frustration, novelty, abuse and wonderful tone - the Xenos Overdrive

[iainpunk]

the Xenos Overdrive



a natural sounding, medium gain overdrive / light distortion. it does some edge of breakup sounds, and someone described its highest gain setting as "tube amp like'', although i won't call it that. i really do think its one of the best overdrives, as it was conceived of through Divine Intervention, hahaha!

i have been sitting on this idea for atleast two years, i was looking in to the data sheet of the CA3130 and CA3160, and came across this:



the output stage of a CA3130 is a CMOS inverter, like the thing that makes it sound great in the tube sound fuzz !?!
the first thought that came in my head was ''that curve seems like something that would sound great, but how can i get it to keep its curve?'' as i realised that the negative feedback of the full opamp linearizes the output massively.

the answer is in here:


i came across it but didnt realise its potential when i saw it first. only after realising that i could put a feedback loop from pin 8 (meant for strobe/compensation) back to an input, i kind of knew it could be done and that it might sound amazing. (spoiler, it does)
So im basically abusing pin 8

looking at the schematic a bit more, i realised that the preamp's input pins change over their (+) and (-) designation if you omit the power amp (inverting CMOS gainstage) from the feedback loop, so the loop had to be terminated on the (+) side, not the (-) side. another consideration is the low current pin 8 can put out, so i choose high value components compared to a more typical feedback loop around an opamp.

i experimented with using one feedback loop from pin 8 to pin 3, and having the input and bias comming in to pin 2
at first i was constantly fighting the bias of the opamp, being biased a little bit off turned the potential ''magic overdrive'' in to a ''fart fuzz'' with low sustain. i had precision trimmers, stopper resistors and every trick i know in the book thrown at it, but nothing did the trick. every time i biased it so it sounded good, it would wander off of that spot the next time i sat down on my desk to tinker with it, and fart again.
i did however in this brief time find out that the amount of gain and the current draw in the feedback loop needed to be kept quite low. a high gain would just sound too distortion-ish, and too high current would make the preamp hard clip.

i had thrown this project aside out of frustration with the bias. i build a kick drum synth, an octave fuzz and even found (yet) another hobby in the mean time (building DIY NERF blasters).
one day i was designing a circuit with and thinking about the self bias a CMOS inverter can do, and then 3 realisations hit me. im like 45% convinced it was divine intervention

1) the CMOS gainstage can provide its own bias if the feedback path from its output to input is DC unity gain. that's why people put capacitors between gainstages of a CMOS inverter design, so the loop is DC unity.

2) that i can give the internal preamp AC gain without DC gain, like we do in dozens of OpAmp pedal designs, with that capacitor in the ground leg of the feedback loop.

3) through taking the CMOS output, extracting its DC via a filter and then putting it back in to the input of the preamp, i can make the CMOS output self bias again, like in the famous CMOS pedals.

the circuit uses 2 feedback loops, one is controllable and determines the preamp's gain (it has the gain control in there), while the other one basically asks an AC gain so high, that the output can be considered to be in open loop operation, while the DC gain it asks to do is unity.



the 3k3 input resistor is not critical and can be anything below 22k and still sound the same, same goes for the input cap, but i haven't experimented with its value. input impedance is about 4.7M, as the inputs of the opamp are MOS inputs

when i first tried this mind worm out on my breadboard, i wasn't sure it would work, i hadn't drawn a schematic of it yet, and no one else i could find exploited this IC in such a way. when i put it on the breadboard at first, it sounded nearly perfect. as i figured out the amount of gain and current in the feedback loop in the earlier experiments.

i was proud as a peacokc when i got it to work! i was so proud that i decided only to announce my discovery in a big post, with full explanation and finished build pictures.
another facet of my pride is my drive to use and access to fancy vintage components.
i used good looking capacitors, a 0.1u miniprint, a 22n tropical fish, and 2 Ducati (yes the motorsport brand) 47n capacitors
the resistors are red 1% or 0.25% resistors with their values written out, not using color code. i used an CA3160 in my build, but it sounds nearly the same using a CA3130, and the tonal spread overlaps hugely




the name came from the enclosure. for a long time i didn't know how to call this circuit. my codename for this project is 'wrath' but that wasn't a name i was planning to put on the eventual pedal. at first i wanted to call it the 3130 Overdrive, but that hat too much of a Peavy / EVH association with it.
there is this Dutch store chain called Xenos and it has this cool lunch box:


cheers,
Iain

[Fancy Lime]

It's finally here! Noice! I need to order some parts...

Andy

[iainpunk]

it took me a while, every small bit of pride comes with a slightly larger bit of insecurity. i also wanted other peoples opinions on it before i finally build it, so a lugged my breadboard around town to gather opinions. most were positive, a friend of mine who builds DIY synths was amazed about the low parts count.

if you want it use it as a more conventional opamp with linear output, but with self biasing features, pin 8 has a DC component on it, and you can filter the signal out of it and use it to generate a Ubias for other parts of the circuit as well.

this circuit can also be used to a similar effect by reducing the 4M7 between out and the 0.1uF to 10k, and making the 0.1uF larger to like 22uF. the node there in-between is a Ubias point as well, if any one thinks its a good idea to add other things to this circuit.

cheers

[Mark Hammer]

Funny, isn't it, how we treat ICs as these black boxes <sic> and only think about the stuff around them, never pondering what's inside them.

Thanks for the prod. 

[iainpunk]

Funny, isn't it, how we treat ICs as these black boxes <sic> and only think about the stuff around them, never pondering what's inside them.

Thanks for the prod. 

You're welcome. This actually isn't the first 'chip abuse' project ive done but its the most satisfying because its smooth and good, instead of harsh and rough.

Cheers

[Rob Strand]

the name came from the enclosure. for a long time i didn't know how to call this circuit. my codename for this project is 'wrath' but that wasn't a name i was planning to put on the eventual pedal. at first i wanted to call it the 3130 Overdrive, but that hat too much of a Peavy / EVH association with it.
there is this Dutch store chain called Xenos and it has this cool lunch box:

Yeah, really cool looking.  The rounded corners and pressed metal gives it a vintage look.    All you need is hammer-tone or a motley-gray pain job.

[iainpunk]

Thanks, i really like them enclosures too. On sale right now so i might buy a bunch more.

Might design a pcb and build a few of em to sell, or just sell the pcb's. Im not experienced in pcb design and distribution tho...

Cheers,
Iain

[Keppy]

That is incredibly cool! Thanks for sharing.

[Elijah-Baley]

Good job.
But there's no a demo?

[iainpunk]

Good job.
But there's no a demo?

I am gonna record some clips this Sunday, while i have access to a the practice space. I wanted to include that in the big post, but forgot to mention it

Cheers, Iain

[imJonWain]

Thanks, i really like them enclosures too. On sale right now so i might buy a bunch more.

Might design a pcb and build a few of em to sell, or just sell the pcb's. Im not experienced in pcb design and distribution tho...

Cheers,
Iain

Could always design the board, do a run for yourself & sell the extras then throw the design up on Oshpark for others?

[Ben N]

Iain, were you purposely looking for a bit of voltage drop in using a 1N4007 rectifier in series connection in the power supply as opposed to a Schottky, or does it not matter?

[ElectricDruid]

Nice work!  I have a few metal CA3130's kicking about, so this would be an excellent circuit for them!

Where did you see that pin 6 can't output much current? I had a look at the datasheet (below) and I can't find that. However, the diagram you posted shows the pin 6 output driving a 5K load to -V, which isn't what I'd call a very light load.

https://pdf1.alldatasheet.com/datasheet-pdf/view/66328/INTERSIL/CA3130.html

What I'm thinking is that all those high value resistors will tend to make the thing noisier than it might be otherwise, so if it were possible to reduce them by a decade or so, it would be a good idea.

[anotherjim]

I've given up wasting Schottky diodes on input protection. If 0.4v or so more drop to +v is bad, there's something wrong with the design.

[PRR]

.... Where did you see that pin 6 can't output much current? I had a look at the datasheet (below) and I can't find that.
https://pdf1.alldatasheet.com/datasheet-pdf/view/66328/INTERSIL/CA3130.html ... ...

Figs 23 24 give output voltage drop as a function of load current. Assuming 9V supply and 1V allowed drop, you can have about 4mA.

The "5K" is one of a family of curves down to 500r, which is near the half-voltage point. On 9V power that is still far over 1 Volt and ample for any sane g-amp slamming.

[iainpunk]

Iain, were you purposely looking for a bit of voltage drop in using a 1N4007 rectifier in series connection in the power supply as opposed to a Schottky, or does it not matter?

I couldn't detecti any audible difference between a 4007 and a Schottky, so i chose the cheapest option.

Nice work!  I have a few metal CA3130's kicking about, so this would be an excellent circuit for them!

Where did you see that pin 6 can't output much current? I had a look at the datasheet (below) and I can't find that. However, the diagram you posted shows the pin 6 output driving a 5K load to -V, which isn't what I'd call a very light load.

https://pdf1.alldatasheet.com/datasheet-pdf/view/66328/INTERSIL/CA3130.html

What I'm thinking is that all those high value resistors will tend to make the thing noisier than it might be otherwise, so if it were possible to reduce them by a decade or so, it would be a good idea.

Pin 8 can't put out much current. Lowering the resistors by 10x changes the smoothness at lower gain settings to much. Its not particularly noisy, except on max gain, but still less noisy than most mu other drive pedals.

Where did you get that 5k load figure from? There's a 4M7 and a 500k resistance in parallel as load on pin 6. Edit: never mind this paragraph, i get it.

Cheers

[iainpunk]

so, Sunday in the rehearsal space, when i tried to record the demo, my dads laptop screwed the pooch, and added ''noise reduction'' to everything, and i had work on Monday which also spilled over to Tuesday morning so i didn't have the energy... so finally here: the Xenos Overdrive Sound Demo.

https://voca.ro/1jdAC4kgBQ3C

two riffs, with progressive gain. the first is the lowest setting, followed by 9, 12, 3 o'clock and full gain at the end.

cheers,
Iain