Mammal

Hak

New member
Hello,
I have a trouble with the mammal PCB.
When the treble is 3 o'clock and more, the gain as well, there is a hight pitch noise, more when the snarl is on "more gain"
I check connection, resistance value ok, and i change capacitor C4,C5,C7,C9 by Wima and i try with a battery, but it is the same isue.
Thanks.
Best regards
 
Hi,
many people have this problem with this noise on few diy Rockett Animal.
For resolve this problem, I add a resistor 1M between the terminal of the Gain potentiometer (resistance 4.7k) and the 0V.
Now it's works correctly.
I don't found a difference sound.
Keep me informed ...
Best regards
 

Attachments

  • 20191013_112243b.jpg
    20191013_112243b.jpg
    176.8 KB · Views: 28
That's an interesting fix. I think I know what it does. Where did you get the idea to do that?

I have four other possible fixes:
1) Connect IC1-5 to Vref. Opamp inputs should not be left floating, weird things can happen. Some multiple opamp chips (don't know if JRC4558 is one of 'em) share the internal bias network between amplifiers. If one goes haywire, it can affect the other(s).
2) Increase the value of C3. C3 doesn't just roll-off the gain at high freq, it also stabilizes IC1's feedback loop. Try 47pF or even 100pF.
3) Try a different opamp part #. Whatever dual opamp you have.
4) connect a 2.2K resistor from IC1-1 to ground. Most opamp output stages are biased class B. There can be a dead spot where both push-pull output transistors are off or nearly off. Adding a resistor from the output to ground biases the output into class A and removes the dead spot.

Turning the treble all the way up loads the opamp with a 47nF capacitor. That by itself will make some opamps go unstable and is generally bad design practice. A few hundred ohms is all that's needed between the opamp and capacitor to eliminate that as a potential problem.
 
Last edited:
Hello,
I am an electronics engineer and I tried several solutions without modifying the original component of the Animal.
I will try your first solution and I will keep you informed.
Thanks you
 
That's an interesting fix. I think I know what it does. Where did you get the idea to do that?

I have four other possible fixes:
1) Connect IC1-5 to Vref. Opamp inputs should not be left floating, weird things can happen. Some multiple opamp chips (don't know if JRC4558 is one of 'em) share the internal bias network between amplifiers. If one goes haywire, it can affect the other(s).
2) Increase the value of C3. C3 doesn't just roll-off the gain at high freq, it also stabilizes IC1's feedback loop. Try 47pF or even 100pF.
3) Try a different opamp part #. Whatever dual opamp you have.
4) connect a 2.2K resistor from IC1-1 to ground. Most opamp output stages are biased class B. There can be a dead spot where both push-pull output transistors are off or nearly off. Adding a resistor from the output to ground biases the output into class A and removes the dead spot.

Turning the treble all the way up loads the opamp with a 47nF capacitor. That by itself will make some opamps go unstable and is generally bad design practice. A few hundred ohms is all that's needed between the opamp and capacitor to eliminate that as a potential problem.

Nice info Chuck. I can’t speak to all of these ideas, but have built this circuit a couple times on vero. The original trace doesn’t connect pins 5,6, and 7, but the original FSB thread recommends grounding pin5, and shorting 6-7. This is how I’ve built mine and the noise is still kind of high at high gains.

I remember I experimented with a higher feedback loop cap without improvement. And I built at least one with a ne5533 which I believe is generally considered to have less noise. As for the treble cap, the noise is there even when the tone knob isn’t maxed, but you’re right it’s not good design. If I ever build another one, I’ll try your other ideas.

FYI this is almost identical to the blue note circuit (only difference is a 50k gain pot and different filters in the hot sw mode). So, I don’t recall if some of what I read was on the blue note fsb thread.
 
Last edited:
In the end, I sort of made the assumption that the high noise compared to other distortion pedals was attributed to the clipping configuration. Animal uses two diodes each way in soft clipping, compared to, say, a rat or ocd using one diode each way in hard clipping. So to generate similar distortion sound, the animal’s op amp gain must be higher, which in turn elevates the noise. Not saying it couldn’t be improved with more tinkering though, but I wasn’t able to.

apparently rockett contracted Brian wampler to design the original topology for the blue note. It’s so simple that if rockett couldn’t figure it out themselves, it makes me doubt their engineering skill, and the design flaws shouldn’t be a surprise.
 
Back
Top