Crackling noise on cd4049 distortion..how i solved it..

[dschwartz]

Hi guys!!
I'm currently working on a new design and i found out the hard way about laying out a 4049 ic..

I'm designing a high gain distortion combining opamps and a couple of inverters.. It worked fine but very noisy on the breadboard..i proceeded to design the layout expecting a lower noise ratio (smd, very short routes)..

I built the prototype and got an awful crackling noise and low freq oscillation..first i thought it was a bad 4049..changed it, no fix..maybe additional filtering...no fix....
Then i dont know why i decided to cut the traces under the IC and use wires to move away from the IC..bingo!!

So..lesson learned...dont trace signal under the ic...

[armdnrdy]

I wonder if you would have experienced the noise with a through hole 4049...since the distance to the IC body and pins would be greater.

[Fender3D]

Did you tie to ground or Vcc unused inverters, didn't you? 

[dschwartz]

Did you tie to ground or Vcc unused inverters, didn't you? 

Yes, unused inputs to ground ..
I have to say that the hint i had was that on the breadboard, i had a resistor from one side to the other of the IC..if the resistor touched the chip, it would make weird noises, so i left it high above the ic..

Now im dealing with hiss as expected..i tried the stone grey trick of a 10pf global feedback..it kills the hiss, but also all dynamics, so maybe the old trick of parallell back to back shottky diodes would do the trick

[Kipper4]

That would be 100pf not 10pf Daniel right ?

[dschwartz]

The stone gray has a 10 pf cap going from the last inverter stage to the inverting input of the first stage (opamp) that for nulling high freq oscillations. It effectively removes any hiss, but it takes away a lot from the distortion dynamics, maybe because the feedback is compressed by the inverters, so it acts like an expander for certain freqs..didn't like it at all

[Kipper4]

Yep Gottya In only one of the several schematics I have. Victor Nerys does have a 15pf global feedback cap.
It might be way off but, What happens if you make the global feedback cap bigger?
I'm sure you have found the best solution but I still gotta ask some dumb questions sorry.......

[dschwartz]

You're right is a 15pF cap..if bigger, it would have more effect..i think smaller is better, i just had a 10pF cap, so that's as small i could get..

I have played so long with the design that I'm starting to feel it is not really good, maybe  i have to give it a rest to refresh my ears...

[anotherjim]

That global feedback cap kind of worries me. The overall path is 180deg inverting, but is there going to be additional phase shifts through the stages? That could be causing oscillations/instability. If setting the high end limit to cut hf noise, I would rather do it in each stage. But it already is cutting hf below guitar top end in each stage - except the op-amp at low gain. When distortion pot R is low (under 300k), 100pF isn't large enough.
What I would try, is split the 200k R5 feeding the first inverter into 2x 100k and ground the centre with 470p. Then the op-amp output always has similar top end cut as the other stages whatever gain it has. The 100p feedback on the op-amp could be reduced - 47p gives a similar top end cut at maximum gain as the other stages, but the existing 100p might sound better overall.

[dschwartz]

Totally agree with you..i ditched the global feedback and went with the usual high freq cut on each stage.. I'm  not building a clone, but by coincidence the architecture is similar to the stone gray..i have an additional input opamp stage for pre- emphasis and additional inverter stages for post active filtering..6 active stages in total..cutting highs on every stage, it is turning to be a really mid focused distortion, very 80's