CMOS 4049 Octaver with Clean Blend help

[Mr. Lime]

I would like to build a octaver based on something I found that looks like a TSF Octaver with a clean blend.
As I never used a CMOS before, I would like to try it with a 4049 instead of the OP-amp.

The gain of the CMOS is smaller than the gain of a TL081, so I will boost it with a Alembic Stratoblaster Booster as Input-Stage.
Please let me know if you see any mistakes and feel free to post suggestions, I will be thankful for anything! 



Best regards

[duck_arse]

the first problem is that we can't see that which is or might be at the musikding you linked.

Code Select Expand

http://forum.musikding.de/vb/filedata/fetch?filedataid=13578

shows as an empty page.

[Mr. Lime]

I'm very sorry for this, to me the schematic is shown normally..

I hope this upload link works:

[duck_arse]

I can now see it. your R2 goes to some sort of supply voltage, yes? does the open end of the tone pot go the same place? are you having a series resistor in the 4049 suppy to limit current? and don't forget to tie the unused inputs to a logic level.

[Mr. Lime]

Thanks for your attention to this topic.

I just drew the signal path, the power supply is clear to me.
I still think about a way to use the unemployed gainstages, that's why they are not mentioned yet.
Maybe I use one stage for buffering the true bypass.

The "Tone" control is basically just a blend between the two caps, Death by Audio used this arrangement in their Octave Clang, so the 3rd leg isn't connected to anything.

I thnik about ditching C2 and R3 and relocate the Tone-control behind c3 and c4 to prevent pot scratching, shouldn't it work that way?

[duck_arse]

either way it is wired, the tone pot will have no dc flow because of the blockings of C3 and C4. I was thinking about the C6//C7 section, though, you might need to split yr signal there. I don't think you will have enough oompff to drive the Tx and the blend pot from the same point. you might take C6 from IC1 and add another inverter to feed C7, but they'll then be outta phase (unless you input from R9 and parallel the IC1 stage). maybe an inverter each, with some tone shaping or something?

[Mr. Lime]

I know, the Tone control seems strange but I want to archive the Octave Clang sound as close as possible, and I think that it's a major part of it..

One of my worries was the gain, you may be right adding another gain stage.
I would put the IC2 Stage infront of C6 // C7 and maybe canel the recovery stage?

My priority is the Octave Clang sound but I still would like to get some Red Lama kind of tones out of the circuit, especially for low gain settings on the amp.

Another option would be a recovery stage with just a tad of gain to not colour the tone too much..
Which feedback resistor R6 should I choose best for a transparent recovery stage, 470k or lower?
Two CMOS gainstages + JFET should get into OP-amp-gain territory, am I right?

the new version:


Here are the schematics of the Octaver, Red Lama and the blend circuit:

[duck_arse]

err, no. you've misunderstood what I was driving at with the splitting. allow me to draw something overnight/day and I'll post here same time tomorrow.

[Mr. Lime]

Thank you, duck_arse for your efforts! 

I think I see your point, you mean:

                             -> IC1 -> IC2 -> Transformer
Input -> JFET ->R9                                                BLEND -> IC3 -> IC4
                                            ->           

I don't really want a "clean" blend meaning mixing the clean signal with the Octaver.
I just would like to blend the octave effect itself to the dirty signal formed by IC1 and IC2..

If we take a look at my attachment concerning the octave blend circuit, it only uses one OP-amp stage driving both paths.
I may be completely wrong, so I'm really interested in what you will show me.

CMOS stuff is something new to me, but I try to get into it..
What's about unsing two buffers parallel feeding C6 and C7?

Thanks a lot!

[duck_arse]

well, the picture I drew is worth at least a thousand words, but my pc went south before I could convert it to a png, so it's only in eagle .sch format at the moment.

I'm certainly no expert on cmos as linears, but if you were to politely ask "deafbutpicky", a user here who recently posted some 3 band distortions based on cmos, he may very well provide some useful tips.

[anotherjim]

You probably do need more gain. The open loop gain of the inverter is somewhere between x8 and x10 at 9volts. That's barely going to exercise the clip diodes without some help from the input. I don't know the gain of your input FET, but it might need another inverter after it to drive the diode clipper.
To drive the transformer, you could go differential or bridge mode. A unity gain inverter after IC2 driving the other end of the primary. You don't need C7 then.
Watch your power consumption though. Could be 5mA per inverter!

[Mr. Lime]

Thanks guys,

well, maybe I should use an Op-Amp instead of the JFET, if there is still not enough gain, would that solve all the driving problems too?

I do not use batteries in pedals so the power consumption isn't such a problem..

How would differential or bridge mode look like in my case?
I just look at different CMOS schematics and count 1 and 1 together means that I'm not that familiar with the circuits in detail.. 

If I use a TL071 input stage, wouldn't it be better to put the diode clipping on that stage instead of the CMOS?

Thank you very much!

[duck_arse]

this is the what I was thinking of, or at least along these lines.


and for the colour adverse:http://i.imgur.com/RH1OjT0.png

using C1//R2 as a highpass, and R4//C4 as a lowpass, you can send the higher freqs to the clipper IC1A without low f muddy, and the lows/fundamentals can be sent to the octaver IC1B, and TX1. the two are isolated by R1 and R4 which also ?set the impedance? of their invertor stages, so you can shape what is fed to each section as you like. the output of each then hangs off the mix pot, feeding IC1C, as shown your original.

some points on your original schem: the feedback caps (my C2, C5, C7 etc) should be hard wired across the inverters, to tame their high freq response. you don't really want to be switching these caps, although you can switch-add larger values to decrease the highs further.

the gain pot (my RV1) should have a series resistor included, so you don't short the inverter input to output at min pot setting. this would be a bad thing. I dunno a value, maybe 10k to start?

diode switching and feedback shapings are all over to you now, as are the correct values for this shown. I haven't tested this, I don't know if it will work as you want or I think, but it's a start. but really, you should listen to jim ^.

[Mr. Lime]

Thank you for your great drawing, I updated my schematic with this.

I replaced the JFET with an op-amp to get the right amount of gain.
Hope this circuit makes sin..
Still with the Red Lama and the Octave Clang in mind, so there isn't much new to the filtering stuff, only a Tone Control at the end.

Here is it: