Sounds like what you're describing is hard wiring a distortion/fuzz/overdrive stage into the pedal, which I think would be good, actually. I wouldn't have to drill new holes in the enclosure for the extra jacks, and I already have an extra switch installed that really doesn't need to be there. I could use it to activate the fuzz. Cheaper and easier.

So this DC blocking cap before R8 - would another .05uf cap be adequate for this purpose?

[Freekmagnet]

So looking at this Klon Buffer here (http://www.beavisaudio.com/techpages/Buffers/Buffer_clon.gif) I can see that only half of the pins are being used, and the same goes for this TS according to this schematic (http://www.geofex.com/article_folders/tstech/tsxtech.gif). So what you're saying is, use half of the the op amp for the buffer and the other half for the TS clipping stage, which makes pretty good sense to me. But, what I don't quite understand is this: is it copacetic to simply wire two leads coming straight from the output of the buffer with one going into R8 and the other going into R7? And would I need to build an output buffer like on the Tube Screamer?

Without checking my back stock, I know I have a bunch of TL072's laying around.

Don't let me stop you from drawing it out, BTW.

[Mark Hammer]

You're on the right track.  I still suggest use of a pair of Schottky diodes (e.g., BAT41, BAT46, 1N5817, etc.) instead of silicon for the clipping stage.  That way you can use 10k instead of 51k, and a 250k pot instead of 500k.  You should get the same clipping, but with a lower output level.

A drawing from my end will have to wait a bit.  Too many bloody errands to run today.

[Freekmagnet]

To add the fuzz to the MXR EF, would it be something like this?



I'm really guessing. Kinda out of my comfort zone here. I've never really hacked two circuits into one before.

Thanks