CD4007-Based Distortion Humming Issue

[Baran Ismen]

Hey everyone.

I have some CD4007UBE's at hand and I had no idea what to do with them, then after some research, I found that they can be used for an overdrive, even distortion circuit.

I've traced the following vero layout, made slight changes, and made my first breadboard circuit.

https://www.parasitstudio.se/stripboard-layouts/nut-cruncher

It worked! It's a true beast, when entered a clean channel of the amp, it growls much like an HM-2, and when on the dirty channel, it growls more. I've also added a modified BMP tone circuit with mid-boost, and it also worked great. Also, when the 4.7uf output cap's positive side is connected to the 13th lug instead of 12, the overall output gain drops and the circuit acts like an overdrive pedal, sort of a booster maybe. I plan to add a switch for that. I also want to play around with other parts more to see their capabilities and how they affect the sound. It's also quite silent... But;

There's this slight yet weird problem.

The gain pot works as it should be, but during its range movement, it hums so much; it is more like a grounding issue, but everything is grounded, and when the gain is set to full or zero, it's working quite normally, no hum or buzz or nothing at all, but when I move the pot, it's humming. Touching it also makes things worse. I tried several different tapered pots with different values, but no luck. Connecting a single resistor works indeed, but then I don't have any gain setting.

What would be the cause do you think? I tried directly grounding the 1st lug of the Drive pot, but still no go.

Attached is the schematic I've prepared according to the layout.

[antonis]

The gain pot works as it should be, but during its range movement, it hums so much and when the gain is set to full or zero, it's working quite normally, no hum or buzz or nothing at all, but when I move the pot, it's humming.

Just to make things clear..

Does the pot hum only during wiper travel or at any setting other than its extremes..??

[Rob Strand]

What would be the cause do you think? I tried directly grounding the 1st lug of the Drive pot, but still no go

You probably need to use coax from the Drive pot to the PCB.   

It's needs to be wired like this.

2xlengths of coax:
- coax 1: 
       shield to input socket and PCB ground
       core from Drive pot pin 2 to PCB
- coax 2:
       shield to PCB  ; easier to find ground point on the PCB.
       core from Drive pot pin 1 to PCB


Instead of coax you can tightly twist two wires together.

It's also a good idea to run the wires close to the metal enclosure between points.

Also you should make sure the metal chassis of the pot connects to ground.

[idy]

touching pot implies it is your antenna. Maybe having it in enclosure will help, maybe just a knob?

[Baran Ismen]

The gain pot works as it should be, but during its range movement, it hums so much and when the gain is set to full or zero, it's working quite normally, no hum or buzz or nothing at all, but when I move the pot, it's humming.

Just to make things clear..

Does the pot hum only during wiper travel or at any setting other than its extremes..??

Only during the travel. Both ends are just fine.

[Baran Ismen]

What would be the cause do you think? I tried directly grounding the 1st lug of the Drive pot, but still no go

You probably need to use coax from the Drive pot to the PCB.   

It's needs to be wired like this.

2xlengths of coax:
- coax 1: 
       shield to input socket and PCB ground
       core from Drive pot pin 2 to PCB
- coax 2:
       shield to PCB  ; easier to find ground point on the PCB.
       core from Drive pot pin 1 to PCB


Instead of coax you can tightly twist two wires together.

It's also a good idea to run the wires close to the metal enclosure between points.

Also you should make sure the metal chassis of the pot connects to ground.

I think you mean coaxial cable right, like one for live, another naked wrapped around it, sort of regular mono instrument cables.

[antonis]

Only during the travel. Both ends are just fine.

Say, you set it at 60% clockwise and leave it alone..
Does it hum..??

In other words, replacing it with a resistive voltage divider should it also hum..??

[Baran Ismen]

Only during the travel. Both ends are just fine.

Say, you set it at 60% clockwise and leave it alone..
Does it hum..??

In other words, replacing it with a resistive voltage divider should it also hum..??

Yes, it does on any setting in between, and when replaced with a resistor it does not hum.

[Baran Ismen]

Delving into the CMOS chips further today, and it led me thinking about whether it'd be possible to make a noise gate with any of CD- the series chips.

This has probably been thought of before by someone else who walks this earth indeed, I ain't Einstein, but I'd like to know the possibilities. While breadboarding above-mentioned circuit, I connected one of the 1M resistors incorrectly, and then when I stopped playing, like hit a single note and muted the strings, the output was going to zero like a noise gate, then shortly after (maybe 20-25ms) it was increasing again. I fixed it by changing the position of that resistor, but this led me to think of a possible noise gate, as this chip has transistors inside, and it's also based on 0-1 logic.

Just trying to fit the pieces of the puzzle and doing some readings now from here.

https://www.tinaja.com/ebooks/cmoscb.pdf

[BJF]

Hi there Baran,

The way the Gain control sits in the circuit makes it an antenna and especially if fed by a high impedance source like an electric guitar. If possible try moving gain control one stage so that input sees guitar and Gain is feed by a transistor. You will have to analyze the first stages of circuit to do this and may have to control gain of input stage to get desired minimum setting. Noise and hum will drop considerably
A variant of this that is less effective is to buffer input. If for buffer using a discrete transistor this will need to be decoupled at Drain or Collector ,series resitor and shunt capacitor to ground because the Power Supply Rejection Ratio is very poor
Consider also that the transistors inside CD4007 also have poor PSRR and will need decoupling if powered from anythingelse than a battery

Wiring
Besides using shielded cables you can probe with finger sensitive components such as perhaps the input capacitor. If a capacitor is very sensitive and there is no low impedance point the outer foil of capacitor can be connected to then a shield made of a solid core wire can be wrapped around the capacitor and grounded in one end ( do not ground both ends or the contraption becomes a transformer).
The wire when used as above creates a sheild for the capacitor.

Generally in design high impedance and high gain is asking for trouble and one can look at scaling down values to lower total impedance but as the Gain sits in the circuit this is not an option.

Have fun
Bjorn Juhl

[Baran Ismen]

Hi there Baran,

The way the Gain control sits in the circuit makes it an antenna and especially if fed by a high impedance source like an electric guitar. If possible try moving gain control one stage so that input sees guitar and Gain is feed by a transistor. You will have to analyze the first stages of circuit to do this and may have to control gain of input stage to get desired minimum setting. Noise and hum will drop considerably
A variant of this that is less effective is to buffer input. If for buffer using a discrete transistor this will need to be decoupled at Drain or Collector ,series resitor and shunt capacitor to ground because the Power Supply Rejection Ratio is very poor
Consider also that the transistors inside CD4007 also have poor PSRR and will need decoupling if powered from anythingelse than a battery

Wiring
Besides using shielded cables you can probe with finger sensitive components such as perhaps the input capacitor. If a capacitor is very sensitive and there is no low impedance point the outer foil of capacitor can be connected to then a shield made of a solid core wire can be wrapped around the capacitor and grounded in one end ( do not ground both ends or the contraption becomes a transformer).
The wire when used as above creates a sheild for the capacitor.

Generally in design high impedance and high gain is asking for trouble and one can look at scaling down values to lower total impedance but as the Gain sits in the circuit this is not an option.

Have fun
Bjorn Juhl

Hi Bjorn,

Thanks for your clarifications. I was suspecting about the gain pot's position indeed, its something unusual seemingly, even when compared to other 4007-based layouts. I'll try to tinker around with it to see what can I do, but to be honest, I don't have much idea how to proceed  The schematic I've shared above is drawn by the layout provided on Parasite's web-site, and there's no other schematic around for that.

[BJF]

Hi there Baran,

Yes I see, the methodical approach is to redraw using the internal diagram of CD4007.

When a totem pair of MOS transistors, which could be a pair out of CD4007 or an inverter or a NAND gate with both inputs connected to each other , receives a feedback from output to input the result can be used as linear amplifier. Note that bandwidth and gain is dependent on supply voltage so gain is highest at 5V but so is bandwidth and gain falls rapidly above 1KHz. Gain is lowest at 15V but that gives widest bandwidth so
For audio 9V to12V is a nice balance.


However for hands on try disconnecting C3 from the junction of R1,R2 and pin 1 of CD4007 and put the loose end of the 2n2 capacitor to the viper of a 100K potentiometer and ground the terminal 1. Take a capacitor of say 0,022uF from junction R1,R2,pin1 of CD4007 to terminal 3 of the 100K potentiometer and you now have moved gain one stage in.

It would be most interesting to know how that works out

At your service
Bjorn Juhl

[Baran Ismen]

Hi there Baran,

Yes I see, the methodical approach is to redraw using the internal diagram of CD4007.

When a totem pair of MOS transistors, which could be a pair out of CD4007 or an inverter or a NAND gate with both inputs connected to each other , receives a feedback from output to input the result can be used as linear amplifier. Note that bandwidth and gain is dependent on supply voltage so gain is highest at 5V but so is bandwidth and gain falls rapidly above 1KHz. Gain is lowest at 15V but that gives widest bandwidth so
For audio 9V to12V is a nice balance.


However for hands on try disconnecting C3 from the junction of R1,R2 and pin 1 of CD4007 and put the loose end of the 2n2 capacitor to the viper of a 100K potentiometer and ground the terminal 1. Take a capacitor of say 0,022uF from junction R1,R2,pin1 of CD4007 to terminal 3 of the 100K potentiometer and you now have moved gain one stage in.

It would be most interesting to know how that works out

At your service
Bjorn Juhl

Hi Bjorn,

I'll give this a try this evening, and write the results, cheers man! But in that case, I'll be using a separate pot, right ? And feed the input via C1 directly to 3rd leg of CD4007 ?

Overall, I did love the sound of it, I was in search of a gnarly fuzz pedal layout, with which I can play some post-rock stuff, then with this project, I've not also found, but made it

The worst, I plan to put a rotary switch to connect various resistors in series, say 5, and use it as a gain divider. Sort of a different approach

[BJF]

Hi there Baran,

Yes I see, the methodical approach is to redraw using the internal diagram of CD4007.

When a totem pair of MOS transistors, which could be a pair out of CD4007 or an inverter or a NAND gate with both inputs connected to each other , receives a feedback from output to input the result can be used as linear amplifier. Note that bandwidth and gain is dependent on supply voltage so gain is highest at 5V but so is bandwidth and gain falls rapidly above 1KHz. Gain is lowest at 15V but that gives widest bandwidth so
For audio 9V to12V is a nice balance.


However for hands on try disconnecting C3 from the junction of R1,R2 and pin 1 of CD4007 and put the loose end of the 2n2 capacitor to the viper of a 100K potentiometer and ground the terminal 1. Take a capacitor of say 0,022uF from junction R1,R2,pin1 of CD4007 to terminal 3 of the 100K potentiometer and you now have moved gain one stage in.

It would be most interesting to know how that works out

At your service
Bjorn Juhl

[Baran Ismen]

Hi there Baran,

Yes I see, the methodical approach is to redraw using the internal diagram of CD4007.

When a totem pair of MOS transistors, which could be a pair out of CD4007 or an inverter or a NAND gate with both inputs connected to each other , receives a feedback from output to input the result can be used as linear amplifier. Note that bandwidth and gain is dependent on supply voltage so gain is highest at 5V but so is bandwidth and gain falls rapidly above 1KHz. Gain is lowest at 15V but that gives widest bandwidth so
For audio 9V to12V is a nice balance.


However for hands on try disconnecting C3 from the junction of R1,R2 and pin 1 of CD4007 and put the loose end of the 2n2 capacitor to the viper of a 100K potentiometer and ground the terminal 1. Take a capacitor of say 0,022uF from junction R1,R2,pin1 of CD4007 to terminal 3 of the 100K potentiometer and you now have moved gain one stage in.

It would be most interesting to know how that works out

At your service
Bjorn Juhl

This trick didnt help, it just helped with the gain, the overall distortion level is dropped a bit, but this hum is still present.

Note that, i just also noticed, guitars volume knob does the same thing as well. Its just fine at full level, but during the travel, theres a humming rising and fading away between two extremities.

[Baran Ismen]

I made some progress. It's much better now.

Drive sweep hum is still there but much much less than before. I think it'll be solved once everything is laid upon a PCB because the jumper cables I use are quite shitey, they play around so much and don't have a stiff grip on the holes of the breadboard. I think I'll get some proper jumper cables with various lengths, much more handy to use. Anyway;

- Added a 1000uf power filter, as it seemed to help with the noise pretty much. I know it's overdoing, but works well. I'll recheck this when I make the PCB with a decent ground network, though.

- Grounded the 3rd pin of the Drive pot via 10k resistor, all of its range is working now with much less humming. When directly grounded, it simply acts like a volume pot. It acts like a mild booster/od on full CCW, and growls like a demon on full CW positions.

- Moved the Drive's wiper to the chip's 1st leg.

- Changed the 2nd 470nf cap with a 1uf N.C. one, yet I have not heard any difference.

- Added a 1M input resistor, because why not? 

One rule I've read today -which I also violated here- is that no pins should be left floating on any CMOS chips, but in my case, pin 3 of the chip (Gate 2) became unused. Any ideas are most welcome.

[Rob Strand]

I think you mean coaxial cable right, like one for live, another naked wrapped around it, sort of regular mono instrument cables.

Yes, coax and coax cable are the mean the same thing.

[Rob Strand]

The detailed schematic is something like this,

[see later post for corrected schematic
https://www.diystompboxes.com/smfforum/index.php?topic=131720.msg1280634#msg1280634]

[Baran Ismen]

The detailed schematic is something like this,

Even tho I can see the schematic with the help of VPN, I see that it's neatly done and mine seems like a messy hairball when compared. Thanks Rob!

The current one I've done last night is a bit different, though. Pin 3 is floating and I could not find anywhere to stick it into

[duck_arse]

but pin 3 is the input of the first pair, how can your circuit work without it connecting? ahh, see, your first stage is not taking part. you must, MUST connect pin 3 to the wiper of the drive pot. no other thing connects to that wiper, just pin 3.

you might get even more demon soon.