Stereo Red Llama decoupling question

[snk]

Hello
I am planning to build a stereo Red Llama.
I know that when you build stereo effects (or two pedals into a single enclosure), you need to decouple the power supplies from each other.

In a previous thread, I got this wise advice : "Wire two 100 ohm resistors from your +9V input, one to each pedal power in, and drop a 47uF cap to ground from the power input to each pedal also. "

However, when looking at the Red Llama schematic, I see a 1k resistor right after the power socket, and then a 100µF cap to ground.

So I am wondering : isn't the circuit already "coupling safe", or should I still add a 100ohm resistor and a 47uF cap to each board?

The schematic is here :
http://beavisaudio.com/schematics/Way-Huge-Red-Llama-Schematic.htm


And, for reference, I might use this layout :
https://tagboardeffects.blogspot.com/2012/06/way-huge-red-llama.html

[Focalized]

I'd like to see the answer here as I have a bunch of two in one pedals and a three in one. I didn't do anything special to the power section, just wired the points per layout. No problems short term.

I'm not sure what you mean by stereo though. Two outputs? You seem to be referring to two different things.

[antonis]

So I am wondering : isn't the circuit already "coupling safe"?

It is..

[snk]

Thank you, Antonis.
This is what I suspected, but wasn't sure
So I can just make two boards without any mod, right?

[duck_arse]

why two boards? you'd use those unused invertors for the other channel, wouldn't you?

[snk]

why two boards? you'd use those unused invertors for the other channel, wouldn't you?

This didn't even occurred to me 

[ElectricDruid]

why two boards? you'd use those unused invertors for the other channel, wouldn't you?

+1 Damn right you would! It'd save no end of messing about, and avoids any problems with the power since it's only one chip.

[snk]

Ok, I need to figure out how to do it.
In the schematic posted on Beavis, the sound goes through pins 14 (A6-output),15 (Q6 -input), 9 (A4-output), 10 (Q4-input).
I guess I could use the Q1(input-pin2) & A1 (output-pin3) pair as well as another pair, but on the Beavis schematic pins 3, 5, 7 are used for the power supply part, so I don't know how I should handle that?

[snk]

If I sum up :

pin 01 is +9V, and pin 08 is ground.
Pins 13 & 16 are unused.
the Red Llama uses the 09-10 pair, as well as the 14-15 pair for processing sound.
So we would have pins 02-03, 04-05, 06-07 and 11-12 left... but the pins 03,05,07 and 11 all seem to go to the power rail.

[snk]

I think I found the anwser :
https://www.reddit.com/r/diypedals/comments/tz6eof/cd4049ube_questions/

There are 2 reasons we tie the unused inverters to either ground or vcc.

    power consumption.

    When a CMOS pair is in high or low state, one of the transistors is blocking all current, as there is no external load on the pair, there is no current at all running through the stage.

2) noise/oscillations and radio

    When an unused CMOS stage is let loose it will often bias itself through leakage, then it will try to either amplify any noise it picks up and/or generates, or it picks up output signal from itself and starts to oscillate. This noise or oscillation leaks into the signal path and makes stuff sound bad.

The CMOS pair can bias itself with only a few external passives, providing external bias is quite hard to get to sound good.

Also, you can't bias one stage with another, they will have different natural 'center points' to them, so decoupling of DC is required between stages.

[ElectricDruid]

If I sum up :

pin 01 is +9V, and pin 08 is ground.
Pins 13 & 16 are unused.
the Red Llama uses the 09-10 pair, as well as the 14-15 pair for processing sound.
So we would have pins 02-03, 04-05, 06-07 and 11-12 left... but the pins 03,05,07 and 11 all seem to go to the power rail.

They've simply tied the unused inverter sections to the power rail to stop them floating about. You can pick any two of those sections and use them for the second drive circuit. Disconnect the input for the ones you need and off you go. The circuit schematic for the second drive will be exactly the same except the pin numbers change for the inverter sections. It'd look something like this (I've used the e and f inverters, but you choose!):

[antonis]

Why not another one, using U1 C & D..??

Or using only one inverter and a mixer, with a degree of phase shift..

[amptramp]

We usually think of CMOS as a low-power logic family but when CMOS gates are operated in the linear region with the input voltage at half the supply voltage, each gate takes about 8 mA.  In this case, you could have a 32 mA drain with four gates operating in the linear region that rapidly goes to almost zero at the peaks of a large signal.  If one channel has a large signal while the other is quiet, the large signal will modulate the quiet signal with the 16 mA drain through the Vcc series resistor.

If you like this sound, use two 4049's, one for each channel and use separate Vcc decoupling with a resistor and an electrolytic capacitor and a ceramic capacitor across the electrolytic since most electrolytic capacitors are deficient at higher audio frequencies.

BTW you might need some resistance in series with C4 just to limit high-frequency noise and gain.

[snk]

Hello all,
Thank you for your help and advices!
I have started working on a schematic and layout using one single 4049 for a dual circuit, and wanted to publish it here, but I haven't finished yet and Electric Druid beat me to it

[snk]

They've simply tied the unused inverter sections to the power rail to stop them floating about. You can pick any two of those sections and use them for the second drive circuit. Disconnect the input for the ones you need and off you go. The circuit schematic for the second drive will be exactly the same except the pin numbers change for the inverter sections. It'd look something like this (I've used the e and f inverters, but you choose!):

Thank you. This is what I understood after reading the Reddit commented posted above, and it made me start working on a new circuit and layout.
On my version, I used the 11-12 and 14-15 pairs, as it was easier for me to make a layout with the other circuit on this side of the chip.
Thanks a lot for the edited schematic !

[snk]

This is the layout I came with.
I borrowed it from here : https://tagboardeffects.blogspot.com/2012/06/way-huge-red-llama.html

Please excuse the rough copy & paste for the pots.
Also, the 68nF cap outside the veroboard is a bit ugly, it's designed like this because in the end I am planning to put two caps on a switch.

Does it seem ok like that?

[snk]

We usually think of CMOS as a low-power logic family but when CMOS gates are operated in the linear region with the input voltage at half the supply voltage, each gate takes about 8 mA.  In this case, you could have a 32 mA drain with four gates operating in the linear region that rapidly goes to almost zero at the peaks of a large signal.  If one channel has a large signal while the other is quiet, the large signal will modulate the quiet signal with the 16 mA drain through the Vcc series resistor.

If you like this sound, use two 4049's, one for each channel and use separate Vcc decoupling with a resistor and an electrolytic capacitor and a ceramic capacitor across the electrolytic since most electrolytic capacitors are deficient at higher audio frequencies.

BTW you might need some resistance in series with C4 just to limit high-frequency noise and gain.

Interesting. Thank you for sharing this information.
I understand it may be a concern is one channel is fed with a full volume and the other rather silent, but since I am planning to use this stereo device on busses and groups from a mixer, I think that most (if not all) of the time the right channel volume will be roughly equal to the left channel volume. In such a case, is it a problem?

[Mark Hammer]

The RL, and similar 4049-based drives, provide a LOT of output; enough that you can accept the passive loss and consider inserting some filtering or tone control.

For example put a pair of 2K2 resistors in series between the 10uf output cap and the volume pot input.  Now run a 10nf cap from the junction of the two added resistors to ground as well as from the junction of the 2nd added resistor and input of the volume pot to ground.  That gets you a nice 2-pole lowpass at around 7.2khz.  Plenty of bite but a bit less hiss.  Now install a DPDT on-off-on toggle, that adds 5600pf in parallel with each 10nf cap, for a 4.6khz rolloff or 12nf in parallel, for a 3.3khz rolloff.  You can play with the parallel cap values to achieve different rolloffs, but the key thing is that, with a 2-pole filter, the different corner frequency settings will be quite noticeable.

That's one of the circuits.  For the other one why not consider a Big Muff style dull-to-bright pan tone control.  In that instance, I'd use the Op-Amp Muff values, since they employ a 10K Tone pot.

[amptramp]

We usually think of CMOS as a low-power logic family but when CMOS gates are operated in the linear region with the input voltage at half the supply voltage, each gate takes about 8 mA.  In this case, you could have a 32 mA drain with four gates operating in the linear region that rapidly goes to almost zero at the peaks of a large signal.  If one channel has a large signal while the other is quiet, the large signal will modulate the quiet signal with the 16 mA drain through the Vcc series resistor.

If you like this sound, use two 4049's, one for each channel and use separate Vcc decoupling with a resistor and an electrolytic capacitor and a ceramic capacitor across the electrolytic since most electrolytic capacitors are deficient at higher audio frequencies.

BTW you might need some resistance in series with C4 just to limit high-frequency noise and gain.

Interesting. Thank you for sharing this information.
I understand it may be a concern is one channel is fed with a full volume and the other rather silent, but since I am planning to use this stereo device on busses and groups from a mixer, I think that most (if not all) of the time the right channel volume will be roughly equal to the left channel volume. In such a case, is it a problem?

You will get some crosstalk between channels.  This may be OK if you do not need complete separation but there will be some right channel signal on the left channel and vice versa.

[snk]

Ok, thank you for the information, Amptramp.
Since I have drawn a layout and it is a low parts count design, I might try to build it this weekend, and see if it is usable. Else, I will build 2 boards.

Did anyone take a look at my layout before I build it? Does it seem ok?