I’m making a noise circuit in which I produce “digital” noise by comparing (through an opamp acting like a comparator) white noise with a steady voltage. Whenever the noise voltage exceeds the steady voltage, I get an instant 12V peak and if I lower the steady voltage the peaks appear more often. This is an iteration from MFOS noise cornucopia.
Now, I want to cv control the voltage by which the noise gets compared.
For that I used the cv in part from the schematic of grids.
The circuit works perfectly as expected but I’m using the 5V regulator and the zenner for the -5V only for that (I really cannot figure out another thing that I can cv control in my noise module).
The whole module uses TL072s and the ±5V parts use too much board space so I was wondering:
Can I produce the same effect using a TL072, as I don’t care about precision (I can adjust the noise level to be compared)? Could I do something like that? (The first opamp doubles the incoming CV voltage and the second it is supposed to perform the same function as the grids cv in schematic but I was told that since the opamp is powered by 0-12V, any negative voltage will fry the chip).
I could use the single rail mcp6002 no problem but are the filter(?) capacitors on the LM1117 are necessary (these are taking the most space on my board)?
Could I replace them with ceramic capacitors, what do they do exactly?
The expected outcome is a voltage between 0 and 5V to the comparators input with the knob determining the center of the voltage (I mean if you set it at 12 o’clock and give it a 0V cv in, the output of the opamp should be @ 2.5V, if you send it a negative voltage at the cv in, it should be lower until reaching 0V, and positive voltages should be incrementing the 2.5V base etc).
Now, precision is not really important in this application.
Christmas tip #1: you are not designing a microcontroller-based module (with a single supply and a relatively low impedance ADC input), so whatever was done for Grids or other digital modules is irrelevant to your problem.
Christmas tip #2: you are not designing an analog module with 2164s and their very low impedance control pins, so whatever was done for Ripples or Veils is also irrelevant to your problem. You are actually dealing with one of the simplest cases: an op-amp input pin, which has a very large input impedance.
Christmas tip #3 (an assortment): some important elements that I don’t see in your circuit (but which are present in the MFOS original): an AC coupling cap somewhere after RAW_NOISE (you want your noise to be centered), a 10pF stability cap in the feedback path of your comparator (you want stable edges), and a low-pass filter on the noise signal sent to the comparator (you probably don’t want HF gates above the audio range).
Christmas tip #4 (your actual question): 3 resistors solve your problem. CV is your CV input (and it should stay grounded even when no patch cable is inserted), POT goes between -12V and +12V (this gives a symmetric control range, from always off, to 50%/50%, to always on). The ratio of the large resistors is 2.4, so an increment of CVs by 5V is equivalent of a turn of the pot voltage by 12V. The resulting voltage COMP_V is conveniently between -1V and +1V, which means that you might not need to amplify the white noise on the other pin of the comparator – this could actually require the opposite, a voltage divider. If you carefully select R2 so that the level of the noise and the range of the threshold match, you won’t need the “amplifier for digital noise” block.
This is still a guess since you did not answer my question about the expected behavior! If you want a unipolar control (output is always off on one side, and 50%/50% on the other side – for example if you intend to keep Ray Wilson’s CD4024 idea, which is good!), POT can simply go between 0V and +12V. And you can also swap what is hooked to the + and to the - pins of the comparator to invert the control scheme.
I didn’t understand what you mean by the expected behavior, this is a video of what I use the cv in for.
The digital noise output should be only positive (I don’t use the negative part of the noise), I take grains only when noise level is above COMP_V (so, I don’t know how to manipulate both positive and negative comp_v values).
This configuration, taken from Thomas Henry Noise Cookbook, produces a dc coupled output (I checked it on the oscilloscope, and it sits symmetrically around 0V.
Is there a feedback path on my comparator???
I saw that with different sets of transistors (actual parts, not types) you get differences on the output levels, and that’s why I have a second amplifier, maybe I could get rid of that somehow, but I really wanted to have control between extra sparse grains and full on avalanche.
See C16 in the MFOS circuit. You can also use a large resistor to add some hysteresis.
I mean matching in terms of order of magnitude… I don’t know what is the order of magnitude of RAW_NOISE, if it’s around 2Vpp, you won’t need an amplifier since the threshold voltage will be in that same order of magnitude (with the resistor values I have suggested). If RAW_NOISE is around 0.1Vpp, it’s easier to simply decrease R2 to get a threshold voltage in that range.
If you want to make things trimmable, you can either put the trimmer in series with R2 to precisely trim the comparison threshold, or as part of a divider that adjusts the level of RAW_NOISE applied to the other pin of the comparator.
All I want to say is that it’s easier to use dividers and compare two small numbers, instead of using an op-amp just for the sake of comparing large numbers.