Meatball, Mutron, Qi-Gon Questions & Brainstorming

[auden100]

I'm taking a plunge into the Meatball Envelope Filter. I've been collecting ideas and schematics while I wait for parts, and I have a couple of questions in hope of help from you all.

1) In the filter section, there is a 2-pole switch hooked up to change the frequency range of the filters. Normally the caps on those poles would be roughly equal. What happens if you separate the two poles and make two switches or use non-equal values for caps?



2) I was looking through an article on state variable filters, and I read that with the addition of a fourth op amp one can make a notch filter. I'm not familiar with any guitar notch envelope filters. Is it useful? Anyone know a pedal I could look up a sound sample on? Would that work with the three inverting opamps as presented in the original circuit? Or would I need to alter the inverting lineup to compensate?
Filter article:
https://www.electronics-tutorials.ws/filter/state-variable-filter.html?fbclid=IwAR2jg6qlNBrt5HzVjXoqj6zc5fA0aPLHvprkm_sWxvK_M0CksUlLlmEYA10

Meat Sphere version from MusicPCB:
http://www.musicpcb.com/pcbs/meat-sphere-filter

As for mods (I'm planning big and working on perfboard, so I expect some pruning might be wise), I'm shooting for:
- Color Mod: tie legs 2 and 3 together on the color pot.
- Inverting Opamp: add a switchable inverting opamp to filters for inversion compensation.
- Qi-Gon Version: makes several changes, including adding an LFO. This is a low priority mod for me.
- Expression Pedal Jack
- Replacing two rotary switches with a on-off-on SPDT (Bandwidth) and a DPDT (Up/Down).

If anyone has extra mods to recommend or input on what I've listed, lay 'em on me.

[iainpunk]

1)
you lower the Q factor of the filter, this means less available resonance and flatter slopes (Q factor stands for Quality factor)

2)
famous pedals use notches that are frequency modulated up and down by an LFO, they are called Phases. the Phase 45 is a single, very narrow notch, the phase 90 has two notches that are about two an a half octave apart.

i wouldn't worry about phase flipping, since the filters will be changing the phase anyways, DC phase becomes irrelevant for the most part.

cheers, Iain

[auden100]

Thanks, Iainpunk. Very clear. I knew phasers did that, but I had assumed they all created a whole array of notches, not just one as in the P45.

Conclusion:
- Keep the cap switching as is. (100% answered. Cheers)
- Notch envelope filter is conceivably worthwhile. Perhaps more so with the Qi-Gon additions.
- Confusion on the inverting opamp. I have read other modders state the inverting opamp mod evens out some volume discrepancies between the filters as they blend into the original signal. You are saying the it won't make a difference, or perhaps it wasn't clear the filtered signals would be blended into the original.

[idy]

If you re-read the thread,  people are saying that one of the positions, LP is inverted in phase and when blended with the clean signal some find it weak, it is canceling out the dry. (I only use the band pass so it doesn't "phase" me...)

Looking at the schematic, and then reading a little about state variable filters, you will find that both HP and LP are inverted, so you would maybe try inverting both of them. The fact that the people talking about this don't notice that suggests maybe that people are "spit-balling" and not responding to a real problem. As the more knowledgeable  poster above pointed out the filter is really changing phase all over and so the cancellation is not as obvious (or undesirable) as you might fear...maybe. But a phase toggle makes sense, I would be curious to hear bandpass inverted to make a notch when mixed back in. So one op-amp with a toggle after the mode selector.

To me the four positions for filter range seem like: medium, high, very high, ridiculously high. An experimenter might try different values for those caps, but probably keep the pairs the same, because the beauty of the filter is that it can get focused. To make it less intense you can: decrease sensitivity, or intensity, or resonance, or blend. Or use the 'bandwidth' at 1/2. Heck even using slower attack makes it less extreme. But it you like experimenting it would not be hard to have the caps on the rotary or a daughter board and socket them.

By far my favorite addition is to use a buffer and a switch to select either pre or post effects loop for the dry side of the blend. It seems a little insipid to mix a wild fuzz auto-wah sound with a clean guitar, but mixing it with original fuzz really makes it "pop" and come alive. The post loop side needs the buffer.

[iainpunk]

it wasn't clear the filtered signals would be blended into the original.

yes. this
i'd add this to the output of the filter, before mixing with the clean again.


there are 3 fx loop points id considder, before the filter, after the filter, the clean signal (parallel to the filter) but keep in mind that most filters work best with a clean signal's envelope and a distorted signal might keep the filter at the same place, instead of sweeping.

cheers, Iain

[ElectricDruid]

One other reason that you might not notice a highpass filter being out of phase so much is that most of the fundamental will be removed, so the place where it would be most obvious has also been removed. You can't have a sound that sounds a bit thin if it's already a bit thin!

[auden100]

@Iain, Thanks for the idea. I'll have to compare the separate inverter addition you've described with the Qi-Gon variation that just changes the first opamp to non-inverting. From what Electric Druid is saying, it might be useful enough, and it would spare a couple opamps to still get a notch filter from the summed low and high passes and a buffer on the post-fx mix.

@idy, I'll probably socket the caps or use a 6-pole switch for a couple extra options. Also, the post-fx blend with buffer seems like a sharp addition.

@ElectricDruid, tastes pretty truthy to me.   But actually, that some people liked an inverter on both filters and some only inverted the low-pass did confuse me a bit. That's good insight. Once this thing is breadboarded, I'll know for certain.

[auden100]

Just wanted to put a quick note in here for posterity.

COLOUR MOD MOD
There's been some interest in the Colour mod, tying together a couple legs of the pot. I've got a version of the Meatball on my breadboard (can't get my MC1458 to work, so I'm using the Qi-Gon Y-trigger with a TL072). It seems like the Colour mod works fine, but I prefer instead no mod and using an anti-log C pot. I happened to have one, and it made a very handy difference in finding the right balance in resonance.

The volume problems with the inverted filter signal is confirmed. As expected, the volume decreases as the blend knob approaches halfway point and is normal volume at either extreme. I think it'll be worthwhile to implement the inverting buffer.

One characteristic I'm trying to compensate for the the curve of the filter movement on decay. It's kind of like a logarithmic downward slope. Starts slow and accelerates down. Playing with the decay control slows the slow part, but doesn't affect the fast descent much. I'm not remotely knowledgeable enough to do more than trade put a few components and hope. So it'll probably just stay that way. It's all still quite useful as is.

The Qi-Gon Y-trigger has worked great in lew of having a working alternative. The Sweep control plays an interesting game with the Intensity control, and it does take a bit to find the pocket. But if you don't like fiddling with knobs, then you shouldn't be building one of these anyway.

[idy]

If you go back to the "building the meatball" thread some bright people explained why the 1458 is such a dicey choice for the trigger, and the fix: adding just shy of 2v bias to its input.

The mutron is bipolar power so bias to ground is okay, but the meatball is single side power but keeps the ground....and the 1458 doesn't play nice close to the - rail.

Without that I found I needed a booster on the guitar to make it trigger easily.

[auden100]

Thanks for the heads up. I'd looked a bit through that thread, but you've clued me in to your nice little summary post. Very helpful.
Post here: https://www.diystompboxes.com/smfforum/index.php?topic=92894.msg1182648#msg1182648

I like that the Qi-Gon has the LFO option integrated. The pedal is so big already that not having need of an additional auto-wah style effect might be helpful. The prime disadvantage IMHO being that given the high interactivity of controls and fine range of favorable sounds, any on-stage fiddling with this effect is not ideal. So many options. I'll keep fiddling.

[auden100]

Qi-Gon Schematic


I received permission to repost this schematic today. So here it is.

Two questions arose. I have it built without the LFO. Everything seems to work fine. But some characteristics appeared that I'm not sure if it's typical or something I need to troubleshoot more.

- I get light distortion from the filter signal. Not from the clean signal. Is that typical? I'm using store-bought vactrols. The indicator LED flashes at the rate of the distortion. I'm not sure what is causing this. Any ideas? I've had the same distortion while using both TL072 and a 4558 on the envelope follower, and it seems like there's some kind of rippling happening in the LEDs that I can't contain.

- The Moog mod doesn't make a positive difference. I tried implementing the Moog mod with the Qi-Gon's non-inverting lowpass filter. It just doesn't sound like it works the same as with the Meatball. It seems to just make the filter less pronounced. Does that seem right? The envelope section works fine with both the original and Qi-Gon versions of the filter. But the moog mod isn't quite amazing enough to warrant the added re-inverting stage that I liked on the Meatball.

MOD ALERT!
I have one tiny mod to the Qi-Gon that was nice. The Sweep knob is pretty cool. As originally described, when it is at the centerpoint, there should be no sweep at all in the LEDs. That was true with the 4558, but not the TL072. I found that I just had to put a trimmer on one side of the Sweep pot, and then I could adjust the location non-sweeping point on the pot accordingly.

[auden100]

Another update.

After trying and comparing the different trigger strategies for the Meatball and Qi-Gon, I've decided to stick with the Qi-Gon. For the Meatball, I tried several ICs and changes, but nothing was quite right. The Qi-Gon has more variability in control of the envelope and as such requires more noodling to get what you want. But I can find just what I want, which became lacking as I switched back to the Meatball version.

Mod Alert:
Clipping Mod: Using the Qi-Gon schematic, I placed a 100kA pot in series with the 10k resistor (R13). Then I place a on/off-switchable silicon diode pair in parallel with the pot. When the pot is a 0, the sound is stock. As the pot increases (diodes off), there is a boost in the low end of the sound. I like the guttural quality of this sound a lot, and it can be subtle or pronounced. With the diodes on, there is clipping introduced. It's not a perfect-sounding clipping, but for the amount of parts and work, it's worth it for me. The issue at the moment is when in Low-Pass mode, this gain creates a boost. That's not ideal, and I'm wondering if it would be worthwhile to include a booster/attenuator at the end of the circuit. This could help because I think the signal is a bit low when some other filter modes are strongly set. This Clipping Mod is grew from the Moog Mod, but I found that even without the capacitor, and just a bit of adjustment, this gain pot gave me the low end bump I liked. Any thoughts on this Clipping Mod that are helpful are appreciated.

Extra note on the Qi-Gon. I'm not yet a fan of the Tremolo mode. Could be personal preference, or it could be my crappy breadboard screwing with the connections. I'll be using at least 6 different cap settings in the filter range, and the tremolo doesn't sound nice enough for the several euros for a 3PDT switch (IMHO).