Messages

I've found breadboarding to be a good corrective for hoarding. There are many circuits that seem exciting...until I breadboard them and get to play with them a little.

Let me know when you have extra boards to sell, I am definitely interested

I sent a DM. Reply here if you don't get it.

Thanks, guys. Your kind words make it easier for me to trust my own instincts.

Six at once?! How did you keep track of everything?

[Mix]

So you have it set up so that when it's in bypass, you're using the dry signal?

Yes: when in bypass the dry signal still hits the input and output buffers. R8 helps to reduce popping when the effect is engaged.

The delay signal is bled to ground through the mosfet, which is acting as a switch. When the effect is engaged, there's no voltage at the gate, so the delay signal sees a resistance to ground of 2 or 3 mega ohms through the drain to the source. When the effect is bypassed, voltage builds up at the gate, which allows current to flow from the drain out the source to ground. The resistance drops to tens of ohms. All that mess between switch 1B and the mosfet gate controls how fast the resistance changes and thus how quickly the Mix and Repeats get turned down. With Tails fully counterclockwise, it's basically instantaneous. Fully clockwise, it's about 12 seconds of tails time.

 was always very happy with the sound of the original version and enjoyed the tails functionality; however, tails combined with the minimal filtering approach that this delay takes meant that there was always more noise from the PT2399 in bypass than I could tolerate.



The enclosure is Aqua Vein from stompboxparts.com. It's beautiful, but difficult to photograph. I used the same cheeky graphics from the original version.



I'm starting to get my PCB layout conventions figured out and I think it shows in this build.



Version 2 implements a slow-opening mosfet switch which allows a "tails time" to be set before the delay signal is turned off. This maintains the tails bypass while eliminating the added noise. Additionally, the input and output sections were streamlined and filtering was increased slightly. The modulation section was revamped to allow for a much greater range and to eliminate the need for a switch. It is essentially the same technology I implemented in the Special Sauce and Trogodyte Reverb.

Here's a demo:

I've got an extra PCB. If you're interested, just PM me. If there's enough interest I can do another run.

I feel like there was a similar panic among studio musicians in the 80's when synths and drum machines hit the scene. Yes; some people did loose their jobs in music, but new jobs were created. We're in the "naive optimism/existential terror" phase of a new technology where the popular belief is that it will "revolutionize the field". Creative and forward thinking people are beginning to explore the possibilities. I imagine in ten years, we'll treat AI like we do synths or drum machines; they are useful in some contexts and not in others. We have synths, but people still hire orchestras. We have drum machines, but many drummers are still employed.


...and honestly, there hasn't been any money in music for decades. People keep making music as a means of self expression, even if they can't sell it afterwards. At worst, commercial music will die, not music in general.

Did you mill your own slots? Looks super clean!

Brilliant idea and really straightforward in its design and application. I love delays with any kind of filtered repeats, like wah, phase, flange, etc. Excellent job!

Thanks! Are you also a fan of Flerb?

Am I remembering correctly that you want reverb to swell in on top of overdrive?

dude, you're absolutely slaying it. Hitting homers one after the other!

That's very kind. My "batting average" goes way down if you include all the hair-brained ideas that never make it past the bread board!

Maybe you could work in parallel instead of series? I've attached a block diagram for what I think might work. Blue arrows represent signal flow. Red represent control signals. When the envelope detector was triggered you want to turn volume 1 down and volume 2 up.

[Frequency]

At its heart, Ouroboros is a PT2399 delay with an auto-wah in the feedback loop. This merits an explanation. PT2399 delays require aggressive filtering to limit the noise and distortion they produce, especially at longer delay times. Some people like this because it keeps the repeats out of the way of the dry signal, but it can result in repeats that are so dark that they becomes indistinct, and get lost in a mix. I wanted a delay that was well filtered to minimize distortion, but would still have clear, distinct repeats. I needed an aggressive filter that also stood out in the mix—what is a wah if not an aggressive filter that cuts through the mix?! Tap tempo and an LFO to sweep the filter were added to fill out the feature set.



The enclosure is tango green from stompboxparts.com. It took a long time to settle on a control layout that was both logical and aesthetically pleasing.



I used pretty much all the available space inside the enclosure, but tried to situate things so that they would be easy to place and solder. There is a specific order of operations that works best.



An input buffer feeds the PT2399 input which uses a pretty standard filtering arrangement. The extra CMOS stage is set up as a multi-feedback band-pass filter. Basically, it's an inductorless wah. The center frequency of the filter is set with the Frequency control. Width toggles between two capacitors in the filter. Up is very narrow. Down is broader, but still pretty aggressive. The filter can also be controlled via LFO. Rate ranges from a super-slow 15 second cycle all the way up into ray gun territory. Depth can sweep the full range of the filter or be set for a subtle glimmering effect. It works in conjunction with Frequency to set the range of the sweep: Depth sets the lowest frequency of the sweep and Frequency they highest.

Tap tempo is implemented using Elecric Canary's Bontempo chipset. This is an excellent tap tempo solution. It can be calibrated to your specific PT2399, it requires minimal extra components, and it has a very robust suite of modulation controls that I didn't even use in this circuit. Best of all, the EEPROM is open source, so you can flash your own micro controller if you want!

My copy is capable of 1300ms of delay—I timed it! This could vary a little between delay chips. An internal trimmer can be used to adjust the onset of self oscillation. Between the filter and output stage, the delay signal can be boosted about 13dB. This means there's enough gain to get the delays above a hot input signal, despite the limited headroom of the PT2399.

I know a wah inside the feedback loop of a delay sounds ridiculous, but it accomplishes the nearly impossible task of producing clear cutting repeats that also stay out of the way of the dry signal.

Here's a demo:

As always, I have a few extra PCBs. If anyone is interested, just DM me.

I did some YouTube watching. Seems like with this, I wouldnt actually need an envelope detector up front? IF it would just be on / off

Actually, NorthCoast makes a good point; if you set up a comparator, you could feed the signal you want into it and set the threshold by adjusting the bias voltage. Or you could set the bias voltage high, and use a boost in front of the signal to get it above the threshold. In either case, you'd need to at least partially rectify the signal hitting the comparator.

If you connected the output of the comparator to the base of a BJT, wet signal output to the collector and the emmitter to ground, you could use it as a switch. When the input signal goes high, the comparator output will go low, causing the resistance between the collector and emmitter to increase and allowing the wet signal to pass. When the input signal goes low, the comparator will go high, increasing the voltage at the BJT's base, which will in turn allow current to flow from the collector to the emmitter, lowering the resistance to ground and muting the signal.

...in theory, at least—the devil is in the details!

I don't know exactly how you'd do that, but I'd start by looking at compressor and noise gate circuits. At the extremes, you'd get silence. I imagine you could use that kind of envelope detection to control an LDR or a mosfet to act as a switch.