New SubOctave Design - Will it work?? (Hacked Blue Box)

[Jaicen_solo]

Ok, the purpose of this post is to get feedback on a new octave down idea i've been thinking about. This is for a friend of mine that wants to sound like Death From Above 1979, ie monstrous bass. I wanted something that is better behaved than the original blue box, and this is the result of my reading (mostly posts by Mark Hammer: Legend :icon_cool:tm).
I can't claim the idea as my own, but I think this is the first time it's actually been drawn up as a schem.
Basically, i've taken the flip-flop circuit from the MXR Blue box, and removed the fuzz. The idea is to have a very high gain in the front end of the pedal which can drive the flip-flop more reliably. The second op-amp gain stage has been set up with lots of gain and a pair of diodes for compression, again this is to make the triggering of the flip-flop more reliable. I considered having an open-loop for the feedback in the second op-amp, but I think that's going overboard.  I also intend to add a variable low-pass filter to the octave down to get a cleaner sound.
The unprocessed guitar sound should be fairly clean, albeit with lots of gain! This will allow lots of gain to be heaped on top of the bass+Clean signal without it turning to mush (as the BlueBox is wont to do).
Obviously the schematic below is very very rushed, I just chopped up pieces of the original Blue box schem from FP's site, but I think it gives the general idea of where I'm going with this.
So what do we think, will it work??

[hank reynolds 3rd]

I'm no whiz, but it looks alright to me (should be easy to breadboard too )..
Maybe use the tone section from the proco rat on the octave bit (it's only about 3 or 4 parts) for a quick tone/filter section...
I did a hand drawn layout for the mutron octave ages ago, and this used a lot of filtering to clean a cleaner sub octave....
going off all the mods from mark hammer  aswell, maybe add a sensitivity pot.....


Sam

[gez]

You need to drive the clock inputs of flip-flops with logic levels, not from the output of an audio amp.  Don't have the data sheet handy, but you might be able to get away with what you've done using the set/reset inputs (that huge 5M resistor would probably cause problems though due to a DC offset)...though you wouldn't be able to divide down with the first stage.

You also need to feed the amp at the output of the flip-flop with something considerably smaller than logic levels...unless you use something like an ICL7621...

Need to do a little redesign.

[hank reynolds 3rd]

http://www.geocities.com/tpe123/folkurban/fuzz/snippets.html

would splicing part of the square wave shaper into the signal going to the 4013 help??

i'm guessing this would change the oct output, but maybe not in an undesirable way??

[gez]

The clock inputs need nice straight +ve edges.  The best way to do this is to use a CMOS device with a Schmidt input.  If you're not using any other of the devices within the chip then a CMOS 555 is ideal instead as it takes up less space (wire it up as a Schmidt trigger).

You can drive some clock inputs directly from op-amps if you drive the hell out of the comparitor bit...it's tempremental though.

PS  There's no gain in the clipper section (it'll act as a follower that limits the signal when the thresholds of the diodes are reached).

[Jaicen_solo]

Yeah I realised my mistake with the second stage almost as soon as I'd posted it. 

To be honest, I don't really fully understand all these logic circuits, (or any for that matter ).
Do i need to bias the driver signal to something like +/- 5v for logic operation?

As I said before, i'd intended to use the second op-amp with an open feedback loop to give a full rail to rail square wave would this provide the necessary conditions to drive the 4013?? Maybe I should check the datasheets

EDIT: Hey, this is my 300th post! Not bad for a young'un eh!

[StephenGiles]

I think perhaps you are trying to reinvent the wheel here. Have a look at the EH deluxe Octave Multiplexer. If you want to stand any chance of an octave down which doesn't sputter and octave jump, you will need to treat the guitar signal in a similar way. The best simple unit I have heard is the Roctave - well documented on this forum.
Stephen

[gez]

Do i need to bias the driver signal to something like +/- 5v for logic operation?

If you change the stage with the diodes in the feeback loop into a comparator that flips between its max output capabilities, or better still rail-to-rail, you're almost there.  Feed the output of this comparator to a schmidtt input device to clean in up and this should be fine to present to the flip-flop's clock input.  So long as all the chips run off the same voltage and the op-amp's output can swing wide enough to trigger the Schmidt, there shouldn't be a problem.

In the back of my mind I think there's a 74 series flip-flop that has a Schmidt clock input, though it runs off a max of 5V or something similar.  You could use a 5V regulator with an op-amp that can run at that voltage and that would save you using another CMOS chip.

The mixing stage would be better done with a simple inverting mixer (the buffer for the first stage is probably superfluous).  Make the feedback resistor smaller than the input resistors.  The blender at Geofex would be a good choice.

Keep plugging away, you'll get there. 

[gez]

I think perhaps you are trying to reinvent the wheel here.

Possibly, but learning how to design is no bad thing and should always be encouraged...even if the design never makes it off the breadboard.

[gez]

Something like this would work well, though I'm sure there are many other choices (first one I came across):

http://www.learn-c.com/74ls74.pdf

"Schmitt-trigger action in the clock
input makes the circuit highly tolerant
to slower clock rise and fall times."

Note that it requires a 5V supply (will need a 5V regulator for the whole circuit).  Use something like ICL7621 amps or sim (quite a few CMOS amps can run as low as 5V) and you should be able to drive the clock input of the first stage directly (get rid of the diodes and give it some gain though).

Edit.  If you don't want to use a regulator for the op-amps there are tricks you can use to interface with the flip-flop running at 5V.

[gez]

PS  As Stephen noted though, you're still going to get yodelling.  A much more complicated circuit would be required to get round this...

[Mark Hammer]

What you've shown is not exactly what I had in mind.  I would encourage you to retain the crude gating capabilities of the Blue Box, since they help to reduce "yodelling" (I like that term in this context!).  The absence of ground connection from the inverting pin in the second stage aside, probably a very good idea to stick a feedback cap in tere to smooth out what the flip flop has to respond to.  Something less than 27pf ought to be fine.

[Jaicen_solo]

Although I appreciate the responses this has got, I must say it seems to have gone beyond the original intention of this design.
The idea is to just modify the original blue box design into something slightly more user friendly, without becoming overly complex.
As Mark has pointed out, there are a number of things missing from the schem, but then that wasn't intended to be complete in any way, merely a design idea.
As far as yodelling goes, I quite like it but I suspect that some additional low pass filtering in the trigger stage could help.
My main issue with the blue box is that it's overly fuzzy which tends to detract somewhat from the low frequencies it generates. I'd like to have a square wave subharmonic (which can then be filtered) blended with the clean guitar sound. Do you guys get what i'm aiming for?

[Mark Hammer]

I get it.  And probably the fastest way for YOU to get it (in the other sense of that two-word phrase) is to make yourself a Rocktave.  Seriously.  The "clean" to be blended in with octave in that design really IS clean.  So much so that it is the exact same signal used when the octave is cancelled for "bypass" mode.

Obviously it is a more complex and costly design, and deprives you of the pleasure of calling a design your own (and believe me, I recognize the pleasure therein!).  But if the overarching goal is to be able to have a clean sound and a nice round-sounding octave below, the Rocktave is your #1 choice.  I might point out that where the Blue Box provides only a choice of divide-by-2 OR divide-by-4, the Rocktave provides level controls for both of those to allow whatever blend you want, as well as a tone control applied only to the divided signals.  As analog octave-dividers go, it gets recommended so much not because we feel any sort of allegiance to PAiA, (the late) John Simonton, or Craig Anderton, or DIY but because it really IS the best analog divider out there in terms of sonic flexibility and tracking.

[gez]

I would encourage you to retain the crude gating capabilities of the Blue Box, since they help to reduce "yodelling" (I like that term in this context!). 

There's a South American bloke (Brazilian?) who is master of the 'Jazz Yodel' - sort of Scat singing but with an Alpine tinge.  I always thought it was a euphemism for being sick, but apparently not.  Having heard a sample of his 'work' I can't tell the difference in all honesty...

[Jaicen_solo]

You know what Mark, I had a feeling that you'd say something like that
I'd normally agree, but this particular divider is going inside a guitar, and that rocktave PCB is pretty big(iirc), not to mention all the offboard.
I'll certainly check it out see how it works at least.
Cheers.

EDIT: It appears that the Rocktave project is no longer hosted on your site Mark, any idea where it's gone??

[Mark Hammer]

It appears that the Rocktave project is no longer hosted on your site Mark, any idea where it's gone??

None, actually.  Steve Morrison at Ampage set up the page for me a couple of years ago, and I think that some things have gone missing once I hit my max storage capacity.  I'll have to speak to him about it, because there is stuff there that even *I* can't get to.