Subdecay - Octasynth

[Liquids]

I don't see any threads about this one yet. It seems like at it's core it is likely a CD4024 for the octave, and probably an OTA (13x00) type for the resonant filter; that's my guess. But the finer details are what I'm curious about!

Here's a good youtube user video:
The manual online shows a low-res guts picture, note the SMD chips: http://subdecay.com/manuals/os2.pdf

Maybe this uses the general EHX Microsynth/Boss OC-2 tracking for octaves, but I know that Brian is a smart guy, and it's not inconceivable that he has come up with some improvements or tweaks for tracking, maybe added a compressor; I'd be interested to see.

These do seem hard to find used, maybe they don't sell all that well, or maybe people who buy them keep them? Does anyone have one and can take some gutshots?

[~arph]

At some points in that demo you can hear a frequency slide/locking effect.. leads me to believe there is a 4046 PLL at the heart with a divider in the phase comparator. The three squarewaves can be taken from that divder (and yes a 4024 is a likely candidate).
The filter sounds nice though... kinda like that famous korg filter.

[Liquids]

Interesting, I will have to listen for that.
If it is a PLL, I'm all them more interested.
I saw Brian Marshall posted elsewhere that the filter cutoff more or less 'sits at a sub-audio frequency' with no input signal and returns there as note fades to make things sound smoother, though of course there is resonance involved...makes sense to me! That's certainly one way to do it!

[~arph]

That could also be what I hear. It seems more obvious in the demo vid on subdecay's site. It certainly has a nice attack. I wonder what happens if you play serveral notes in quick succession without muting the sound. This way we can find out whether it's envelope controlled or not. (I may need to listen to the samples better)

[Brian M]

Hi guys,

The octasynth uses a CD4013, similar to the blue box. The prefiltering and comparitor prior to the flipflop circuit are different/better though.

The filter is LM13700 based.

I tried using several other IC's for the octave including the CD4047, and a couple counter IC's that I have around. All of these ended up either having problems or being more complicated than the small footprint allowed for.

This basic idea goes back to a protpype I built in 2006. Over the years many variations of it were worked on. Most of the new products we've released in the last two years have been a ton of work, with code and programming being needed, loads of prototypes, and redesign after redesign. At the time I kinda needed some "instant gratification" in the design process. I basically gathered all my notes from prior work on octave and filter effects and went from placing parts on a breadboard to finishing the PCB artwork in a matter of hours. I wish all projects went like that.

[Liquids]

Thanks for the info Brian. Agreed that the Blue Box filtering and squaring are not worth duplicating.

Did you use something similar to the Boss OC-2/Microsynth scheme for prefiltering and comparator into the 4013? OR another route such as the EBS Octobass?

Please do not be offended - I greatly respect your company, and don't want you to share more than you are comfortable with. But, if indeed there is no harm in asking - would you be willing to provide a schematic, or even portions of the schematic (prefiltering, resonant filtering)? Or even just high resolution gut shots, so we can try to deduce more of the inner workings for ourselves if you'd rather not hand it over on a silver platter?

Out of curiosity, do these pedals sell well for you or are they more the occasional novelty sell to those who are going beyond acquiring or cycling through an encyclopedic fuzzbox collection?

Again - I hope you are not insulted by my asking these questions and if so, I do apologize.

Matthew

[Brian M]

Thanks for the info Brian. Agreed that the Blue Box filtering and squaring are not worth duplicating.

The bluebox is actually pretty cool, but glitchy. Your better off embracing the glitchyness than wishing it did something else.

Did you use something similar to the Boss OC-2/Microsynth scheme for prefiltering and comparator into the 4013? OR another route such as the EBS Octobass?

I don't know anything about either of those pedals.

Please do not be offended - I greatly respect your company, and don't want you to share more than you are comfortable with. But, if indeed there is no harm in asking - would you be willing to provide a schematic, or even portions of the schematic (prefiltering, resonant filtering)? Or even just high resolution gut shots, so we can try to deduce more of the inner workings for ourselves if you'd rather not hand it over on a silver platter?

I won't share a schematic, but I don't mind sharing how things work. I don't think anyone has ever traced one of our pedals here... still waiting....
I'll add that the filter section is based on the LP/BP filter in the 13700 datasheet. I just added some changes to feedback for resonance.

Out of curiosity, do these pedals sell well for you or are they more the occasional novelty sell to those who are going beyond acquiring or cycling through an encyclopedic fuzzbox collection?

To be honest I expected this pedal to be a slow seller. I know synthy pedals aren't for everyone. That's the biggest reason that I didn't turn it in to a product 5-6 years ago. Given it's simplicity, I do think the controls and functionality it is pretty approachable. It's pretty surprising to me, but since it came out, I think we've sold more of them than anything other than the proteus.

Again - I hope you are not insulted by my asking these questions and if so, I do apologize.
Matthew

No worries. I'll talk shop with you guys when I have time.

I'll also add that I really did not want to use the CD4013. Sometimes we spend a lot of time trying to reinvent the wheel, but in this case none of those ideas made any practical sense.
I really wanted to use the CD4047, because we already use it in the noisebox, and it would be one less part to stock. There seems to be some inconsistencies with the internal logic diagram for that IC and what it actually does. I guess that happens sometimes when you try to use something in a way they design didn't have in mind.

[culturejam]

I don't think anyone has ever traced one of our pedals here... still waiting....

And probably nobody will (need to) if you keep up with your willingness to share design details.

I don't need to know *exactly* every single connection in a cool pedal. Knowing the big-idea functional blocks and how they interact is plenty enough. And if the designer is giving you that information, the need to do a full scorched-earth trace diminishes greatly.

[Liquids]

Did you use something similar to the Boss OC-2/Microsynth scheme for prefiltering and comparator into the 4013? OR another route such as the EBS Octobass?

I don't know anything about either of those pedals.

Those circuits basically use an gain of 1-20, which is followed by a 3-pole filter (passive r-c plus a 2-pole active filter) around a buffering op amp stage in order to get a LPF with a sharp cutoff that drops sharply after 700hz or so, in order to get a 'simple' and functional tracking/fundamental extraction as far as pre-filtering the harmonic content of a guitar signal.

I'll add that the filter section is based on the LP/BP filter in the 13700 datasheet. I just added some changes to feedback for resonance.

Yeah, it's hard to go wrong with utilizing the available OTAs for a resonant filter. Minimal additional components needed, great sound, easy to work with, hard to blow up, built in buffering etc, etc, etc. Its an oft-used chip for such a purpose for those reasons not in mass-produced and DIY resonant filtering alike. There are few reasons to make it too much more complicated than that (IMO), unless you are Moog Inc, or target market is a significant number of people with $$$ and a lust for a discrete, mojo, diode-ladder style resonant filter circuit.

To be honest I expected this pedal to be a slow seller. I know synthy pedals aren't for everyone. That's the biggest reason that I didn't turn it in to a product 5-6 years ago. Given it's simplicity, I do think the controls and functionality it is pretty approachable. It's pretty surprising to me, but since it came out, I think we've sold more of them than anything other than the proteus.

Awesome! I'm pretty surprised that the proteus is a big seller for you....I can't tell you how glad I am to hear that somehow you've avoided (for the time) paying the bills with an OD and everything else is specialty. Of note, my local dealer said he is out of the OctaSynth!

I'll also add that I really did not want to use the CD4013. Sometimes we spend a lot of time trying to reinvent the wheel, but in this case none of those ideas made any practical sense.
I really wanted to use the CD4047, because we already use it in the noisebox, and it would be one less part to stock. There seems to be some inconsistencies with the internal logic diagram for that IC and what it actually does. I guess that happens sometimes when you try to use something in a way they design didn't have in mind.

That's funny to me, because it seems like the CD4047 thing came about from a typo in a mouser order, if I remember your post correctly! Haha. Thomas Henry is an amazing reference/resource/consultant when it comes to going beyond what the fabricators reveal and dare I say even understands about manipulating specialty chips. I wonder if he would be able to sort some of that out for you.

I don't think anyone has ever traced one of our pedals here... still waiting....

I have noticed that! Is that a challenge? Even if it were, while I have some money burning a hole in my pocket right now plus a fender champ to sell off, alas, my homebrew hobby has me focused on spending cash on a refrigerator instead of music gear!

I won't share a schematic, but I don't mind sharing how things work.

Cool....well than, here's some things I'm curious about, and if you're willing to answer them, all the nicer, if not, oh well. :
do you utilize passive or active filtering for the pre-octave fundamental extraction?
Does an OTA (apart from the resonant filtering after the sub-octave) do any of the pre-filtering?
Do you feed the 4013 with a simple distorted/rectified square wave, an op-amp configured as a comparator, or a real comparator?

And hey, I'm relly not trying to blow smoke up your other end here, but thanks for being a generous with your circuit knowledge and experiences.

[Brian M]

Cool....well than, here's some things I'm curious about, and if you're willing to answer them, all the nicer, if not, oh well. :
do you utilize passive or active filtering for the pre-octave fundamental extraction?
Does an OTA (apart from the resonant filtering after the sub-octave) do any of the pre-filtering?
Do you feed the 4013 with a simple distorted/rectified square wave, an op-amp configured as a comparator, or a real comparator?

The OTA is only used for the filter on the output.

There LPF on the input before it hits any actives.
I think there is an integrator cap in the feedback loop on the first gain stage... then another RC LPF before the comparator. The comparator is just one of the opamps from the quad which is a dirt cheap LM324.
I believe when I was tinkering with it I set all the LPF values roughly the same, but I think in the end it worked better for them to not be. I'd really have to go back and look to find out.
LM324s work really well as comparators by the way. Most audio opamps like a TL07X with a jfet input gets all screwy if your inputs get too close to the rail.
The 4013 just gets the comparator output.

Anyway, I'd generally never use an LM324 for audio, but the only audio you hear from it is the comparator output. The only other thing we use it in is the Prometheus DLX where it's used to make 5V switching signals go r/r. again, used as a comparator.

[Liquids]

LM324s work really well as comparators by the way. Most audio opamps like a TL07X with a jfet input gets all screwy if your inputs get too close to the rail.

Awesome info.

I've heard others say that jfet input op amps make the worst op-amp comparators, however, haven't found much in the way of what might be better, and I'm lazy and a bit cynical about swapping op amp types and listening for differences...the microsynth triggering section uses 4558s as comparators. I was originally determinted to re-work it to use LM311 types in there, but it seemed silly quite quickly...so I was going to go with the stock 4558s (or comparable non-jfet input op amp) but I'll be sure to use, or at lest try out the LM358 (dual). Thanks!