Driving a BBD (V3208) via a CD4047 via TAPLFO

[KarenColumbo]

God(ess) bless all in here!

There's one thing that's been conctantly on my mind: doing a BBD chorus that get's the clock from a CD4047 that is driven by Electric Druid's TAPLFO. Must be fun finding a good LFO waveform that makes a comfortable chorus wobblery. Here's the datasheet of the TAPLFO IC: https://electricdruid.net/datasheets/TAPLFO3Datasheet.pdf

Now I've come up with this here - wchich is a mixture of Tom Wiltshire's excellent "Flangelicious" project and the CD4047/MN3208 section of the VFE "Choral Reef" chorus pedal:



In the Choral Reef schematic the CD4047 is driven by a 2-op-amp-LFO circuit: https://www.madbeanpedals.com/projects/_folders/VFE/pdf/VFE_ChoralReef.pdf

I guessed i "could" use one of the TAPLFO's output (I think it could be pin 5 including the filtering shown on page 7 of the TAPLFO3 datasheet) and put it into the CD 4047's pin 1 as seen in the Choral Reef schematic. So I did, as seen in my schematic.

But I guess there must be tons of caveats.

I already breadboarded the thing as shown in my schematic. Doesn't work (not surprisingly, given my unterstanding of advanced electronics). If I drive the CD4047 via pin 7 of the TAPLFO, I get some kind of distorted delay out of the BBD V3208.

What do I miss here? Could you give this mess an expert glance?

[ElectricDruid]

I think it looks good. The Flangelicious-based audio path is tried-and-tested, so that should be ok. The clock circuit is originally from the EHX Small Clone, as far as I know. Again, tried and tested.

I had a crack at something very similar myself, except I used the StompLFO rather than the TapLFO. Since they're both spitting out a string of 5V pulses, it doesn't make much odds.

The problem I found was getting the passive filtering to play nicely with the clock. The three stage filter from the TAPLFO datasheet is almost certainly the problem. If you think about it, it's putting 10K+100K+1Meg in series with the LFO. The VFE circuit has between 15K and 115K in that same position, so that might be more what we're after.
I'd drop the third stage of the filter altogether, and I'd reduce the resistors by a factor of ten and increase the capacitors by ten to compensate. That'll give you 1K/470n, 10K/47n for the filter. You might also need a resistor between the filter and the clock. I had a 39K in that position because that's what the Small Clone uses, but I have no idea where that value came from, so experiment. Maybe it isn't even necessary.

Even without the LFO, you should be able to get the clock running, and if the clock is running, you should be able to hear delayed signal, so make sure you've got all that working before you start the experiments with the LFO.

Good luck!

[ElectricDruid]

Oh, I've just spotted something!

The BBD needs to be powered from 9V not 5V. In the Flangelicious it's running from 5V because that's the voltage that the  PIC uses, so we have 5V clock pulses. In this circuit, you've got a special power supply for the clock and BBD based around Q2, so you should use that - otherwise the poor BBD is running on 5V but getting nearly 9V clock pulses! Ouch!

[KarenColumbo]

Whoa, thanks for the spotting, Tom! And for your help!

[KarenColumbo]

Well, the BBD is putting out a good loud sawtooth self-resonating tone and suddenly my 9V rail on the breadboard has 7,2V. Ah well. I think I'll take tomorrow off breadboarding this again. Hell of a lot of jumper cables. And of course it's not exactly nice Eagle-ing this one on a neat PCB. I think I may have bitten off quite a mouthful with this one But it's gotta work. I'm pure 80ies deep down, so I need chorus to get me through the days.

[KarenColumbo]

Work got in the way. But i managed to go at this from another angle, boarding a "Zombie Chorus" (MN3007 & 4046) and gradually inserting all those buffer, bells and whistles (like a Ce-2 for example) to ensure stable operation under any circumstances. Not that the venerable Zombie is not stable, tried and tested. But wanna make sure it doesn't explode or poison my dog when I confront it with a digitally generated LFO signal.

Zombie:


CE-2

[KarenColumbo]

Gotta share this

Finally arrived at a compromise: I boarded the "Zombie" (mis-)using Tom's filtering in his "Flangelicious" project. When i first heard it, it was nearly perfect in my ears. After tinkering with various LFOs - including TAPLFO and VCLFO, which worked well but turned out a bit too much - I always returned to this first working version:



Now it's (almost) dead quiet without sacrificing too much treble, not really "lush" but very pleasing to my ears. A bit like a PT2399 thing, but warmer.

There MAY STILL be some errors in the schematic. They will be killed in their sleep.

[bluebunny]

"Zombielicious".  Cool. 

[duck_arse]

ergh. you've got some wrong connections on yer input - the pulldown shouldn't go to "Vcc" [which you probably should call Vcc/2 or V/2] but to ground. you've connected the !C1B input to the outside world, which will also put DC on the input/tip in. you might want a seperate Vcc/2 to supply the oscillator section [without a 10uF cap], as sharing that point between osc and audio will/may/can thump you silly. but mostly, your V3207 pins 7 and 9 are connected to +9V suppply, and seemingly feeding that 9V to the following filter stage as audio. is this correct?

ask antonis for one of his coffeess.

[KarenColumbo]

updated the last schematic - thanks again Duck!

There MAY be some errors in the schematic

Well, "some" is like calling a World War a minor tussle. I dunno why I just post crap like this, but in my euphoric state I had to share 

Corrected what I could find. I had no bleed-through at all from LFO & clock - but I build it on a HUGE breadboard and tried to manage the whole power suplly lines as good as it got. I just manically switched caps around and got rid of the 3n3 cap that originally was at the junction C3/R15/R16/R17 because I got strangest noises in the path. So I went ahead until the whole thing was as good as quiet except the usual breadboard crap and single coil hum. Just thought: "The more treble content goes into the BBD, the more I cut off afterwards - raising C9 to 2n2 and C10 to whopping 10nF seemed to do the trick. Hiss decreased by 99%.

There's a question packaged in this here post: Why the Anti-Aliasing and Reconstruction Filters @ all the CE-2 clones? I breadboarded one of those and checked with a "audio probe" after each step. The signal's volume got so low after the Antialias-Filter that I thought "There can't be enough signal left for the BBD to shift it across thousands of caps", so I ripped the crap off the board. Tom's 8 kHz Butterworths seem to do all that's necessary ... or am I getting deaf at my old age?

And maybe another question: What's the clock-use-related difference between CD4046 and CD4047? NEVER could get anything working with a '47. Putting a '46 in - voilá! Is the 4046 inferior? Some kinda "Idiot's Clock Chip"?

[ElectricDruid]

You've still got the outputs of the BBD tied to 9V, not ground. 9V is the MN3007, ground for the MN3207.

What's C5 doing there? It looks like it's either a left over, or it's trying to be part of a 3-pole filter, but is missing a resistor.

Anti-aliasing filtering shouldn't reduce the level massively, so if that happened, something's wrong. What it's supposed to do is remove any higher frequencies that the BBD won't be able to handle given the clock rate we're intending to use. That does usually imply trimming the treble a bit. My 8KHz filter was a choice made because the lowest clock I was intending to use in the Flangelicious was 25KHz. In theory, you could go up to 12.5KHz (25KHz / 2) but with BBDs dividing by 3 is a more conservative option. I tried 10KHz and it was a bit noisy, so on the next revision I dropped it down to 8KHz.
If your chorus uses a lowest clock frequency that is higher than 25kHz, you could tweak the filters to let more treble through.

The 4046 is a phase locked loop chip, whereas the 4047 is an astable multivibrator, so they're both completely different. The 4046 includes a VCO, which is handy when what we need is a voltage-controlled clock. The 4047 includes an oscillator and can be made to have voltage-control with a bit of effort.