U-boat. NEW Smooth sub-octave / synth effect!

[deadastronaut]

cheers luke, i'll check it out..

[deadastronaut]

?

at the low pass filter (2nd to last opamp) there is a cap to ground....2.2n or 22n ?


edit: nevermind, 22n... just cross referenced..

[deadastronaut]

ok, i have this on breadboard, using a j201 (waiting on some j112's) nice one merlin, this is lots of fun

using cd4013be.

i have the tracking pretty good, and octave...(and have been messing with the low pass filter cap just for synthy fun as you can hear in the clip)

https://dl.dropboxusercontent.com/u/7464107/uboattest1.mp3

(its been a nightmare breadboarding with one eye as i have a really bad eye infection which has totally closed my left eye and swollen my face big time..off to docs/hosp later )

anyway..

like i say the tracking is pretty cool....but where should i be looking/tinkering to get the long unbroken sustain to fade?...

[bluebunny]

Yeah, go and get your eye sorted, mate!     You know it makes sense.

Can you imagine how long your project tinkering threads would be if you can't actually see what you're doing...?  

[mattoverse]

How many of you have had good luck with this?  I have breadboarded it at least 6 different times using all three variations of Merlin's schematic as well as the THC Customs build schematic, and even combinations of all of the above and the best I have got is pretty variable tracking quality, depending on string, with the tracking jumping octaves, bubbling, sputtering and fuzzing, to fizzy clean signal side issues, and ticking, distortion etc.. Have tried a variety of diodes, TL072's, lm833's, J112s, different envelope follower cap values etc etc..  Just not having any good luck, my OC2 still tracks much better.  I was so excited when I found this design as it is exactly what I am looking for.  I have built pedals to replace all my mass manufactured pedals, delay, chorus. compressor, distortion, etc.. and a simple octave down was the last one to finish but don't think I can handle anymore frustration at this point.  Guess I'll have to try the chopped OC2.

[LightSoundGeometry]

very cool, and I have to agree with the artwork, I like it .

[thehallofshields]

Merlinb. Why not use all Tl072? Can they not output tight squares or something?

[~arph]

read:  http://www.valvewizard.co.uk/uboat.html

Tracking was the best with a LM833. Still works with TL072 and R19 installed

[thehallofshields]

Thanks. I missed that. I can't wait to get this circuit on breadboard.

I've got a Japanese OC-2 that is a little janky since I modded it...

Still. Octave-Downs are my favorite effects. I've got high hopes for this project.

[thehallofshields]

I can't understand why R19 helps.

It's going to bring up the Inverted+Rectified signals DC Offset so that...?... Compared to the Inverted Signals 4.5V Bias it helps...?...Because the Tl072 is...?

Anyone care to help me out?

[merlinb]

I can't understand why R19 helps.

Precisely why it works is not clear even to me. I tested a lot of opamps and found the LM833 to work best. The LM833 is a BJT opamp, so it draws a significant input current which the TL072 (JFET) does not. This input current seems to help with tracking, presumbly by creating a (veriable?) offset between the rectified signal and the original which causes it to cut-off more quickly as the note fades out, which reduces spluttering. R19 is an attempt to duplicate this effect when using high-impeance opamps like the TL072. The result is not identical to the LM833, but it's not bad.

[thehallofshields]

To me it would seem like Inverted+Rectified signal gets folded down, so that it's positive peak only briefly touches 4.5V, throwing the Vavg low.

Did you ever experiment with Decoupling and Re-biasing the Rectifier's Output?

[merlinb]

Did you ever experiment with Decoupling and Re-biasing the Rectifier's Output?

I'm not sure what you mean...

[Granny Gremlin]

I'm going to be taking a stab at this thing in a bit.  Anybody tried using 1N914s instead of all those 1N4148s?

[merlinb]

I'm going to be taking a stab at this thing in a bit.  Anybody tried using 1N914s instead of all those 1N4148s?

I would expect them to work fine -they're practically the same as 4148s.

[Granny Gremlin]

That's what I thought, but apparently there's a bit of a difference with max fwd current.  Thanks Merlin (hoping to get to the Eng Thumb soon to).

[merlinb]

Sounds great, Merlin. Playing around with a perf layout. Just wanted to check if a j201 could replace the j112?

Yes, in fact J201 should work better as it has a smaller Vgs(off).

[PRR]

> 1N914s instead of all those 1N4148s?

Today they ARE (probably) the same thing.

The '914 is older and may have a lower max-current number on the sheet.

Diode fabrication is so good today that they probably make one small-signal doped-Si diode which covers nearly all older specs, and supply it under whatever type-number you order.

However here the max current is limited by the opamps, and happens only in quick spikes, never steady. And *most* of the diodes are input protection, behind 10K resistors, so can't be exposed to even 1mA.

I suspect for most of them you could use the C-B junctions of any Silicon transistors you have excess of.

The only one which may be fussy is D3. I too can't grok why that opamp is critical and why a 10Meg helps. While you might try any transistor or diode you find in your pants-cuff, probably best to stick with a '914/'4148 type as Merlin did.