CD4049 Revisited - i.e. Mark Hammer

[Nasse]

Adjusting supply voltage of inverter stage is another forgotten trick - try from 5 to 10 volts. Some brit magazine fu,zz did it straight from 10k pot Viper - avoid funny smell

And lately been thinkin If you could insert cmos stage in Tube screamer...

[Rob Strand]

There's also the Laney amp version,  notice the 4049's are powered from 5V
https://elektrotanya.com/laney_hcm65r_sch.pdf/download.html

Another trick was the resistor from the input to +V or gnd to shift the bias point.

I've never got the x3 inverters in series from the app note to work. I found all 3 needed their own feedback resistor in order to bias (can be really high >1M so as not to affect gain) and then the overall feedback resistor as per the note.
The app note is using the obsolete National 74C04 so that could be the difference, although I'm not sure what it is.

It might work on an oscillator but highly unlikely to work for a linear amp.   You only need to be a small mount of voltage off and it hit's the rails.

[aron]

Can you read the value of the input caps on the inverters? I can't read the values.

[Rob Strand]

Can you read the value of the input caps on the inverters? I can't read the values.

Do you mean for the Laney?

The PDF is much clearer than the preview pic.

10n, 10n, 100n, 22n

IIRC, other Laney models have different part values.  I think one model had crunch and distortion with different 4049 configurations.

I posted some info here,
https://www.diystompboxes.com/smfforum/index.php?topic=122849.msg1161336#msg1161336

[anotherjim]

There's also the Laney amp version,  notice the 4049's are powered from 5V
https://elektrotanya.com/laney_hcm65r_sch.pdf/download.html

Another trick was the resistor from the input to +V or gnd to shift the bias point.

I've never got the x3 inverters in series from the app note to work. I found all 3 needed their own feedback resistor in order to bias (can be really high >1M so as not to affect gain) and then the overall feedback resistor as per the note.
The app note is using the obsolete National 74C04 so that could be the difference, although I'm not sure what it is.

It might work or an oscillator but highly unlikely to work for a linear amp.   You only need to be a small mount of voltage off and it hit's the rails.

Korg used x3 of 4069UB inverters as ring oscillator for ensemble chorus modulation in the Delta. They fitted individual feedback resistors. I've copied the design. It works well giving x3 120deg phase-shifted sine outputs for BBD delays (Delta only had x2 at 120deg apart) -  it has to have the feedback resistors or it won't start. That design provides two LFO's for fast & slow chorus and much more compact than the LFO + phase shifters used in the Solina and others. One chip - six outputs!

[Rob Strand]

Korg used x3 of 4069UB inverters as ring oscillator for ensemble chorus modulation in the Delta. They fitted individual feedback resistors. I've copied the design. It works well giving x3 120deg phase-shifted sine outputs for BBD delays (Delta only had x2 at 120deg apart) -  it has to have the feedback resistors or it won't start. That design provides two LFO's for fast & slow chorus and much more compact than the LFO + phase shifters used in the Solina and others. One chip - six outputs!

Neat circuit.  I guess you have to trust their better judgment that saving a couple of resistors isn't worth possible headaches in production (even if it did work).  Resistors on each stage helps symmetry.

[Mark Hammer]

The Korg Delta chorus.  Certainly a different kind.  Not simply counterswept.

[anotherjim]

Same circuit I have Mark. Note the Rin/Rf inverter resistors are setting gain around x3. That's critical - too much gain and it distorts - too little gain it takes ages to start; but additionally, the Rin value has to match with the feedback cap for the required frequency! Rob seems to know that already 

Also note the extra measures setting supply & reference voltages to have oscillator outputs in the DC control range for the 4046 VCO's since LFO speed requires DC coupling to avoid large coupling capacitors. I would have thought you can take advantage of the high impedance of the VCO control pin to AC couple the LFO's via 1uF caps and with a large value resistor setting the midpoint from a reference and simply mix the 2 LFO's into them with resistors.

On top of all that, the opamps used to mix the LFO's have reduced gain on the negative out swing due to those diodes D5 & 6. I wonder why they did that?

[Rob Strand]

The Korg Delta chorus.  Certainly a different kind.  Not simply counterswept.

The 120 deg phase shift on the LFO is a very interesting idea.    Especially in that it's used on a mono chorus.   It's a slightly different spin to the 180 deg  (up/down) idea used on the Boss Dimension C.    I did some DSP experiments using the dimension C idea but it didn't do what I wanted.  I think I tried 90deg as well but that didn't work out either.  Maybe the missing link is the 120deg- damn!

On top of all that, the opamps used to mix the LFO's have reduced gain on the negative out swing due to those diodes D5 & 6. I wonder why they did that?

I'm pretty sure that's a simplified hyper-sine LFO (like the black box Ross phaser).    I know other pedals have used this, maybe Mutron?  The shaper needs to be DC coupled so it only clips the negative swing, so that's probably why they kept the LFO DC coupled all the way.

Places like Korg had a lot of unsung heros.    Novel idea and a nice clean circuit - and off the shelf parts.

[moosapotamus]

FWIW, I've done the parallel inverters...

http://moosapotamus.net/ideas/tube-steak-fuzz/

[aron]

What difference in sound did the parallel inverters make?
Thanks!

[Nasse]

Uraltone sells micro mixer kit, its 4049.

I have seen active bax type tone control schem done with inverters, and sum/difference

[Mark Hammer]

The Korg Delta chorus.  Certainly a different kind.  Not simply counterswept.

The 120 deg phase shift on the LFO is a very interesting idea.    Especially in that it's used on a mono chorus.   It's a slightly different spin to the 180 deg  (up/down) idea used on the Boss Dimension C.    I did some DSP experiments using the dimension C idea but it didn't do what I wanted.  I think I tried 90deg as well but that didn't work out either.  Maybe the missing link is the 120deg- damn!

There doesn't seem to be any means of changing speed in the Delta circuit.  Are the two clocks synced-but-staggered by 120 degrees, or are they unsynced, with the 120 degrees simply adding some aperiodicity?

[Ben N]

And lately been thinkin If you could insert cmos stage in Tube screamer...

Like this? http://www.runoffgroove.com/ubescreamer.html

[Electron Tornado]

There's also the Laney amp version,  notice the 4049's are powered from 5V
https://elektrotanya.com/laney_hcm65r_sch.pdf/download.html

Another trick was the resistor from the input to +V or gnd to shift the bias point.

The Laney HCM30R also uses the 4049 for some distortion. The schematic (and entire service manual) for that amp used to be online.

[anotherjim]

The Korg Delta chorus.  Certainly a different kind.  Not simply counterswept.

The 120 deg phase shift on the LFO is a very interesting idea.    Especially in that it's used on a mono chorus.   It's a slightly different spin to the 180 deg  (up/down) idea used on the Boss Dimension C.    I did some DSP experiments using the dimension C idea but it didn't do what I wanted.  I think I tried 90deg as well but that didn't work out either.  Maybe the missing link is the 120deg- damn!

There doesn't seem to be any means of changing speed in the Delta circuit.  Are the two clocks synced-but-staggered by 120 degrees, or are they unsynced, with the 120 degrees simply adding some aperiodicity?

Delta is 2/3rds of a string synth ensemble chorus. Conventionally, there are x2 LFO's and each goes through phase shifters to put the x3 BBD clock modulation 120deg apart. The LFO's could be adjustable, but the phase shift is only  120deg at the designed frequency. The Delta oscillator frequencies are as hard to make adjustable as a PSO, you could think of the Ring oscillators as a PSO with each of its x3 3 RC networks buffered. However, if you do adjust the frequency, the phase shift will still be 120deg.
Why 120deg? Many stereo recordings of rotary speakers don't use the obvious 180deg mic placement. Different engineers have different ideas about mic placement, but an average separation would be 120deg to give a sense of rotation without a ping pong effect and sounds better mixed to mono. I've no idea why 90deg wouldn't work - I guess it's too close for some reason (phase cancellation?)

In all ensemble chorus, the two LFO's are free of each other. The fast LFO is nearly always x10 faster than the slower. They are chosen close to the two-speed Leslie, but a little slower. But each of the phase-shifted outputs keeps the same relationship.

[Electron Tornado]

And lately been thinkin If you could insert cmos stage in Tube screamer...

Like this? http://www.runoffgroove.com/ubescreamer.html

Here's another idea from the folks at runoffgroove:

http://www.runoffgroove.com/doubled.html


I have the app note linked in a previous post. Are there any sources that discuss using the inverters in a filter, or how to create filters between two inverter stages?

[Rob Strand]

There doesn't seem to be any means of changing speed in the Delta circuit.  Are the two clocks synced-but-staggered by 120 degrees, or are they unsynced, with the 120 degrees simply adding some aperiodicity?

That circuit is doing two things.  You have the dual LFO, which is fair enough and adds it's own character and isn't in itself a new idea.   Then you have the dual BBDs which are fed by phase-shifted clocks.  To simplify the though process it's best the separate out the dual LFO and the phase-shifted LFO.    If you think of one LFO frequency the each of the BBD clocks is getting modulate as usual, by they are at different frequencies.   If the clocks tracked there wouldn't be much point to having the two BBDs.   The Dimension C shifts one BBD up and one BBD down.   One BBD will have an upward pitch shift and one will have a downward pitch shift, that makes the chorusing effect thicker.   The Korg is doing the same thing but it's not simply one up and one down.

Why 120deg? Many stereo recordings of rotary speakers don't use the obvious 180deg mic placement. Different engineers have different ideas about mic placement, but an average separation would be 120deg to give a sense of rotation without a ping pong effect and sounds better mixed to mono. I've no idea why 90deg wouldn't work - I guess it's too close for some reason (phase cancellation?)

The 90deg and 180deg shifts might not have worked because of the specifics of the idea I had.   I was thinking back to the project and in the back of my mind i may have tried 120deg as well,  However I think I added three sources not two like the Korg and that created the same problems I had with the 90deg and 180deg cases.   The key feature of the Korg is there's  only two BBD's but they are 120deg shifted.  That set-up breaks the symmetry and might be the key to added "goodness".

I think the difference in the pitch-shifts make it difficult to compare to any mic set-ups, which would phase-related not pitch.

Thanks for bringing this one up.    There's always something new out there,  even though it's actually old  .

[Harry Muff]

all i know is the schmitt trigger produces a square wave that has a nice fuzz tone to it lol

also has a super high Zin which I believe is good for guitar

[jonny.reckless]

I did some experiments with various unbuffered CMOS inverters a few years back, but never was satisfied with the results I could obtain. They are very noisy (hiss) and sounded quite grainy to my ears. I could alleviate the noise somewhat by using 4 in parallel for each gain stage, but I couldn't ever get rid of the grainy sound. This is true to a certain extent for all MOSFETs. They work well for super high gain, saturated metal tones, but I was never able to get a sweet mild overdrive sounding right. I even tried removing the feedback resistor and directly biasing the input to VDD/2 to smooth out the transfer curve, to no avail.

What's the secret to getting good tone out of these?