Hiss/white Noise in in Zombie Chorus MN3207

[percyhornickel]

Hi, time ago I wrote about about my ZC and how I've got rid of thicking/distortion problems to cero. The problem I have is with the hiss when I set the switch mode from 1n to 2n (even with 1n the hiss is there).

Testing the circuit I've desoldered the 4046 VCO part (PINS 6-7) and instaled pin headers to try another cap values, my surprise was with no caps attached, nice sound almost like CE-2 with cero hiss (NOTHING).

I soldered a three position switch in a way I can select  330p / - / 1n.

With 2.2n cap selected the hiss is too strong, worse if I select the vibe switch. With no caps the "small detuning" is very very small.

By now my default setting is with no caps for VCO unless somebody could give me a trace about the amount of hiss I am getting in here, I've really would appreciate that.

I hope I explained myself well, my english is not that good.

My build:

BBD MN3207
CLOCK 4046 (with pins 14/16 attached to zener pin 15)
OSC 4558
AUD 4559

100 ohm/47u to isolate the oscilator 4558.
Cable from the 3PDT switch (metal part) to output ground.


Saludos from Venezuela

Percy

[Mark Hammer]

Your English will always be better than my Spanish.

The Zombie has very little lowpass filtering, compared to other chorus units.  That's good, because it gives a nice crisp sound.  But using a higher-value capacitor to change the delay range makes the clock frequency lower, and brings the clock pulse into the audio range.  And the little lowpass filtering means that HF clock pulse will be easy to hear.

One thing you can do is this.  There is a resistor on the output of the op-amp that does the filtering.  That 10k resistor mixes delayed signal with the dry signal coming on the other 10k resistor.  Replace that 10k resistor with two resistors that add up to around 10k in series (e.g., 5k1 and 4k7).  You can now run a small-value capacitor to ground from the junction/middle of those two resistors.  Use a SPDT on-off-on toggle to add a little bit of capacitance or more capacitance to remove hiss.  When that toggle is off, you go back to the standard filtering the Zombie provides.

With 5k1, a 10nf capacitor will provide a 6db/oct rolloff around 3100hz.  A 15nf cap will do the same, starting just over 2khz.

[percyhornickel]

Mark I think I could try this in my board, I have modified the layout I am showing in here but essentially is the same. I could replace the 10K resistor at the bottom for 5.1K + 4.7K and connect the cap to ground, I don't know if is a good idea to put another couple of wires for the cap selector switch because my board looks like a power transformer instead a circuit board because the quantity of switches, pots, bottoms, etc etc...    ..maybe I am generating a tesla bobine instead a zombie chorus  !!

I hope to have time to try your solution today/tonight, tomorrow I have to go back to work until the next week.

I'll be posting the results as soon as I can, and thank you for the answer.   

By now I just put a 330p/470p on the pin headers (my las test) and is not much different, maybe 680p or 820p could work out to get a little less hiss, so..  .. 330p / - / 470p my mode switch right now.

BTW, I have build the CE-2 and small clone already and they both sounds great, but I am kind of obsessed to make my board to work great!.

Saludos


Percy

[percyhornickel]

Hi, Mark finally I did the trick with 5.1K + 4.7K + 10n cap to improve the filtering / hiss. I was searching what kind of topology (1/2/3 order??) are related with that advice you wrote which worked pretty well with the sound, not completelly with hiss but sounds pretty well.

For testing I installed a separated caps (180p and 470p) for the 4046 VCO (pins) and the hiss is very very low, with 180p sounds like CE-2 (tested side by side) with a little less headroom (I think is b/c I´m using a 3207 instead 3007 BBD in this build). I think I´m gonna keep this way. Using 1n cap the hiss is bigger so for my build is not that usefull.

For thicking, I cover both sides of aluminum tape with plastic tape, open a small hole and completelly cover the rate pot (with solder points even) = no thicking at all, better covering input/output jacks this way too and a small cable from the metal part of the 3pdt switch to output ground.

I am pretty happy with the sound of this build right now (3207 version), I have another 3207 version already in eagle. I hope to find the time later to eatch the board and solder the components.

Percy

[Mark Hammer]

Excellent!